2013-05-01 06:54:29 <build64-2-4> failed to build v2.4.10-42-g66847be. http://distfiles.alpinelinux.org/distfiles/build64-2-4.txt
2013-05-01 06:57:06 <ncopa> build64-2-4: retry
2013-05-01 07:57:30 <build64-2-4> files from v2.4.10-42-g66847be uploaded
2013-05-01 09:04:12 <ncopa> build-2-4: retry
2013-05-01 10:11:26 <build-2-4> files from v2.4.10-42-g66847be uploaded
2013-05-01 10:29:28 <ncopa> here comes rc1
2013-05-01 11:03:27 <ncopa> algitbot: build master
2013-05-01 11:03:59 <ncopa> algitbot: build master
2013-05-01 11:06:05 <ncopa> algitbot: build master
2013-05-01 12:09:55 <ncopa> so rc1 is out
2013-05-01 12:10:02 <ncopa> i wonder what to do with alpine-vserver
2013-05-01 12:10:07 <ncopa> the kernel lags behind
2013-05-01 12:10:26 <rnalrd> didn't we ship lxc with 2.6?
2013-05-01 12:10:41 <ncopa> we do
2013-05-01 12:11:15 <ncopa> oh we should probably kill the alpine-scst iso too
2013-05-01 12:11:24 <rnalrd> yes
2013-05-01 12:11:51 <ncopa> but i would rather not kill vserver yet
2013-05-01 12:12:06 <ncopa> maybe we could go back to 3.4 kernel for vserver
2013-05-01 12:12:11 <ncopa> which it lts
2013-05-01 12:20:16 <rnalrd> sounds good
2013-05-01 12:29:17 <Barthalion> ltsi
2013-05-01 12:36:26 <ncopa> ltsi is still at 3.4.25
2013-05-01 12:36:33 <ncopa> they lag behind?
2013-05-01 12:36:52 <ncopa> i also wonader what stuff tehy backport
2013-05-01 13:09:54 <Barthalion> Afaik embedded, and all patches are sooner or later merged with upstream 3.4
2013-05-01 17:02:33 <hl> ncopa: the mkinitfs patch should have minimal impact on existing boot scenarios so rc2 testing should be okay
2013-05-02 04:45:54 <Mp5shooter> hi guys
2013-05-02 04:46:00 <Mp5shooter> is whoever packaged openswan in here
2013-05-02 04:46:02 <Mp5shooter> ?
2013-05-02 06:05:58 <ncopa> Mp5shooter: thats probably me
2013-05-02 06:06:35 <ncopa> hm
2013-05-02 06:07:07 <ncopa> no, looks like it was Danielo initially
2013-05-02 11:38:52 <ncopa> hl: question re nfs and pxe
2013-05-02 11:39:38 <ncopa> i wonder if we could use ALPINE_MNT=/media/${ALPINE_DEV##*/} for nfs too
2013-05-02 11:49:27 <ncopa> or if we shoud mount it as /nfs/<host>/
2013-05-02 12:04:59 <StarWarsFan> hi
2013-05-02 12:05:05 <ncopa> hi
2013-05-02 12:05:28 <StarWarsFan> ncopa: for http://lists.alpinelinux.org/alpine-devel/2475.html it seems that uclibc has to be build with UCLIBC_HAS_LOCALE='yes'
2013-05-02 12:05:40 <ncopa> hm
2013-05-02 12:05:48 <StarWarsFan> but then every package has to be recompiled
2013-05-02 12:05:56 <ncopa> yeah :-/
2013-05-02 12:06:05 <ncopa> well
2013-05-02 12:06:07 <ncopa> does it?
2013-05-02 12:06:10 <ncopa> will it break abi?
2013-05-02 12:06:15 <StarWarsFan>           uClibc now has full ANSI/ISO C99 locale support (except for
2013-05-02 12:06:15 <StarWarsFan>           wcsftime() and collating items in regex).  Be aware that enabling
2013-05-02 12:06:15 <StarWarsFan>           (169 UTF-8 locales, and 144 locales for other codesets) will enlarge
2013-05-02 12:06:15 <StarWarsFan>           this option will make uClibc much larger.
2013-05-02 12:06:15 <StarWarsFan>           Enabling UCLIBC_HAS_LOCALE with the default set of supported locales
2013-05-02 12:06:17 <StarWarsFan>           uClibc by around 300k.  You can reduce this size by building your own
2013-05-02 12:06:19 <StarWarsFan>           custom set of locate data (see extra/locale/LOCALES for details).
2013-05-02 12:06:47 <ncopa> will it break libiconv?
2013-05-02 12:07:01 <ncopa> i suppose we can still install libiconv if we want
2013-05-02 12:07:22 <StarWarsFan> i'm not that expert on that...
2013-05-02 12:07:28 <StarWarsFan> but seems to be worth a try...
2013-05-02 12:07:39 <ncopa> last time i tried it missed some locale data files
2013-05-02 12:07:44 <ncopa> which i could not find any place
2013-05-02 12:07:49 <StarWarsFan> i c
2013-05-02 12:07:52 <ncopa> that was some time ago
2013-05-02 12:11:31 <StarWarsFan> what do you think, is the size of the resulting uclibc a go/nogo?
2013-05-02 12:11:47 <StarWarsFan> i think, proper utf-8 support is a must have...
2013-05-02 12:19:35 <ncopa> depends on how big the difference is
2013-05-02 12:19:50 <ncopa> i suppose it would still be lot smaller than glibc
2013-05-02 12:20:23 <ncopa> and if the locale data files could be installed as a separate package, then its probably ok
2013-05-02 12:29:28 <StarWarsFan> ack
2013-05-02 15:50:00 <royger> ncopa: are you around?
2013-05-02 15:50:33 <royger> re the PXE change that you committed, does this mean we will start distributing the alpine installer as a kernel+initram also?
2013-05-02 15:55:32 <hl> ncopa: currently nfs alpine_dev goes to /media/alpine_dev. I would think adding an alpine_mnt override would be pretty simple.
2013-05-02 21:55:39 <kaniini> yes, we will offer PXE media, but not for 2.6
2013-05-03 06:02:53 <ncopa_> morning
2013-05-03 06:03:38 <ncopa> hl: i was mor thinking of replacing /media/alpine_dev with /media/<nfshost-or-ip-addr>/
2013-05-03 06:22:05 <ncopa> rnalrd: what do you think about syncing our ucarp init script with gentoo?
2013-05-03 06:22:08 <ncopa> http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/net-misc/ucarp/files/
2013-05-03 06:22:19 <ncopa> the .initd and .confd
2013-05-03 06:24:47 <rnalrd> ncopa, move also the up/down scripts location?
2013-05-03 06:27:01 <ncopa> where to?
2013-05-03 06:27:09 <ncopa> you mean those:
2013-05-03 06:27:10 <ncopa> # Specifies the command to run after ucarp has successfully become master.
2013-05-03 06:27:11 <ncopa> UCARP_UPSCRIPT=/usr/libexec/ucarp/vip-up-default.sh
2013-05-03 06:27:17 <rnalrd> yeah
2013-05-03 06:27:28 <rnalrd> we have them in /etc/ucarp iirc
2013-05-03 06:27:36 <ncopa> ok
2013-05-03 06:27:44 <ncopa> well, yes, i think that makes more sense
2013-05-03 06:27:49 <rnalrd> since they are customized
2013-05-03 06:27:54 <ncopa> the scripts are config
2013-05-03 06:27:58 <rnalrd> we want them into LBU by default
2013-05-03 06:28:01 <ncopa> yes
2013-05-03 06:28:17 <ncopa> looks good otherwise?
2013-05-03 06:28:28 <rnalrd> yup
2013-05-03 06:28:35 <ncopa> thank!
2013-05-03 06:28:39 <ncopa> i'm glad i asked :)
2013-05-03 06:28:53 <rnalrd> yw
2013-05-03 11:15:17 <build-2-2> files from v2.2.3-226-ga564f0e uploaded
2013-05-03 11:15:43 <build64-2-2> files from v2.2.3-226-ga564f0e uploaded
2013-05-03 12:19:22 <build-2-3> files from v2.3.6-201-g622b89d uploaded
2013-05-03 12:28:38 <build64-2-4> files from v2.4.10-44-g97558aa uploaded
2013-05-03 12:28:53 <build-2-4> files from v2.4.10-44-g97558aa uploaded
2013-05-03 12:38:24 <build64-2-5> failed to build v2.5.4-59-g508b920. http://distfiles.alpinelinux.org/distfiles/build64-2-5.txt
2013-05-03 12:45:54 <build-2-5> files from v2.5.4-60-ga396023 uploaded
2013-05-03 12:48:44 <StarWarsFan> ncopa: on the utf-8 problem we talked yesterday
2013-05-03 12:48:56 <StarWarsFan> it seems to be a problem with the core-utils
2013-05-03 12:49:10 <ncopa> aha
2013-05-03 12:49:11 <StarWarsFan> they do not support utf-8 at the moment
2013-05-03 12:49:37 <StarWarsFan> because the 'ls' of busybox works correct
2013-05-03 12:49:54 <StarWarsFan> we're digging into this at the moment... ;-)
2013-05-03 12:50:07 <ncopa> heh
2013-05-03 12:50:19 <ncopa> some things works in xfce4-terminal
2013-05-03 12:50:24 <ncopa> i can type ø
2013-05-03 12:50:40 <ncopa> but it behaves strange in tmux
2013-05-03 12:50:53 <ncopa> the char is printed, but cursor jumps 2 chars ahead
2013-05-03 12:50:59 <ncopa> instaed of the expected 1
2013-05-03 12:51:25 <ncopa> $ echo $CHARSET
2013-05-03 12:51:25 <ncopa> UTF-8
2013-05-03 12:51:41 <ncopa> i dunno if that envvar matters
2013-05-03 12:52:11 <StarWarsFan> ok, i will let you know how it goes ahead
2013-05-03 13:34:57 <ncopa> who was about to look at phpmyadmin?
2013-05-03 13:35:12 <ncopa> seems like phpmyadmin-4.0.0 is out
2013-05-03 13:35:18 <Barthalion> 
ACTION 

2013-05-03 13:35:35 <Barthalion> 4.0.0 was released, I'll have time tommorow to update it
2013-05-03 13:35:41 <ncopa> super
2013-05-03 13:35:53 <ncopa> feel free to take over maintainership there
2013-05-03 14:06:28 <ncopa> goidn home now
2013-05-03 14:06:32 <ncopa> have a nice weekend
2013-05-03 18:55:02 <vkrishn> can something like grml-debootstrap be adapted for alpine
2013-05-03 18:55:23 <vkrishn> then new alpine can installed on the specified device or directory, or directly into an image file suitable for Virtual Machine use
2013-05-03 18:55:30 <vkrishn> :)
2013-05-03 18:55:50 <vkrishn> ?
2013-05-03 18:55:51 <kaniini> i wrote something for that called appliancekit
2013-05-03 18:55:59 <kaniini> apk add py-appliancekit-ng@testing
2013-05-03 18:56:14 <vkrishn> requires py
2013-05-03 19:11:29 <vkrishn> what does it mean when it says, `ApplianceKit has become unmaintainable`
2013-05-03 19:11:39 <vkrishn> https://bitbucket.org/tortoiselabs/appliancekit-ng
2013-05-03 19:13:51 <vkrishn> ah.. appliancekit-ng
2013-05-03 19:32:51 <vkrishn> looks like deps would be much less, but don't know about features
2013-05-03 19:51:02 <kaniini> appliancekit-ng is effectively an AST engine
2013-05-03 19:51:08 <kaniini> it could be written in any language easily
2013-05-03 21:36:19 <vkrishn> I guess appliancekit-ng flexibility lies in xml templating
2013-05-03 21:36:39 <vkrishn> could not get to https://bitbucket.org/tortoiselabs/ADL.md
2013-05-04 23:55:10 <kaniini> hi
2013-05-04 23:55:41 <kaniini> ncopa: i have built a prototype migration tool for vservers -> lxc
2013-05-04 23:56:27 <kaniini> (this might be something we can collaborate with debian on, actually)
2013-05-05 08:12:18 <fcolista> kernel 3.8.11-0-grsec segfaults on my laptop
2013-05-05 08:12:29 <fcolista> __chech_object_size
2013-05-05 08:13:06 <fcolista> PAX: kernel memory leak attempt detected from $memory_address
2013-05-05 20:41:55 <Mp5shooter> are there any xmpp servers packaged?
2013-05-05 20:42:52 <Barthalion> There is prosody in testing
2013-05-06 01:33:32 <Dave1029> Do we have a place for bug reports?  I'd like to help in testing with regards to Xen, and I believe I have found some things that might be bugs...http://bugs.alpinelinux.org ?  Is that it?
2013-05-06 01:47:53 <Mp5shooter> Dave1029: bugs.alpinelinux.org
2013-05-06 01:48:00 <Mp5shooter> yes
2013-05-06 01:48:01 <Mp5shooter> :P
2013-05-06 01:48:15 <Mp5shooter> hey Barthalion, there's a leftover Makefile in /etc/prosody/certs
2013-05-06 01:48:17 <Mp5shooter> for prosody
2013-05-06 01:49:35 <Mp5shooter> also default private key is world readable :x
2013-05-06 02:44:45 <Mp5shooter> ugh
2013-05-06 02:44:48 <Mp5shooter> prosody is being annoying
2013-05-06 02:44:58 <Mp5shooter> I told it to listen only on one ip but it still listens on 0.0.0.
2013-05-06 02:45:00 <Mp5shooter> 0.0.0.0
2013-05-06 07:01:06 <Barthalion> Mp5shooter: Well, it's in testing!
2013-05-06 07:01:16 <Barthalion> Dave1029: Yes, it's corrext place
2013-05-06 07:01:22 <Barthalion> Mp5shooter: Patches welcome :p
2013-05-06 08:39:22 <clandmeter> ncopa: in order to have a boot from ram system based on linux-vanilla one should always build its own iso?
2013-05-06 08:39:59 <ncopa> so far yes
2013-05-06 09:17:05 <royger> ncopa: I plan to test the Xen iso this afternoon, I've got all my test boxes busy right now :(
2013-05-06 09:17:14 <royger> sorry for not realizing the change in syslinux
2013-05-06 09:22:02 <ncopa> i think it was my fault
2013-05-06 09:22:04 <ncopa> or
2013-05-06 09:22:10 <ncopa> it doesnt matter whos fault it was :)
2013-05-06 09:22:21 <StarWarsFan> hi
2013-05-06 09:22:36 <ncopa> hi StarWarsFan
2013-05-06 09:23:07 <ncopa> royger: i did a quick fix in the Makefile, i think it is ok but i havent test booted it
2013-05-06 09:23:14 <ncopa> not sure if it will boot in qemu
2013-05-06 09:23:37 <StarWarsFan> is it possible that the 32bit-iso from the webpage could not be used for virtualization on xen?
2013-05-06 09:23:41 <royger> ncopa: I think it should, with no virt-extensions detected, but it should be able to boot fine
2013-05-06 09:24:03 <ncopa> ok, I'll verify that it boots here then
2013-05-06 09:24:14 <ncopa> this is the makefile change: http://sprunge.us/haZL
2013-05-06 09:24:18 <royger> I remember I was able to boot it on virtualbox
2013-05-06 09:24:30 <royger> never tried qemu
2013-05-06 09:25:03 <ncopa> StarWarsFan: i think it should be possible to use it under xen...
2013-05-06 09:25:04 <ncopa> hm
2013-05-06 09:25:11 <ncopa> or maybe only 64bit works?
2013-05-06 09:25:20 <ncopa> due to missing PAE in 32 bit kernels
2013-05-06 09:25:34 <royger> it won't boot without PAE
2013-05-06 09:25:42 <ncopa> ah, yes
2013-05-06 09:25:56 <ncopa> so only x86_64 iso will boot in xen
2013-05-06 09:26:06 <StarWarsFan> :-(
2013-05-06 09:26:09 <royger> I think we alreary had this discussion, why is PAE not enabled in 32bit?
2013-05-06 09:26:21 <ncopa> we have (had?) some older boards
2013-05-06 09:26:26 <ncopa> that we want support
2013-05-06 09:26:34 <ncopa> i think the reasoning was something like
2013-05-06 09:26:46 <ncopa> if you have that kind of old hardware,what distro would you pick?
2013-05-06 09:26:48 <royger> and PAE breaks the support for those boards?
2013-05-06 09:26:52 <ncopa> yes
2013-05-06 09:27:06 <ncopa> could be alix boards
2013-05-06 09:27:41 <ncopa> i know there was other board that broke, but i forgot the name of it
2013-05-06 09:27:54 <ncopa> what i have been thinking recently is
2013-05-06 09:28:06 <ncopa> that we should probably enable PAE, and require i586 or something
2013-05-06 09:28:13 <royger> what does Debian or other diestros do?
2013-05-06 09:28:18 <ncopa> and then have an mini kernel
2013-05-06 09:28:23 <ncopa> i think they have multiple kernels
2013-05-06 09:28:46 <StarWarsFan> in my case I would like to setup some xen-base build-vm's and for this I need both versions, 32 and 64 bit
2013-05-06 09:29:10 <royger> you could add PAE to the i386 config file and recompile the kernel and the isos
2013-05-06 09:29:10 <ncopa> i think we could have an mini/nano kernel for the older boards
2013-05-06 09:29:20 <ncopa> we could strip off lots of kernel configs
2013-05-06 09:29:27 <ncopa> lomem
2013-05-06 09:29:31 <ncopa> no smp
2013-05-06 09:29:33 <ncopa> etc etc
2013-05-06 09:29:37 <ncopa> no pae
2013-05-06 09:29:49 <ncopa> then we could actually ship both on the iso
2013-05-06 09:29:57 <ncopa> and let syslinux detect if pae is there
2013-05-06 09:30:06 <StarWarsFan> sounds like a plan ;-)
2013-05-06 09:30:09 <ncopa> and automatically select the right kernel
2013-05-06 09:30:12 <royger> that might be interesting, 32bit PV guests are faster than amd64 on Xen
2013-05-06 09:30:19 <StarWarsFan> oh, brb, lunchtime...
2013-05-06 09:30:51 <ncopa> i dont know we we want the nopae kernel with grsec
2013-05-06 09:31:07 <ncopa> if we drop grsec there, we save us for lots of work
2013-05-06 09:31:11 <ncopa> or some work at least
2013-05-06 09:31:23 <ncopa> the vanilla kernel is easy peasy to maintain
2013-05-06 09:31:42 <ncopa> we would probably also not have any 3rd party mods for the nopae kernel flavor
2013-05-06 09:31:44 <ncopa> or very few
2013-05-06 09:32:24 <royger> but you will still need grsec for the pae kernel, how is having the nopae kernel without grsec going to save time?
2013-05-06 09:32:25 <ncopa> the main reason for not having such kernel atm, is that it is a big job to configure the kernel
2013-05-06 09:33:05 <ncopa> when i upgrade the current linux-vanilla kernel, i do: abump linux-vanilla-<newversion>
2013-05-06 09:33:07 <ncopa> and thats it
2013-05-06 09:33:17 <ncopa> i spend 10 sec on updating it
2013-05-06 09:33:24 <ncopa> grsec kernel, requires manualy work
2013-05-06 09:33:32 <ncopa> download the patch
2013-05-06 09:33:39 <ncopa> check that it applies
2013-05-06 09:33:47 <ncopa> sometimes there are new config options
2013-05-06 09:33:59 <royger> but the only difference between nopae, pae and amd64 is the kernel config right?
2013-05-06 09:34:20 <royger> you will still need to apply the patch for amd64 and pae i386
2013-05-06 09:34:29 <ncopa> yes
2013-05-06 09:34:29 <ncopa> hm
2013-05-06 09:34:47 <ncopa> we could maybe build the nopae in same apkbuild then
2013-05-06 09:34:48 <ncopa> hm
2013-05-06 09:35:22 <royger> that was what I was thinking of, just add a new config file, but dunno how that will play with the build system
2013-05-06 09:35:27 <royger> because is like adding a new arch
2013-05-06 09:35:35 <ncopa> yes
2013-05-06 09:36:01 <ncopa> i think thats also what other distros do
2013-05-06 09:36:06 <ncopa> from same build scripts
2013-05-06 09:36:10 <ncopa> makes sense
2013-05-06 09:36:28 <royger> yes, I guess it's the option with less overhead and duplication
2013-05-06 09:36:50 <ncopa> the only drawback with that approach is that it slows down when i'm working on the kernel
2013-05-06 09:36:53 <ncopa> on the apkbuild
2013-05-06 09:37:01 <ncopa> if need to rebuild and rebuild
2013-05-06 09:37:04 <royger> also, I would liek to ask you about the recent PXE kernel, is alpine planning to distribute a installer with kernel + initramfs only and no iso?
2013-05-06 09:37:19 <ncopa> then it builds the kernel twice...
2013-05-06 09:37:31 <royger> well, yes, there will be 3 arches now, but for your personal buildbox you could set up ccache
2013-05-06 09:37:39 <ncopa> i think we want pxe installer at some point
2013-05-06 09:37:47 <ncopa> but i dont think we will make it for v2.6
2013-05-06 09:38:32 <ncopa> but with v2.6 you should be able to set up your tftp server and your initramfs image
2013-05-06 09:38:33 <royger> mmm, ok, it will be really nice, so I don't have to extract the kernel + initramfs and attach the iso to the DomU every time I want to setup a Xen DomU
2013-05-06 09:38:39 <ncopa> so you could have an iso less install
2013-05-06 09:38:59 <ncopa> speaking ov kernels
2013-05-06 09:39:08 <ncopa> i would *really* like to use the 3.9 kernel
2013-05-06 09:39:14 <ncopa> the 3.8 is EOL
2013-05-06 09:39:24 <ncopa> but no grsecurity patch for 3.9 yet
2013-05-06 09:40:04 <royger> ahh, 3.9 has been released already
2013-05-06 09:40:11 <royger> is 3.9 LTS?
2013-05-06 09:41:02 <ncopa> i think nobody knows
2013-05-06 09:41:12 <ncopa> what we know is that 3.8 is *not* LTS
2013-05-06 09:41:29 <ncopa> and yes, i think we want aim for LTS kernels in future
2013-05-06 09:41:32 <ncopa> and try stick to those
2013-05-06 09:41:43 <royger> last LTS is 3.4?
2013-05-06 09:41:55 <ncopa> yes
2013-05-06 09:42:03 <ncopa> localhost:~# uname -a
2013-05-06 09:42:03 <ncopa> Linux localhost 3.3.8-grsec #2-Alpine SMP Wed Jun 20 15:18:49 UTC 2012 x86_64 Li
2013-05-06 09:42:03 <ncopa> nux
2013-05-06 09:42:04 <ncopa> hm
2013-05-06 09:42:10 <ncopa> 3.3.8?
2013-05-06 09:42:38 <royger> ?
2013-05-06 09:43:04 <ncopa> that was the xen image i booted
2013-05-06 09:44:13 <ncopa> ok
2013-05-06 09:44:16 <ncopa> local issue
2013-05-06 09:44:46 <royger> wrong aports?
2013-05-06 09:45:27 <ncopa> must have done something wrong when building the iso
2013-05-06 09:45:29 <ncopa> rebuidling it now
2013-05-06 09:45:50 <ncopa> btw
2013-05-06 09:46:01 <royger> probably the syslinux version in that iso was also wrong?
2013-05-06 09:46:07 <ncopa> i wonder if it would be worth having a virt-pxe initramfs image
2013-05-06 09:46:33 <ncopa> with the ehternet drivers to xen guests
2013-05-06 09:46:43 <royger> that's what I would aim for, but we need to pack busybox and install scripts inside the initramfs
2013-05-06 09:47:17 <ncopa> hm
2013-05-06 09:47:23 <ncopa> i think we also want the nomodeset
2013-05-06 09:47:26 <ncopa> for xen
2013-05-06 09:50:38 <ncopa> i dont think we need the install scripts in the initramfs
2013-05-06 09:50:56 <ncopa> they can be installed with alpine_repo=http://....
2013-05-06 09:51:00 <ncopa> we have ip addr early
2013-05-06 09:51:04 <ncopa> and wget with busybox
2013-05-06 09:51:18 <ncopa> so we can download the installscript from http
2013-05-06 09:51:32 <ncopa> thats what the current initramfs does
2013-05-06 09:51:55 <royger> well, current initramfs gets them from the CD, right?
2013-05-06 09:52:21 <ncopa> with rc2 it can also get them from http
2013-05-06 09:52:21 <ncopa> or nfs
2013-05-06 09:52:33 <ncopa> given you build your initramfs image with the needed thernet drivers
2013-05-06 09:52:43 <ncopa> ethernet*
2013-05-06 09:52:53 <ncopa> hl's PXE work
2013-05-06 09:53:19 <ncopa> while here
2013-05-06 09:53:26 <ncopa> do we want xen boot be so noisy?
2013-05-06 09:53:39 <royger> could be provide a initramfs with common drivers? like bnx2, tigon, e1000...?
2013-05-06 09:53:48 <ncopa> should we add 'quiet' boot opt by default?
2013-05-06 09:53:59 <ncopa> royger: thats what i'm thinking
2013-05-06 09:54:07 <royger> well, I generally prefer it so users can at least post the trace
2013-05-06 09:54:25 <royger> but it's completelly optional
2013-05-06 09:54:35 <ncopa> its a question about adding 'noquiet' at the boot prompt
2013-05-06 09:55:54 <ncopa> its kinda nice to make it visible that this is a xen hypervisor
2013-05-06 09:55:56 <ncopa> and dom0
2013-05-06 09:56:06 <royger> I tend to think alpine users are not novices, so they like to see the verbose boot dmesg
2013-05-06 09:57:09 <ncopa> np, i'll keep it like that
2013-05-06 09:57:38 <ncopa> localhost:~# uname -a
2013-05-06 09:57:38 <ncopa> Linux localhost 3.8.11-1-grsec #2-Alpine SMP Thu May 2 15:21:35 UTC 2013 x86_64
2013-05-06 09:57:38 <ncopa> Linux
2013-05-06 09:58:05 <ncopa> localhost:~# rc --sys
2013-05-06 09:58:05 <ncopa> XENU
2013-05-06 09:58:52 <royger> nice :)
2013-05-06 09:59:30 <ncopa> alpxen:~# rc --sys
2013-05-06 09:59:30 <ncopa> XEN0
2013-05-06 09:59:38 <ncopa> thats after running setup-alpine
2013-05-06 09:59:57 <ncopa> i think it mounts /proc/xen
2013-05-06 09:59:59 <vkrishn> I came across gcc hardening flags, would this useful for nano kernels with grsec off?
2013-05-06 10:00:20 <ncopa> dunno
2013-05-06 10:00:21 <vkrishn> or they already being used
2013-05-06 10:00:24 <royger> yes, I guess it detects the privileges after mounting /proc/xen
2013-05-06 10:00:42 <ncopa> vkrishn: we have some gcc hardeneing enabled by default
2013-05-06 10:00:48 <vkrishn> ok
2013-05-06 10:00:53 <ncopa> but for kernel some parts are disabled
2013-05-06 10:01:01 <clandmeter> ncopa: any reason intel liva driver is still in testing?
2013-05-06 10:01:09 <clandmeter> libva
2013-05-06 10:01:13 <ncopa> ok xen iso works
2013-05-06 10:01:14 <ncopa> hm
2013-05-06 10:01:20 <ncopa> dunno clandmeter
2013-05-06 10:01:30 <clandmeter> do you mind if i move it?
2013-05-06 10:01:30 <ncopa> feel free to move it to main
2013-05-06 10:01:40 <clandmeter> k thx
2013-05-06 10:01:46 <ncopa> lunch
2013-05-06 10:01:55 <royger> bon apetit!
2013-05-06 10:06:36 <vkrishn> also if lxc is being preferred that btrfs, zfs, apparmor compliment it usage
2013-05-06 10:06:47 <vkrishn> than*
2013-05-06 11:30:43 <clandmeter> ncopa: im just testing 2.6 but it freezes when i start x, and i cant find any info in the logs.
2013-05-06 11:38:56 <ncopa> with udev? mdev? what graphics driver?
2013-05-06 11:39:06 <ncopa> can you log in via ssh from other box and troubleshoot whats going on?
2013-05-06 11:39:18 <ncopa> xorg-server is a bit tricky to set up manually
2013-05-06 11:39:45 <ncopa> i have ended up rebooting box 3-5 times before i get it right
2013-05-06 11:44:25 <clandmeter> it locks up without access
2013-05-06 11:44:32 <clandmeter> kernel crashes i guess
2013-05-06 11:44:47 <clandmeter> network also down
2013-05-06 11:44:58 <clandmeter> udev intel
2013-05-06 11:47:18 <clandmeter> weird thing is, it worked the first time.
2013-05-06 12:13:54 <clandmeter> ncopa: if the kernel hard locks and there is no log anywyere, is there a way to debug that?
2013-05-06 12:43:40 <ncopa> clandmeter, maybe via remote gdb or something
2013-05-06 12:43:41 <ncopa> i dunno
2013-05-06 13:15:13 <royger> clandmeter: set up a serial cable and force a panic
2013-05-06 13:30:58 <clandmeter> royger: that works even if I dont have console access over vga?
2013-05-06 14:13:09 <royger> clandmeter: yes, you get a console on the serial cable
2013-05-06 14:17:13 <royger> clandmeter: and I think you can also use kgdb over the serial console, but I have never used it
2013-05-06 14:27:14 <clandmeter> royger: thx, ill give it a go when im back home. dont have serial here.
2013-05-06 14:39:26 <vkrishn> ncopa, package update is frozen?
2013-05-06 14:39:59 <ncopa> package update?
2013-05-06 14:40:09 <vkrishn> zmq , nginx has newer versions
2013-05-06 14:40:23 <vkrishn> I mean for v2.6
2013-05-06 14:40:39 <ncopa> whats the new nginx ver?
2013-05-06 14:40:41 <vkrishn> nginx has SPDY, I would like to test
2013-05-06 14:40:46 <vkrishn> :)
2013-05-06 14:40:48 <ncopa> theoretically its frozen
2013-05-06 14:40:53 <ncopa> but we can do exceptions
2013-05-06 14:41:01 <vkrishn> improves website by 40%
2013-05-06 14:41:08 <ncopa> i think nginx got updated already
2013-05-06 14:41:11 <vkrishn> thanks
2013-05-06 14:41:12 <ncopa> shoudl have nginx-1.4
2013-05-06 14:41:15 <vkrishn> ok
2013-05-06 14:41:31 <ncopa> what zmq version do you want?
2013-05-06 14:41:37 <ncopa> we have 3.2.2 currently
2013-05-06 14:41:49 <vkrishn> some minor bug fixes
2013-05-06 14:42:14 <ncopa> ok, thanks
2013-05-06 14:42:22 <ncopa> i'll throw in 3.2.3
2013-05-06 14:42:23 <vkrishn> let me check if they are tagged critical
2013-05-06 14:42:30 <ncopa> nah
2013-05-06 14:42:31 <vkrishn> ok
2013-05-06 14:42:37 <ncopa> i can do minor version update
2013-05-06 14:42:38 <ncopa> np
2013-05-06 14:44:02 <Barthalion> I'm trying to push nginx with spdy, but git.al.o dislikes me
2013-05-06 14:44:17 <Barthalion> I'm unable to connect
2013-05-06 14:44:22 <ncopa> what is missing?
2013-05-06 14:44:32 <ncopa> can you sprunge the diff?
2013-05-06 14:44:41 <ncopa> or git format-patch HEAD^
2013-05-06 14:44:49 <Barthalion> sure
2013-05-06 14:44:58 <ncopa> and sprunge < 0001-*.patch
2013-05-06 14:45:05 <ncopa> i'll git am it from here
2013-05-06 14:45:09 <Barthalion> ncopa: http://sprunge.us/dRSJ
2013-05-06 14:45:47 <Barthalion> Ah, connect
2013-05-06 14:45:49 <Barthalion> I'll push
2013-05-06 14:45:56 <Barthalion> connected*
2013-05-06 14:46:27 <Barthalion> No idea why couldn't before :c
2013-05-06 14:46:53 <ncopa> ok. i do rc3?
2013-05-06 14:48:44 <vkrishn> btrfs tools are in testing, any issues with them
2013-05-06 14:49:09 <vkrishn> when I last test some 6months back, I did find some issues, but cannot recall it now
2013-05-06 14:50:40 <vkrishn> if someone has tested it, pls post bugs
2013-05-06 14:50:45 <ncopa> have they made any release?
2013-05-06 14:50:52 <ncopa> i think you need use git snapshow
2013-05-06 14:50:59 <ncopa> eg current development for it
2013-05-06 14:51:21 <Barthalion> We can use 0.20rc1
2013-05-06 14:51:31 <ncopa> oh 0.20rc1 is out
2013-05-06 14:51:31 <Barthalion> We made our tarball in Arch
2013-05-06 14:52:01 <ncopa> i'd be ok with that
2013-05-06 14:52:05 <Barthalion> ftp://ftp.archlinux.org/other/btrfs-progs/btrfs-progs-0.20-rc1.1.tar.xz
2013-05-06 14:52:27 <vkrishn> https://www.kernel.org/pub/linux/kernel/people/mason/btrfs-progs/
2013-05-06 14:52:33 <ncopa> i'd be ok just use the arch tarball
2013-05-06 14:52:42 <ncopa> ah
2013-05-06 14:52:43 <ncopa> there it is
2013-05-06 14:52:52 <ncopa> ok
2013-05-06 14:53:00 <ncopa> lets do that after rc3
2013-05-06 14:53:14 <Barthalion> When 2.6 will be released? 8th?
2013-05-06 14:53:16 <ncopa> im doin rc3 now
2013-05-06 14:53:21 <ncopa> it should be tomorrow
2013-05-06 14:53:35 <Barthalion> Ok, so I'll update xfce4 too
2013-05-06 14:53:41 <Barthalion> There are bugfix releases
2013-05-06 14:53:49 <Barthalion> 4.10.0 → 4.10.1
2013-05-06 14:53:55 <ncopa> yeah
2013-05-06 14:54:01 <ncopa> do that thanks
2013-05-06 14:55:14 <Barthalion> ncopa: Will you update and move btrfs?
2013-05-06 14:55:40 <ncopa> i'm on it
2013-05-06 14:57:48 <ncopa> ok i gotta go
2013-05-06 14:58:09 <ncopa> Barthalion: if you want work on btrfs-progs, this what i got so far: http://sprunge.us/UUhF
2013-05-06 14:58:23 <Barthalion> not found
2013-05-06 14:58:34 <Barthalion> I'll finish xfce and then make btrfs, no problem
2013-05-06 14:59:42 <ncopa> i got it
2013-05-06 15:00:00 <Barthalion> vkrishn: Could you test it?
2013-05-06 15:00:11 <ncopa> thanks for your help and feedback everyone!
2013-05-06 15:01:00 <vkrishn> rc3 iso out ?
2013-05-06 15:02:06 <clandmeter> missing zlib-dev?
2013-05-06 15:02:11 <Barthalion> Probably
2013-05-06 15:02:42 <vkrishn> :)
2013-05-06 15:05:48 <ncopa> im'm late
2013-05-06 15:06:01 <ncopa> could somebody please email that rc3 is out?
2013-05-06 15:06:10 <ncopa> see u
2013-05-06 15:06:40 <Barthalion> see you
2013-05-06 15:57:00 <Barthalion> i created a new btrfs partition and spammed it with dd
2013-05-06 15:57:07 <Barthalion> Everything works so far
2013-05-06 15:57:29 <Barthalion> Could someone check if it's possible to boot Alpine from btrfs? I gtg soon
2013-05-06 20:06:43 <vkrishn> mod_fcgid 2.3.6 considered stable, I see even wheezy has not moved to 2.3.7, thought 2.3.8 is available in svn
2013-05-06 20:06:49 <vkrishn> ?
2013-05-06 20:07:19 <vkrishn> 2.3.7 is released in 2012-04-23
2013-05-07 07:35:03 <ncopa> hi
2013-05-07 07:37:43 <vkrishn> wddx had error since 2.4x
2013-05-07 07:37:56 <vkrishn> I am unable to check in v2.6rc3
2013-05-07 07:38:04 <vkrishn> slow network
2013-05-07 10:52:38 <vkrishn> is it possible to use lbu to save apkovl when the installation type is 'data' ?
2013-05-07 11:00:30 <ncopa> vkrishn, yes
2013-05-07 11:03:51 <vkrishn> same as setup-lbu, lbu ci ?
2013-05-07 11:19:59 <ncopa> yes
2013-05-07 11:24:18 <vkrishn> it say 'not found in fstab'
2013-05-07 11:26:15 <ncopa> what do you have in /etc/lbu/lbu.conf?
2013-05-07 11:37:39 <vkrishn> LBU_MEDIA=usb
2013-05-07 11:37:55 <vkrishn> change it to /dev/sda2 ?
2013-05-07 11:38:30 <vkrishn> /dev/sda2 is /var in "data mode" installation
2013-05-07 11:41:12 <vkrishn> or just sda2 ?
2013-05-07 11:43:01 <vkrishn> has some tested http://wiki.alpinelinux.org/wiki/Install_Alpine_in_Qemu with v2.6rc3
2013-05-07 11:43:09 <vkrishn> ?
2013-05-07 11:43:26 <ncopa> where do want save your apkvol?
2013-05-07 11:43:39 <ncopa> apkovl*
2013-05-07 11:44:33 <vkrishn> /dev/sda2 if possible, same as data disk
2013-05-07 11:44:41 <vkrishn> but this is running using kvm
2013-05-07 11:44:59 <ncopa> ok
2013-05-07 11:45:11 <ncopa> i think lbu requires a mount under /media
2013-05-07 11:45:23 <ncopa> so i suppose you could add /dev/sda2 /media/sda2
2013-05-07 11:45:27 <vkrishn> yes, I tried doing a bind
2013-05-07 11:45:34 <vkrishn> saving woks, but
2013-05-07 11:45:46 <vkrishn> works, but lbu list does not
2013-05-07 11:46:14 <ncopa> hm
2013-05-07 11:46:20 <ncopa> i wonder if it will find it when it boots up too
2013-05-07 11:46:29 <ncopa> yo boot from cd or disk?
2013-05-07 11:47:06 <vkrishn> kvm iso -hda qcow.disk
2013-05-07 11:47:25 <ncopa> ok
2013-05-07 11:48:11 <ncopa> i wonder if it might be an idea to make a boot disk
2013-05-07 11:48:30 <ncopa> and boot from that instead om directly from iso
2013-05-07 11:49:09 <vkrishn> for now tried http://wiki.alpinelinux.org/wiki/Install_Alpine_in_Qemu but seems a problem
2013-05-07 11:49:20 <vkrishn> with v2.6rc3
2013-05-07 11:50:48 <vkrishn> lbu list works,
2013-05-07 11:51:04 <vkrishn> lbu list-backup sda2 not
2013-05-07 11:52:58 <vkrishn> seems iso is not able to syslinux the qemu disk
2013-05-07 11:53:30 <vkrishn> I am creating qemu2 disk using v1.0, hope the format has not changed
2013-05-07 11:57:22 <vkrishn> would give try later(night), when net speed improves :)
2013-05-07 12:01:21 <vkrishn> can apkovl boot option pick ovl file from disk too ?
2013-05-07 12:02:33 <hl> vkrishn: no, use ovl_dev for that
2013-05-07 12:03:17 <vkrishn> than, in 'data mode' it may be possible to pick the ovl saved to it ?
2013-05-07 12:03:36 <hl> ovl_dev is used to load an ovl when booting
2013-05-07 12:03:52 <hl> so if you specify ovl_dev=/dev/sda1, say, then it will search for an ovl on that device
2013-05-07 12:04:04 <hl> you can override the fstype like ovl_dev=/dev/sda1:ext3
2013-05-07 12:04:14 <vkrishn> :)
2013-05-07 12:04:23 <hl> incidentally, this is all implemented as a shell script
2013-05-07 12:04:43 <hl> you can see how it works under the covers by looking at the init script in /usr/share/mkinitfs
2013-05-07 13:13:29 <ncopa> i think i'll postpone alpine v2.6.0 til we ahve a grsec patch for 3.9 :-/
2013-05-07 13:19:22 <vkrishn> ncopa, I had send some initial abuilds, can I get some feedback pls
2013-05-07 13:21:12 <vkrishn> like is code stable, bugs or some alternate available
2013-05-07 13:22:25 <vkrishn> php-zmq, patch for varnish
2013-05-07 13:23:08 <ncopa> ah
2013-05-07 13:23:37 <ncopa> sorry about that
2013-05-07 13:23:43 <ncopa> q
2013-05-07 13:26:15 <ncopa> zmq.ini: FAILED
2013-05-07 13:26:16 <ncopa> md5sum: WARNING: 1 of 1 computed checksums did NOT match
2013-05-07 13:27:21 <ncopa> i think we want the extrension enabled by default in .ini
2013-05-07 13:27:55 <ncopa> http://sprunge.us/ZHUH
2013-05-07 13:28:10 <vkrishn> ok
2013-05-07 13:28:20 <ncopa> also
2013-05-07 13:28:27 <ncopa> you dont need libzmq in depends
2013-05-07 13:28:34 <ncopa> it will automatically pull it in
2013-05-07 13:28:43 <ncopa> >>> php-zmq*: Tracing dependencies...
2013-05-07 13:28:43 <ncopa>         php
2013-05-07 13:28:43 <ncopa>         libzmq
2013-05-07 13:28:43 <ncopa>         so:libc.so.0.9.32
2013-05-07 13:28:43 <ncopa>         so:libzmq.so.3
2013-05-07 13:28:57 <ncopa> the so:libzmq.so.3 dep
2013-05-07 13:29:01 <vkrishn> hmm..
2013-05-07 13:30:23 <ncopa> i'll apply it with this diff: http://sprunge.us/dLVW
2013-05-07 13:31:49 <vkrishn> ok
2013-05-07 13:32:56 <ncopa> vkrishn, what was the opterh patch you sent?
2013-05-07 13:33:47 <vkrishn> varnish, but at time of sending seems to be working, now I get errors
2013-05-07 13:33:56 <ncopa> argh
2013-05-07 13:33:59 <ncopa> i pushed wrong
2013-05-07 13:35:52 <ncopa> i cannot push the libc-dev -> uclibc-dev change
2013-05-07 13:35:56 <vkrishn> I think varnish is not suited to lowmem kvm setup
2013-05-07 13:38:38 <vkrishn> gcc libc-dev
2013-05-07 13:40:38 <vkrishn> also I added pkgusers,pkggroups
2013-05-07 13:41:58 <ncopa> that is good
2013-05-07 13:43:06 <vkrishn> there was report in mailing list 'samba does not startup properly if /var/run is on tmpfs'
2013-05-07 13:43:20 <vkrishn> hope that is not true with one I set in for varnish
2013-05-07 13:43:24 <ncopa> i think that was fixed
2013-05-07 13:43:27 <ncopa> it probably is
2013-05-07 13:43:33 <vkrishn> /var/run/varnish
2013-05-07 13:44:25 <vkrishn> then varnishd.initd may require some changes
2013-05-07 13:44:28 <ncopa> look how i fixed samba.initd script: http://sprunge.us/YQdR
2013-05-07 13:44:30 <ncopa> yes
2013-05-07 13:44:45 <ncopa> you need to run checkpath --directory ... from there
2013-05-07 13:46:39 <vkrishn> ok
2013-05-07 13:47:43 <ncopa> i'm writing a review of that email now
2013-05-07 13:55:14 <vkrishn> resend varnish patch ?
2013-05-07 14:00:08 <ncopa> i'm posted the review
2013-05-07 14:00:19 <ncopa> thanks for working on it
2013-05-07 14:02:58 <ncopa> Barthalion: thanks for emailing about rc3
2013-05-07 14:04:29 <vkrishn> thanks
2013-05-07 14:21:11 <vkrishn> would it be ok if I start a HTPC page on wiki.a.o ? Presently I don't know if alpine has all the deps needed.
2013-05-07 14:21:43 <vkrishn> I would mark it as WIP
2013-05-07 14:21:53 <vkrishn> Home Theater PC
2013-05-07 14:21:56 <vkrishn> :)
2013-05-07 14:23:05 <vkrishn> I know some app, like on screen-keyboard, gesture control not yet available
2013-05-07 14:24:57 <vkrishn> remote control apps, wiimote(may not be needed)
2013-05-07 14:44:39 <vkrishn> jbilyk, nangel , this one of the appliances I was talking about some time back
2013-05-07 14:45:01 <jbilyk> vkrishn: i think it's no problem to start the page
2013-05-07 14:45:20 <jbilyk> i'd love to have an alpine htpc (especially on arm)
2013-05-07 14:45:40 <hl> ncopa: mini-patch: setup-timezone: line 50: change "t)" to "z)". bugfix.
2013-05-07 14:46:49 <clandmeter> vkrishn: xbmc not an option?
2013-05-07 14:47:09 <vkrishn> yes, its most integral part
2013-05-07 14:47:21 <clandmeter> if you have intel then it should work
2013-05-07 14:47:26 <clandmeter> intel GPU
2013-05-07 14:48:11 <vkrishn> geebox has a lighter version
2013-05-07 14:48:27 <vkrishn> of 'media center'
2013-05-07 14:48:41 <clandmeter> anyway you need gpu acceleration
2013-05-07 14:49:03 <clandmeter> and currently your only choise is intel
2013-05-07 14:49:18 <vkrishn> hmm..
2013-05-07 14:50:08 <clandmeter> the nvidia and amd blobs are not supported
2013-05-07 14:50:18 <vkrishn> That should not be an issue if we package in partnered hardware brand box applicance
2013-05-07 14:50:49 <clandmeter> like?
2013-05-07 14:50:55 <vkrishn> so xbmc is not a good option?
2013-05-07 14:51:08 <clandmeter> has nothing to do with xbmc
2013-05-07 14:51:30 <clandmeter> you can however play if you put a lot of cpu power on it
2013-05-07 14:52:34 <vkrishn> If I remember I was able to run xbmc in kvm
2013-05-07 14:52:39 <clandmeter> not sure what cpu you need to play 1080p
2013-05-07 14:52:53 <clandmeter> it uses opengl for its interface
2013-05-07 14:52:57 <vkrishn> rasPi supports that
2013-05-07 14:53:10 <Barthalion> rpi has its own chip for decoding
2013-05-07 14:53:12 <clandmeter> not really
2013-05-07 14:53:29 <clandmeter> it has a differnet implementation
2013-05-07 14:53:36 <clandmeter> and it uses openmax
2013-05-07 14:53:50 <clandmeter> but we have no arm port...
2013-05-07 14:54:02 <vkrishn> true
2013-05-07 14:54:14 <clandmeter> so your only logical option atm is intel
2013-05-07 14:54:39 <vkrishn> what about those nano boards, atom powered
2013-05-07 14:54:54 <vkrishn> they support 1080p
2013-05-07 14:55:08 <clandmeter> its not about hardware
2013-05-07 14:55:26 <clandmeter> its the blobs that breaks it on alpine
2013-05-07 14:55:56 <clandmeter> sign an NDA and get them from nvidia ;-)
2013-05-07 14:57:43 <vkrishn> so build successful alpine htpc a "partnered hardware" would be needed ?
2013-05-07 15:05:06 <jbilyk> vkrishn: if you wanted nvidia drivers yes, since they're not compiled against uclibc
2013-05-07 15:08:47 <clandmeter> you need an OS driver which can be build against uclibc
2013-05-07 15:08:57 <clandmeter> which intel has
2013-05-07 15:09:47 <clandmeter> rpi also has it, but its not truely open.
2013-05-07 15:19:05 <vkrishn> blobs are referred to binary modules specific to hardware ?
2013-05-07 15:20:26 <vkrishn> then unless hardwares become fully open-source'd its would continue to be that way
2013-05-07 15:21:20 <jbilyk> vkrishn: some vendors only distribute compiled drivers as kernel modules
2013-05-07 15:21:27 <jbilyk> vkrishn: aka "binary blobs"
2013-05-07 15:22:29 <vkrishn> Not so bright side of it is "economic reasons"
2013-05-07 15:22:34 <jbilyk> yup
2013-05-07 15:22:50 <jbilyk> it makes sense for the company, just sucks if your kernel can't use their modules
2013-05-07 15:22:58 <vkrishn> how you expect to employ, earn, run economy
2013-05-07 15:23:35 <vkrishn> yes, than making licensing problems easier is the solution
2013-05-07 15:23:42 <clandmeter> the rpi loads the blob on boot
2013-05-07 15:24:04 <vkrishn> so rpi solves easier licensing issues
2013-05-07 15:24:47 <clandmeter> maybe soon we will have AMD support
2013-05-07 15:25:00 <clandmeter> they releases an OS driver last month
2013-05-07 15:25:10 <vkrishn> Not so bright side again: large part of economy would stumble it would be opensourced
2013-05-07 15:25:27 <clandmeter> http://www.phoronix.com/scan.php?page=article&item=amd_opensource_uvd&num=1
2013-05-07 15:28:27 <vkrishn> I think there could be a GPL's for blobs ;)
2013-05-07 15:29:23 <vkrishn> one that say, once a vendor has issued a license to certain distributor on certain terms, that terms is also applicable for other seeking distributor
2013-05-07 15:29:34 <vkrishn> ;))
2013-05-07 15:30:50 <vkrishn> from the article seems AMD is now htpc'd for alpine
2013-05-07 15:32:19 <clandmeter> going home. nice evening.
2013-05-07 15:32:29 <vkrishn> :)
2013-05-07 15:32:54 <vkrishn> gtg too, dinner
2013-05-07 15:38:01 <Barthalion> Would it be possible to provide /etc/os-release file?
2013-05-07 15:38:12 <Barthalion> I mean by default
2013-05-07 16:03:58 <tdtrask> Barthalion: in addition to /etc/alpine-release ?
2013-05-07 16:21:25 <Barthalion> http://www.freedesktop.org/software/systemd/man/os-release.html
2013-05-07 16:22:14 <Barthalion> I can add it on my own to my chroots, but maybe someone finds this useful too
2013-05-07 17:14:49 <vkrishn> clandmeter, I could not find any supporting article for phoronix at amd.com
2013-05-07 20:03:14 <build64-2-5> files from v2.5.4-61-g2b15056 uploaded
2013-05-08 06:37:36 <ncopa> hi
2013-05-08 06:37:42 <ncopa> i didnt know about os-release
2013-05-08 06:37:46 <ncopa> i think we can provide it
2013-05-08 06:38:52 <kaniini> why though
2013-05-08 06:39:01 <ncopa> but i am kinda sceptic to things that comes from people who needs to put "we hate what you do" in their FAQ
2013-05-08 06:39:01 <kaniini> systemd will never be added to alpine :P
2013-05-08 06:39:36 <kaniini> where does lennartp say this
2013-05-08 06:39:46 <kaniini> this sounds like an intriguing FAQ to read
2013-05-08 06:39:50 <ncopa> if they get that as a frequest asked question... maybe there is a reason for it...
2013-05-08 06:39:54 <ncopa> http://0pointer.de/blog/projects/os-release
2013-05-08 06:40:29 <ncopa> but i kinda like the concept
2013-05-08 06:40:32 <ncopa> it is a standard
2013-05-08 06:41:02 <kaniini> aha
2013-05-08 06:41:46 <kaniini> i wouldn't mind systemd if it didn't bring in dbus
2013-05-08 06:41:51 <ncopa> Barthalion: why do you want it?
2013-05-08 06:41:56 <ncopa> kaniini: agree
2013-05-08 06:42:02 <kaniini> i don't want dbus as part of alpine-base group
2013-05-08 06:42:07 <ncopa> agree
2013-05-08 06:42:08 <ncopa> +1
2013-05-08 06:42:14 <hl> i dunno
2013-05-08 06:42:21 <kaniini> i don't mind udev there, but dbus is too much
2013-05-08 06:42:30 <hl> systemd seems like a monopolist, so I'm not sure dbus is my only objection to it
2013-05-08 06:42:53 <kaniini> yes, that's the other point, alpine likes providing choice and flexibility (such as allowing separate-/usr mount)
2013-05-08 06:42:53 <ncopa> i also miss the ttys
2013-05-08 06:43:03 <hl> it is also tightly coupled
2013-05-08 06:43:10 <ncopa> bah dont get me started with /usr separation
2013-05-08 06:43:14 <hl> (why on earth does systemd have a login manager)
2013-05-08 06:43:42 <ncopa> one of the arguments was "it is not possible to boot a modern system with less than 500MB nowdays anyway"
2013-05-08 06:44:41 <ncopa> i think systemd has many good ideas but..
2013-05-08 06:44:49 <ncopa> like hl said, why a login manager???
2013-05-08 06:44:54 <ncopa> why depend on dbus?
2013-05-08 06:44:58 <ncopa> dbus??? hello??
2013-05-08 06:45:01 <hl> hopefully someone else will take them and run with them and then we can have the nice ideas without all the bizarreness.
2013-05-08 06:45:16 <ncopa> i have my on bizarre ideas :)
2013-05-08 06:45:19 <ncopa> just not time to implement
2013-05-08 06:45:45 <hl> and as for the tight couping I mentioned
2013-05-08 06:46:12 <hl> I use arch for a desktop which is systemd-based, it implements /etc/crypttab by generating systemd unit files from it at boot time
2013-05-08 06:46:33 <hl> i have yet to determine how this generation process can be run without rebooting
2013-05-08 06:46:38 <kaniini> hl: indeed, i think that systemd would remove a lot of the magic that makes alpine alpine
2013-05-08 06:46:54 <ncopa> hl i feel your pain
2013-05-08 06:47:13 <kaniini> interestingly, debian is switching to openrc
2013-05-08 06:47:25 <hl> in other words, it seems to me like systemd is going to turn booting into a theoretically understandable but in practice impractical to understand black box of a system that brings us to the Windowsesque "reboot to fix it"
2013-05-08 06:47:29 <kaniini> based partially on our work
2013-05-08 06:47:33 <ncopa> one thing isi for sure, alpine will not switch to systemd for the v2.6 release!
2013-05-08 06:47:39 <ncopa> really?
2013-05-08 06:47:41 <ncopa> cool!
2013-05-08 06:48:10 <ncopa> hl: we dont want alpine linux end up there...
2013-05-08 06:48:12 <hl> I'm sure humans could understand why issue X is fixed until you reboot windows, the point is it is "beyond economic understanding" and the only solution offered is "reboot".
2013-05-08 06:48:15 <hl> ncopa: exactly.
2013-05-08 06:48:55 <ncopa> kaniini, do you ahve any links?
2013-05-08 06:49:06 <kaniini> just things being thrown around their irc channels
2013-05-08 06:49:17 <kaniini> it's a summer of code project
2013-05-08 06:52:21 <kaniini> "dbus? udev? on my server?! and you expect a linux 3.0+ kernel? waaah!" -- plauer on systemd in Debian
2013-05-08 06:52:26 <kaniini> who incidentally, is now on tech-ctte
2013-05-08 06:52:37 <kaniini> so i don't see systemd being successful in debian
2013-05-08 06:52:56 <hl> it is a shame, since the idea of .ini files seems a lot more appealing to me than init scripts
2013-05-08 06:53:10 <hl> especially since systemd can do process supervision and inetd-like activation
2013-05-08 06:54:08 <ncopa> ok enough of systemd
2013-05-08 06:54:22 <ncopa> any objections against /etc/os-release?
2013-05-08 06:54:39 <ncopa> Barthalion: why do you want it?
2013-05-08 06:55:13 <ncopa> i'm not against it
2013-05-08 06:55:21 <kaniini> it seems ok
2013-05-08 06:55:22 <ncopa> but i dont want add it unless it actually solves a real problem
2013-05-08 06:55:31 <kaniini> but, i don't see the use-case
2013-05-08 06:55:37 <kaniini> we're not LSB compliant, and won't ever be
2013-05-08 06:56:15 <ncopa> we could replace the /etc/alpine-release with the /etc/os-release
2013-05-08 06:56:24 <ncopa> for compatibility
2013-05-08 06:56:42 <ncopa> if lets say llvm/clang looks for it
2013-05-08 06:56:45 <kaniini> sure
2013-05-08 06:57:23 <kaniini> really that's my question
2013-05-08 06:57:27 <kaniini> what realworld problem is solved by this
2013-05-08 06:57:31 <ncopa> exactly
2013-05-08 06:57:36 <ncopa> thats what i want to know too
2013-05-08 06:57:46 <ncopa> Barthalion?
2013-05-08 07:09:26 <vkrishn> how to get the name of os and version from command line ?
2013-05-08 07:11:10 <ncopa> uname -a ;)
2013-05-08 07:13:58 <vkrishn> is the string consistent for grep to get `os name` + version ?
2013-05-08 07:14:09 <vkrishn> on knoppix its
2013-05-08 07:14:12 <vkrishn> Linux Microknoppix 3.3.7-64 #40 SMP PREEMPT Tue May 22 08:47:38 CEST 2012 x86_64 GNU/Linux
2013-05-08 07:16:44 <vkrishn> can lxc have containers from other linux distros ?
2013-05-08 07:16:54 <ncopa> yes
2013-05-08 07:17:25 <vkrishn> then one usage for os-release(though name sounds not so ok) - is a profiler that gives info for OS's installed, consistently
2013-05-08 07:17:46 <vkrishn> any thoughts on AF_BUS ?
2013-05-08 07:18:01 <ncopa> its the dbus in kernel thingy?
2013-05-08 07:18:15 <ncopa> sounds ok to me
2013-05-08 07:18:30 <vkrishn> I have not read much about it, patch was refused upstream
2013-05-08 07:18:35 <hl> no, that's kdbus isn't it
2013-05-08 07:18:38 <hl> AF_DBUS is more neutral
2013-05-08 07:18:45 <hl> err, AF_BUS i mean
2013-05-08 07:19:07 <ncopa> http://projects.genivi.org/afbus-dbus-optimization/about
2013-05-08 07:19:29 <vkrishn> can AF_DBUS replace dbus in systemd ?
2013-05-08 07:19:45 <ncopa> hopefully
2013-05-08 07:20:01 <ncopa> AF_BUS sounds like AF_UNIX with multicase
2013-05-08 07:20:05 <ncopa> multicast
2013-05-08 07:20:52 <vkrishn> then systemd people should try to release dbus a pluggable module with AF_DBUS as alternate
2013-05-08 07:21:21 <ncopa> heh.. do you really believe that they will allow anything "alternate"?
2013-05-08 07:21:26 <hl> the systemd people simply do not care
2013-05-08 07:21:32 <vkrishn> or maybe we send them a patch ;)
2013-05-08 07:21:38 <hl> they have indicated that they would actively refuse patches to e.g. make systemd work on bsd, etc.
2013-05-08 07:21:46 <hl> do you really want a hostile upstream?
2013-05-08 07:21:57 <ncopa> which is one of the bigget issues with systemd
2013-05-08 07:22:06 <ncopa> they dont want anything alternate
2013-05-08 07:22:14 <ncopa> do it our way or go away
2013-05-08 07:22:20 <ncopa> "no.. dont go away"
2013-05-08 07:22:24 <ncopa> "do it out way"
2013-05-08 07:22:30 <hl> ncopa: wait that's a good slogan for systemd, it's catchy
2013-05-08 07:22:33 <hl> "do it our way, or go away"
2013-05-08 07:23:12 <ncopa> if they were open to "alternate" ways to do things we would probably already been on systemd
2013-05-08 07:23:24 <vkrishn> I think publishing a patch that provides alternate is a mater of getting acceptance by opensource community
2013-05-08 07:23:32 <vkrishn> it accepted, rest is history
2013-05-08 07:23:34 <vkrishn> if*
2013-05-08 07:24:15 <ncopa> kdbus is entirely different from AF_BUS.
2013-05-08 07:24:31 <vkrishn> having a say in open-source - a patch awaited
2013-05-08 07:24:42 <vkrishn> having a say in open-source - is a patch awaited
2013-05-08 07:25:02 <vkrishn> ooops sounds like a QUOTE
2013-05-08 07:25:04 <vkrishn> ;)
2013-05-08 07:25:32 <hl> anyway, AF_BUS was rejected
2013-05-08 07:25:48 <hl> as I understand it the kernel sockets maintainer was unconvinced that you could do reliable multicast
2013-05-08 07:26:08 <vkrishn> AF_BUS patches are in the LTSI kernel releases
2013-05-08 07:26:41 <kaniini> good for that?
2013-05-08 07:26:51 <kaniini> we're trying to *drop* patches in our kernels
2013-05-08 07:27:01 <vkrishn> :)
2013-05-08 07:27:29 <kaniini> we don't want to add more patches, we want to get rid of them
2013-05-08 07:27:35 <kaniini> it's why vserver is being deprecated
2013-05-08 07:27:51 <ncopa> yes
2013-05-08 07:28:05 <ncopa> once kdbus is in mainline kernel we will probably have it
2013-05-08 07:28:06 <ncopa> and support it
2013-05-08 07:28:22 <kaniini> incidentally openvz stuff is switching to namespaces
2013-05-08 07:28:25 <ncopa> looks like they are going for kdbus rather than AF_BUS
2013-05-08 07:28:43 <hl> how politically neutral is this 'kdbus' thing?
2013-05-08 07:29:22 <kaniini> so, we can add openvz support with 2.7 as experimental (using namespaces mode)
2013-05-08 07:39:38 <ncopa> so openvz will not require kernel patch?
2013-05-08 07:39:42 <ncopa> thats something...
2013-05-08 07:42:54 <kaniini> http://openvz.org/Vzctl_for_upstream_kernel
2013-05-08 07:42:57 <kaniini> ncopa: ^
2013-05-08 07:46:31 <ncopa> aha
2013-05-08 07:46:42 <ncopa> This article describes using OpenVZ tool vzctl as an alternative to LXC tools.
2013-05-08 07:47:59 <ncopa> ok
2013-05-08 07:48:02 <ncopa> that lookspromising
2013-05-08 07:52:32 <ncopa> fcolista: are you ok with this change: http://sprunge.us/KQRS
2013-05-08 07:52:52 <ncopa> ah
2013-05-08 07:52:59 <ncopa> needs set perms on /var/rancid
2013-05-08 07:53:04 <ncopa> ownership
2013-05-08 07:53:10 <fcolista> +1 ncopa
2013-05-08 07:53:31 <fcolista> should add pkgusers
2013-05-08 07:53:37 <fcolista> and pkggroups
2013-05-08 07:54:53 <ncopa> http://sprunge.us/jUca
2013-05-08 07:54:54 <ncopa> ok?
2013-05-08 07:57:05 <fcolista> yes
2013-05-08 07:57:39 <fcolista> cloginrc is not used?
2013-05-08 07:57:48 <ncopa> hm
2013-05-08 07:58:06 <fcolista> it contains config file
2013-05-08 07:58:42 <ncopa> i assuemd --enable-conf-install fixed that..
2013-05-08 07:59:26 <ncopa> fedora does not install the cloginrc
2013-05-08 07:59:28 <ncopa> http://pkgs.fedoraproject.org/cgit/rancid.git/tree/rancid.spec
2013-05-08 08:01:20 <fcolista> ok
2013-05-08 08:01:23 <fcolista> i will try it
2013-05-08 11:54:53 <Barthalion> ncopa: Otherwise systemd-nspawn is unable to chroot into system
2013-05-08 11:56:07 <ncopa> it refuses to chroot if there is no os-release?
2013-05-08 11:56:19 <Barthalion> Generally, yes
2013-05-08 11:56:20 <ncopa> it refuses to chroot if there is no systemd in new chroot?
2013-05-08 11:56:28 <Barthalion> If there is no os-release file
2013-05-08 11:56:36 <ncopa> what does it need it for?
2013-05-08 11:56:40 <ncopa> print pretty name?
2013-05-08 11:56:41 <Barthalion> Basically touch etc/os-release is good workaround, it's not crucial
2013-05-08 11:56:51 <Barthalion> For some metadata, I don't care to be honest
2013-05-08 11:57:11 <ncopa> if touch etc/os-release "solves" it
2013-05-08 11:57:17 <Barthalion> Just asking if it could be done from our side, I'm not big fan of workarounds
2013-05-08 11:57:22 <ncopa> then it sounds like systemd needs a tiny patch
2013-05-08 11:57:32 <hl> yeah
2013-05-08 11:57:49 <ncopa> instead of bail out with error, continue as if it was empty
2013-05-08 11:58:00 <Barthalion> Yeah, possibly
2013-05-08 11:58:07 <Barthalion> But it's systemd :p
2013-05-08 11:58:20 <ncopa> i know i know...
2013-05-08 11:58:24 <ncopa> ugh
2013-05-08 11:59:04 <ncopa> sillyness like that is a good reason to why not switch to systemd
2013-05-08 11:59:24 <ncopa> they dont do it for tech reason, by for political reasons
2013-05-08 11:59:59 <ncopa> not strange they get so many "we hate you" that they add it to a FAQ...
2013-05-08 12:00:11 <ncopa> ok
2013-05-08 12:00:25 <ncopa> Barthalion: I'm ok if you add it to alpine-base
2013-05-08 12:00:35 <Barthalion> Great, thank you
2013-05-08 12:00:41 <Barthalion> Will do later
2013-05-08 12:00:56 <ncopa> echo "DATA=<value" > "$pkgdir"/etc/os-release
2013-05-08 12:01:08 <ncopa> just make sure that you get the correct values
2013-05-08 12:01:22 <ncopa> i suspect it can be difficult to change it later
2013-05-08 12:17:44 <clandmeter> ncopa: do you mind if i bump v4l-utils?
2013-05-08 12:19:34 <ncopa> # Maintainer: Francesco Colista <francesco.colista@gmail.com>
2013-05-08 12:19:37 <ncopa> ask fcolista
2013-05-08 12:19:49 <ncopa> if its abi compat i suppose its ok
2013-05-08 12:20:06 <ncopa> if you need rebuild 5 pkgs due to abi break i'm ok too
2013-05-08 12:20:16 <ncopa> but if you need rebuild 120... then better not :)
2013-05-08 12:20:42 <clandmeter> ok
2013-05-08 12:21:10 <clandmeter> what was the cmd to check abi?
2013-05-08 12:21:43 <clandmeter> checkapk :)
2013-05-08 12:24:19 <clandmeter> fcolista: do you mind if i bump it and clean it up a little?
2013-05-08 13:01:14 <vkrishn> just did cgit pull to update, show some files as untracked, is it ok?
2013-05-08 13:11:03 <vkrishn> ok now
2013-05-08 13:33:12 <vkrishn> ncopa, send varnish patch, I still have v2.5 dev setup, no libc-dev trans pak, it compiles ok with uclibc-dev though
2013-05-08 13:34:01 <vkrishn> would appreciate if to know if they are ok, I even cannot test the compiled in kvm
2013-05-08 13:37:17 <grimeton> hi
2013-05-08 13:37:20 <grimeton> check the mirrors http://www.welivesecurity.com/2013/05/07/linuxcdorked-malware-lighttpd-and-nginx-web-servers-also-affected/
2013-05-08 13:39:42 <ncopa> hm
2013-05-08 13:39:49 <ncopa> is the executable file modified?
2013-05-08 13:39:57 <ncopa> then we can just check with apk verify
2013-05-08 13:40:11 <grimeton> just reading, didn't finish the main report yet
2013-05-08 13:40:23 <grimeton> just read it on a german news portal and copied the link in here
2013-05-08 13:41:55 <ncopa> The following image shows the assembly listing of the reverse connect back shell invocation from the three types of backdoored binaries we were able to analyze:
2013-05-08 13:42:16 <ncopa> so unless the backdoor was compield from source, we should be able to detect it with apk verify
2013-05-08 13:42:28 <jbilyk> apk audit --system ?
2013-05-08 13:43:21 <grimeton> what scares me a bit is the fact that it doesn't come in via a scripting language like php or python that is used on the servers
2013-05-08 13:43:25 <grimeton> but still not finished reading
2013-05-08 13:45:09 <ncopa> jbilyk: yes
2013-05-08 13:45:55 <ncopa> Before going any further, one point needs to be clear about Linux/Cdorked.A. We still don’t know for sure how this malicious software was deployed on the web servers.
2013-05-08 13:46:32 <grimeton> the audit only returns the modules.alias, modules.symbols and modules.dep are not original (maskes sense imho)
2013-05-08 13:46:40 <ncopa> yup
2013-05-08 13:47:04 <ncopa> at the end of the article they have sha1 hashes of the analyzed binaries
2013-05-08 13:47:20 <grimeton> i wonder if binary deltas that one could use to patch the software would even work on alpine as of the uclibc
2013-05-08 13:47:26 <Barthalion> doesn't mean they won't change
2013-05-08 13:47:32 <ncopa> so, yes apk audit --system should pick it up
2013-05-08 13:47:46 <grimeton> i guess the entry points are different in software compiled/linked against uclibc
2013-05-08 13:48:03 <ncopa> they could probably make it run on uclibc
2013-05-08 13:48:17 <grimeton> sure, but not ootb and not when targeting the glibc folks :D
2013-05-08 13:48:18 <ncopa> what i wonder is how it got depolyed on the server in first place
2013-05-08 13:49:03 <grimeton> maybe they have been deployed by patched packages because they also modified the dns servers of those server
2013-05-08 13:49:04 <ncopa> pax and that kind of stuff is supposed to prevent that
2013-05-08 13:49:17 <grimeton> so maybe they redirected their update attempts to a server with a changed package...
2013-05-08 13:49:32 <ncopa> i think they already had control of the servers
2013-05-08 13:50:11 <grimeton> yeah something like that
2013-05-08 13:50:19 <grimeton> only way to logon to my mirror via ssh is by key or kerberos auth
2013-05-08 13:50:33 <ncopa> or security hole in something...
2013-05-08 13:50:56 <jbilyk> all alpinelinux.org infra that we manage has been checked
2013-05-08 13:51:06 <ncopa> goody
2013-05-08 13:51:08 <ncopa> thanks!
2013-05-08 13:51:10 <jbilyk> np
2013-05-08 13:51:17 <grimeton> yeah dl-5 is clean too
2013-05-08 13:53:15 <clandmeter> jbilyk: you checked alpine-proxy?
2013-05-08 13:53:41 <clandmeter> i did, but i didnt see you logged in
2013-05-08 13:54:16 <jbilyk> clandmeter: i just checked the vserver host
2013-05-08 13:54:29 <jbilyk> 88.159.20.184
2013-05-08 14:01:18 <grimeton> lol, a us judge filed a court order based entirely on star trek against some lawyers that tried to earn some money by sueing people that downloaded a porn video
2013-05-08 14:01:27 <grimeton> this is so awesome http://www.popehat.com/wp-content/uploads/2013/05/PendaSanctionsOrder.pdf
2013-05-08 14:01:42 <grimeton> he calls the attourneys "redshirts"
2013-05-08 14:02:21 <grimeton> "It was when the Court realized Plaintiffs engaged their cloak of shell companies and fraud that the Court went to battlestations.
2013-05-08 14:02:24 <grimeton> "
2013-05-08 15:41:15 <fcolista> hi clandmeter
2013-05-08 15:41:21 <fcolista> ok for me
2013-05-09 08:13:40 <vkrishn> I missed prefix in subject line "main/varnish: "
2013-05-09 08:14:00 <vkrishn> :(
2013-05-09 08:17:32 <ncopa> np
2013-05-09 08:17:48 <ncopa> you'll remember it next time now ;)
2013-05-09 08:18:06 <vkrishn> lesson learnt
2013-05-09 08:18:19 <vkrishn> found some tool, http://www.kraxel.org/blog/linux/fbida/
2013-05-09 08:18:24 <ncopa> good work on that patch btw
2013-05-09 08:18:24 <ncopa> thanks
2013-05-09 08:18:47 <vkrishn> thanks for review
2013-05-09 08:19:14 <vkrishn> Reading a lot about other distros switching to MariaDB
2013-05-09 08:19:24 <ncopa> i have seen it too
2013-05-09 08:19:33 <ncopa> maybe for v2.7 we do the same
2013-05-09 08:19:39 <vkrishn> its a dropin replacement for mysql
2013-05-09 08:20:17 <vkrishn> alternatives are percona
2013-05-09 09:42:22 <clandmeter> ncopa: in provdes so:xxx=1.2.3 is 1.2.3 abi version?
2013-05-09 09:43:27 <ncopa> yes i think so
2013-05-09 09:43:30 <ncopa> yes
2013-05-09 09:43:44 <clandmeter> when will it be added and when not?
2013-05-09 09:44:29 <clandmeter> when there is another package that provides it?
2013-05-09 09:55:33 <clandmeter> ncopa: any idea?
2013-05-09 09:56:00 <ncopa> yes, when other package provides it
2013-05-09 10:00:11 <clandmeter> oh i was probably miss reading it. it always provides the abi version in p lines
2013-05-09 17:49:51 <Mp5shooter> hey guys
2013-05-09 17:49:58 <Mp5shooter> bugs.alpinelinux.org SSL cert expired yesterda
2013-05-09 17:50:00 <Mp5shooter> yesterday
2013-05-09 18:18:54 <fabled> Mp5shooter, yes, we have new certs already, but they are just not pushed out yet. ncopa ^^^
2013-05-09 18:20:21 <Mp5shooter> oh ok
2013-05-09 18:20:22 <Mp5shooter> also
2013-05-09 18:20:27 <Mp5shooter> fabled: do you know who packaged lua-zlib?
2013-05-09 18:25:22 <fabled> Mp5shooter, i suppose ncopa; he at least made the original commit adding the APKBUILD
2013-05-09 18:25:28 <Mp5shooter> oh
2013-05-09 18:25:28 <Mp5shooter> mm
2013-05-09 18:25:43 <Mp5shooter> i'm using prosody xmpp server (in testing)
2013-05-09 18:25:54 <Mp5shooter> http://prosody.im/doc/depends#lua-zlib
2013-05-09 18:26:15 <Mp5shooter> apparently there's a zlib library out there called 'brimworks'
2013-05-09 18:26:26 <Mp5shooter> should i request that it be packaged?
2013-05-09 18:26:50 <Mp5shooter> normal lua-zlib doesn't work
2013-05-09 18:26:53 <Mp5shooter> (with prosody)
2013-05-09 18:27:31 <fabled> i think so. sounds like it's completely separate implementation/project with just same name as the other one
2013-05-09 18:28:24 <Mp5shooter> ok
2013-05-09 18:28:35 <Mp5shooter> i'll bring my laptop to work and request there... gotta go now
2013-05-09 19:00:48 <Mp5shooter> ok i'm back
2013-05-10 06:06:54 <ncopa> aww
2013-05-10 06:06:58 <ncopa> certs expired..
2013-05-10 06:32:22 <ncopa> Mp5shooter, bugs.a.o certs are updated
2013-05-10 06:46:47 <ncopa> i'm restarting bugs.a.o
2013-05-10 06:47:49 <ncopa> argh
2013-05-10 06:54:30 <ncopa> ok
2013-05-10 06:54:38 <ncopa> i have a tense relationship with redmine...
2013-05-10 06:54:52 <ncopa> simply doing ssec upgrades makes it fail to start up...
2013-05-10 07:17:23 <build64-2-5> files from v2.5.4-62-ga513864 uploaded
2013-05-10 07:26:06 <build64-2-5> files from v2.5.4-63-gb3feba3 uploaded
2013-05-10 07:35:10 <build64-2-3> files from v2.3.6-201-g622b89d uploaded
2013-05-10 07:43:40 <build64-2-5> files from v2.5.4-64-g774f912 uploaded
2013-05-10 07:45:37 <clandmeter> redmine down?
2013-05-10 07:45:57 <clandmeter> ncopa: did you break it?
2013-05-10 07:46:44 <ncopa> yup :-(
2013-05-10 07:46:47 <ncopa> i did apk ugprade
2013-05-10 07:47:24 <clandmeter> are you fixing it?
2013-05-10 07:47:25 <ncopa> there was security updates for ruby-rails ruby-activerecord
2013-05-10 07:47:28 <ncopa> i'm on it
2013-05-10 07:47:32 <ncopa> its a nightmare...
2013-05-10 07:47:33 <clandmeter> again?
2013-05-10 07:47:39 <ncopa> no
2013-05-10 07:47:41 <ncopa> those are old
2013-05-10 07:47:57 <clandmeter> i think i already did that some time ago.
2013-05-10 07:48:08 <ncopa> i pushed them to the v2.5 repo some time ago
2013-05-10 07:48:12 <ncopa> but i never updated bugs.a.o
2013-05-10 07:49:12 <clandmeter> why doesnt it start?
2013-05-10 07:49:33 <ncopa> dependency hell
2013-05-10 07:49:45 <ncopa> some component has a hard dep on rails=3.11
2013-05-10 07:49:49 <ncopa> 3.2.11
2013-05-10 07:49:54 <ncopa> which is old an vunlerable
2013-05-10 07:50:20 <clandmeter> ah i see
2013-05-10 07:50:30 <ncopa> now
2013-05-10 07:50:42 <ncopa> there are like 1000 deps that needs be handled
2013-05-10 07:50:51 <ncopa> so much for stable releases...
2013-05-10 07:51:59 <ncopa> i suspect we have tup upgrade redmine
2013-05-10 07:52:13 <ncopa> i manually edited the gemsfile to require rails=3.2.13
2013-05-10 07:53:00 <ncopa> redmine is just painful to maintain...
2013-05-10 07:53:56 <build64-2-5> files from v2.5.4-65-gf78d955 uploaded
2013-05-10 07:55:08 <ncopa> its up
2013-05-10 07:55:12 <ncopa> phew
2013-05-10 07:55:32 <clandmeter> are you using bundler now?
2013-05-10 07:55:43 <ncopa> partially
2013-05-10 07:55:48 <clandmeter> ouch
2013-05-10 07:56:18 <ncopa> it installs whats missing
2013-05-10 07:56:26 <ncopa> when it tries to compiel something
2013-05-10 07:56:31 <ncopa> i update the apkbuild
2013-05-10 07:56:35 <ncopa> as no gcc is there
2013-05-10 07:57:21 <clandmeter> yes
2013-05-10 07:57:41 <clandmeter> i update the needed apk's myself so gem is not needed
2013-05-10 07:57:46 <clandmeter> and disable bundler
2013-05-10 07:58:00 <ncopa> ok
2013-05-10 07:58:05 <clandmeter> else you mix apk's with gems
2013-05-10 07:58:13 <ncopa> yyeah..
2013-05-10 07:58:16 <ncopa> thats what i did
2013-05-10 07:58:25 <clandmeter> thats why i said ouch :)
2013-05-10 07:58:31 <ncopa> i panicked.. sorry
2013-05-10 08:01:07 <clandmeter> ncopa: any reason alpine-iso is not in aports?
2013-05-10 08:01:28 <ncopa> it is a makefile..
2013-05-10 08:01:38 <ncopa> not really something we can package
2013-05-10 08:01:51 <clandmeter> on the ml somebody has issues usign it
2013-05-10 08:02:02 <ncopa> i saw
2013-05-10 08:02:04 <clandmeter> and i think its because its not backwards compat
2013-05-10 08:02:06 <ncopa> and i think you are right
2013-05-10 08:02:13 <ncopa> yes
2013-05-10 08:02:17 <ncopa> i think we should branch it
2013-05-10 08:02:20 <ncopa> every release
2013-05-10 08:03:28 <ncopa> clandmeter: do you know if there is some way to know what gem is from bundler and which is from apk?
2013-05-10 08:04:07 <clandmeter> only by checking apk info and gem list afaik
2013-05-10 08:05:01 <ncopa> http://dpaste.org/BEmEQ/
2013-05-10 08:05:06 <build64-2-5> files from v2.5.4-71-g27c8280 uploaded
2013-05-10 08:05:06 <ncopa> all those Intalling...
2013-05-10 08:05:09 <clandmeter> gem list also shows multiple versions installed
2013-05-10 08:05:13 <build-2-5> files from v2.5.4-71-g27c8280 uploaded
2013-05-10 08:05:16 <ncopa> shoud probably be packaged
2013-05-10 08:06:55 <clandmeter> we could remove all the gems and run apk fix on them
2013-05-10 08:07:37 <ncopa> Installing rack (1.4.5)
2013-05-10 08:10:04 <ncopa> i wonder if we should try use versioned deps on the apks
2013-05-10 08:10:10 <ncopa> on the ruby apks
2013-05-10 08:15:49 <ncopa> clandmeter: it woudl be nice if you could try set up a test redmine using v2.6
2013-05-10 08:16:02 <ncopa> just to make sure we have all deps we need
2013-05-10 08:26:22 <clandmeter> ncopa: i still need to look into the deps, or did you take care of those?
2013-05-10 08:29:45 <ncopa> clandmeter: i think i took care of most of them, if not all
2013-05-10 14:27:08 <vkrishn> which is the preferred driver nv or nouveau ?
2013-05-10 14:27:54 <ncopa> i dont have that hw but i woudl guess nouveau
2013-05-10 14:28:07 <Barthalion> if nouveau works for you, pick it
2013-05-10 14:28:33 <vkrishn> so it depends on cards supported ?
2013-05-10 14:28:58 <vkrishn> in each package
2013-05-10 14:28:59 <Barthalion> video-nv is afaik deprecated by nvidia
2013-05-10 14:29:40 <Barthalion> even if it's developed (doubt it), it works only on older gpus
2013-05-10 14:29:55 <vkrishn> great then,  alpine-iso/alpine-desktop.packages has nv package in it
2013-05-10 14:30:25 <Barthalion> alpine-desktop should fallback to vesa, but right, there should be nouveau too
2013-05-10 14:33:36 <vkrishn> wikipedia.org/Graphics_hardware_and_FOSS.html says noveau is marked stable
2013-05-10 14:33:51 <Barthalion> it depends on its mood
2013-05-10 14:34:31 <ncopa> alpine-iso/alpine-desktop.packages has not been touched for long time
2013-05-10 14:35:58 <vkrishn> from article nv drivers sounds gloomy
2013-05-10 14:38:31 <vkrishn> ncopa, I was thinking to enhance the sprunge package, make it more generic
2013-05-10 14:38:39 <vkrishn> enhancing*
2013-05-10 14:39:10 <vkrishn> http://dpaste.org/about/ also provides similar curl interface, and I believe there are other pasteboard too
2013-05-10 14:40:24 <vkrishn> so like maybe a small app like pasty, textshare... name yet to decide
2013-05-10 14:41:14 <ncopa> vkrishn, we have pastebinc in testing
2013-05-10 14:41:17 <vkrishn> this app would take second param
2013-05-10 14:41:21 <vkrishn> oh
2013-05-10 14:41:30 <ncopa> which tries to be generic
2013-05-10 14:41:43 <ncopa> i like sprunge because it takes not params
2013-05-10 14:41:52 <ncopa> we coudl add a second script dpaste
2013-05-10 14:42:53 <vkrishn> ok
2013-05-10 14:43:06 <vkrishn> a common app would be more nice
2013-05-10 14:43:26 <ncopa> use symlinks then
2013-05-10 14:43:36 <ncopa> i like spru<tab> < file
2013-05-10 14:43:48 <ncopa> no args makes it quick to use
2013-05-10 14:44:00 <vkrishn> so set a default to sprunge
2013-05-10 14:44:09 <Barthalion> 
ACTION uses sprunge shell function from oh-my-zsh

2013-05-10 14:46:05 <ncopa> http://dpaste.de/NWahh/
2013-05-10 14:46:05 <vkrishn> the app would be shell script
2013-05-10 14:46:44 <ncopa> what do you think abute that http://dpaste.de/NWahh/
2013-05-10 14:46:49 <ncopa> it was slower than sprunge
2013-05-10 14:47:09 <ncopa> i think pastebinc is a generic app in c
2013-05-10 14:47:18 <vkrishn> strunge ?
2013-05-10 14:47:47 <ncopa> ups
2013-05-10 14:48:59 <ncopa> was this better:  "http://dpaste.de/ShhSF/"
2013-05-10 14:49:06 <vkrishn> but I guess shell is preferred here
2013-05-10 14:49:24 <vkrishn> just to make a quick hack and add own pasteboard
2013-05-10 14:49:40 <vkrishn> or maybe a conf file with list of pasteboards
2013-05-10 14:49:55 <ncopa> thats what pastebinc does
2013-05-10 14:50:03 <ncopa> i just dont like the glib dep
2013-05-10 14:50:17 <vkrishn> then lean shell
2013-05-10 14:50:34 <ncopa> http://dpaste.de/3dPNB/
2013-05-10 14:50:37 <ncopa> hm
2013-05-10 14:50:51 <ncopa> i dont like that dpaste api encapsulate it in ""
2013-05-10 14:51:15 <ncopa> and dpaste thinks its python...
2013-05-10 14:51:42 <ncopa> http://dpaste.de/ShhSF/raw/
2013-05-10 14:51:47 <ncopa> there is a raw too
2013-05-10 14:51:57 <ncopa> i think i would prefer the raw mode
2013-05-10 14:52:19 <ncopa> because then can you 'curl <therurl> | git applyl'
2013-05-10 14:52:27 <vkrishn> raw mode good for script fetch
2013-05-10 14:52:36 <ncopa> yes
2013-05-10 14:54:30 <vkrishn> I was also thinking a shorter version
2013-05-10 14:54:40 <vkrishn> like past<tab> < file 2
2013-05-10 14:54:53 <ncopa> pastie
2013-05-10 14:54:55 <vkrishn> so don't have to recall the pasteboard name
2013-05-10 14:55:08 <ncopa> paste
2013-05-10 14:55:21 <ncopa> and it'd pick a random service :)
2013-05-10 14:55:25 <vkrishn> ok with me
2013-05-10 14:55:49 <ncopa> or configiurable
2013-05-10 14:55:57 <ncopa> fell free to make an app of it
2013-05-10 14:56:04 <ncopa> i'm ok with lua, C and shell
2013-05-10 14:56:12 <ncopa> nor ruby or python
2013-05-10 14:56:16 <ncopa> not*
2013-05-10 14:56:31 <vkrishn> deps become huge installs
2013-05-10 14:59:57 <ncopa> yup
2013-05-10 18:37:40 <jmelis> hi guys, any non-interactive way to change a user's password in alpine?
2013-05-10 18:43:50 <Barthalion> What about chpasswd?
2013-05-10 18:46:57 <jmelis> :)
2013-05-10 18:47:03 <jmelis> exactly what I was looking for
2013-05-10 18:47:22 <jmelis> I was writing a small script in ruby (my alpine has ruby installed)
2013-05-10 18:47:24 <jmelis> thanks :)
2013-05-10 18:47:25 <jmelis> btw
2013-05-10 18:47:39 <jmelis> is there any documentation page for these basic system commands?
2013-05-10 18:47:50 <Barthalion> I don't think so
2013-05-10 18:47:55 <jmelis> ok
2013-05-10 18:47:59 <jmelis> thanks :)
2013-05-10 18:48:09 <Barthalion> But don't ask me, I used to google for command I'm looking for
2013-05-10 18:48:20 <jmelis> oh
2013-05-10 18:48:43 <Barthalion> This is what I did now :P
2013-05-10 18:49:11 <Barthalion> "linux change password non-interactively" and then tested found command in Alpine chroot
2013-05-10 21:13:37 <Mp5shooter> could someone package libreswan for me pretty please :P
2013-05-11 00:31:25 <kaniini> ncopa: i propose switching to kmscon (packaged as kmscon-light and kmscon) from fbcon so we can have antialiased console
2013-05-11 00:31:28 <kaniini> :D
2013-05-11 00:31:42 <kaniini> kmscon-light would be kmscon built without ttf or opengl support
2013-05-11 00:32:04 <kaniini> this is also a big win for when we fix utf-8 in uclibc
2013-05-11 00:32:10 <kaniini> because we'll have a console which can display CJK
2013-05-11 00:32:11 <kaniini> lol
2013-05-11 03:41:07 <Mp5shooter> why is logrotate a dependency of dovecot packages? o.O
2013-05-11 17:16:30 <fcolista> Barthalion, are you around?
2013-05-11 17:17:10 <Barthalion> o/
2013-05-11 17:17:30 <fcolista> hi
2013-05-11 17:17:42 <fcolista> i saw that you're the maintainer of ffmpeg on archlinux
2013-05-11 17:17:57 <fcolista> is the latest package supoprting akamai patch for FMLE auth?
2013-05-11 17:18:03 <Barthalion> Minor, but yes
2013-05-11 17:18:03 <fcolista> is pretty ot :)
2013-05-11 17:18:20 <fcolista> ok, thx
2013-05-11 17:18:35 <Barthalion> It's not answer for your question :p
2013-05-11 17:18:41 <Barthalion> Could you link to this patch?
2013-05-11 17:18:51 <fcolista> umh, not now really
2013-05-11 17:18:55 <fcolista> i can send it to you
2013-05-11 17:18:56 <fcolista> l8r
2013-05-11 17:19:21 <Barthalion> No need, I'll check git log then
2013-05-11 17:20:02 <Barthalion> I don't see anything relevant, so I guess no
2013-05-11 17:25:13 <Barthalion> fcolista: Ah, btw, in Arch we're numbering pkgrel from 1, not 0
2013-05-11 21:52:16 <Mp5shooter> isolated:~# iptables -t nat -A POSTROUTING -s 10.0.0.0/24 -d 0/0 -j SNAT --to-source 192.95.14.15
2013-05-11 21:52:16 <Mp5shooter> iptables: No chain/target/match by that name.
2013-05-11 21:52:18 <Mp5shooter> wtf
2013-05-11 21:52:21 <Mp5shooter> it's in the list
2013-05-11 22:04:29 <Mp5shooter> why is irclogger_com in here
2013-05-11 23:41:02 <duck_> how can I get some APKBUILDs added to alpine edge?
2013-05-11 23:41:40 <kaniini> duck_, submit them to the alpine-devel list
2013-05-11 23:43:10 <duck_> I've noticed some strange things from a desktop perspective, xfce not having printer configuration manager, pidgin not supporting irc, nobody packaging libotr/pidgin-otr, lack of x2go, lack of any cups printer drivers
2013-05-11 23:43:38 <duck_> Is there an interest in those packages/features, or have they been neglected due to lack of resources?
2013-05-12 00:03:12 <duck_> how can I force abuild to build as root?
2013-05-12 00:33:35 <kaniini> duck_: there is interest, but, it's not enough for any of us to do it :P
2013-05-12 00:33:47 <kaniini> duck_: most of us are running alpine on server
2013-05-12 00:34:04 <duck_> I want to run alpine on my phone, beagle bone, laptop, and server
2013-05-12 00:34:14 <kaniini> your phone is x86 ?
2013-05-12 00:34:35 <duck_> No, I expect I'll have to start a cross compiler at some point.
2013-05-12 00:34:57 <kaniini> well, setting up an arm builder should be easy
2013-05-12 00:35:08 <kaniini> i haven't had access to arm hardware that i can afford to brick :P
2013-05-12 00:35:30 <duck_> If I give you shell access to an arm box, would you like to set it up as a build server?
2013-05-12 00:35:55 <duck_> If it gets bricked, I'm okay with that.
2013-05-12 00:36:05 <duck_> I have 3 spares
2013-05-12 02:31:21 <kaniini> duck_: sounds intriguing
2013-05-12 03:29:41 <kaniini> brb
2013-05-12 05:00:55 <Mp5shooter> gonna try packaging libreswan myself lo
2013-05-12 05:02:10 <Mp5shooter> hm
2013-05-12 05:02:19 <Mp5shooter> kaniini: http://puu.sh/2RK7R.txt
2013-05-12 05:02:30 <Mp5shooter> what license would i put down for libreswan?
2013-05-12 06:00:09 <Mp5shooter> abuild-apk: Not a member of group abuild
2013-05-12 06:00:09 <Mp5shooter> wat
2013-05-12 06:00:16 <Mp5shooter> i get this when doing "abuild -r"
2013-05-12 06:01:52 <kaniini> add yourself to abuild group duh
2013-05-12 06:02:56 <Mp5shooter> guide said we didn't have to :() but yeah it worked
2013-05-12 06:18:47 <Mp5shooter> ugh
2013-05-12 06:18:50 <Mp5shooter> it won't build
2013-05-12 06:18:52 <Mp5shooter> too tired to deal with this
2013-05-12 06:45:09 <vkrishn> Mp5shooter, I has requested for irclogger_com, I have a slow internet connection that disconnects during regular irc chat.
2013-05-12 06:45:17 <vkrishn> I had*
2013-05-12 06:47:16 <Barthalion> Yeah, I was wondering why he's sitting here
2013-05-12 06:47:18 <Barthalion> Good to know
2013-05-12 14:37:12 <Barthalion> \o/
2013-05-12 17:51:13 <grimeton> i just upgraded my firewall from 3.6.6 to 3.8.12 and now it's not booting anymore
2013-05-12 17:51:34 <grimeton> coming up with mdadm complaints and the usb keyboard isn't working anymore in the console
2013-05-12 17:51:37 <grimeton> kinda messed up
2013-05-12 17:54:36 <Mp5shooter> guys
2013-05-12 17:54:43 <Mp5shooter> docbook-xml = docbook-dtds?
2013-05-12 17:59:19 <Barthalion> I wouldn't say so
2013-05-12 17:59:33 <Barthalion> But who knows
2013-05-12 18:07:12 <Mp5shooter> :/
2013-05-12 18:09:43 <grimeton> how do i change the rootfs of allpine?
2013-05-12 18:09:49 <grimeton> just chaning the init line didn't work
2013-05-12 18:13:10 <Barthalion> What do you mean? Partition to boot from?
2013-05-12 18:13:24 <grimeton> yeah it was a raid1 partition before the upgrade
2013-05-12 18:13:27 <Barthalion> Apart from bootloader configuration, you need to change fstab
2013-05-12 18:13:39 <grimeton> and now the raid isn'tcoming up, so i tried to boot via the scsi device directly
2013-05-12 18:13:43 <grimeton> (sda)
2013-05-12 18:13:49 <grimeton> but still the same error
2013-05-12 18:13:57 <grimeton> i rebuilt initramfs with mkinitfs
2013-05-12 18:14:12 <grimeton> if it would concern about fstab i'd be happy
2013-05-12 18:14:15 <grimeton> don't come that far
2013-05-12 18:14:33 <Barthalion> uh
2013-05-12 18:14:52 <grimeton> the initramfs tries to mount root and dies
2013-05-12 18:15:02 <grimeton> and the usb keyboard isn't working on the initramfs console
2013-05-12 18:15:08 <Barthalion> Sounds like bugs
2013-05-12 18:15:16 <grimeton> indeed
2013-05-12 18:15:23 <grimeton> the main bug is the version diff of mdadm
2013-05-12 18:15:28 <grimeton> that's why the raid stopped working
2013-05-12 18:21:33 <grimeton> mounting root
2013-05-12 18:21:35 <grimeton> and that's it
2013-05-12 18:23:02 <grimeton> it's my firewall
2013-05-12 18:25:21 <grimeton> where does initramfs get the root device from?
2013-05-12 18:27:02 <grimeton> thanks for your help...
2013-05-12 19:24:41 <kaniini> i've had that one laying around in my aports tree for a while :P
2013-05-12 20:09:24 <Mp5shooter> hm
2013-05-12 20:09:28 <Mp5shooter> so apparently
2013-05-12 20:09:28 <Mp5shooter> [14:20:59] <bleve> Mp5shooter: xmlto needs docbook-dtds to do anything sensible.
2013-05-12 20:09:28 <Mp5shooter> [14:21:11] <bleve> Mp5shooter: so it sounds like your distro is broken.
2013-05-12 20:09:34 <Mp5shooter> :*
2013-05-12 20:09:35 <Mp5shooter> :(
2013-05-12 20:16:41 <kaniini> Mp5shooter: apk add docbook-xml
2013-05-12 20:51:36 <Mp5shooter> kaniini: did.. didn't work though
2013-05-12 20:52:13 <Mp5shooter> libreswan is so complicated :(
2013-05-12 21:49:15 <Mp5shooter> hey Barthalion
2013-05-12 21:49:22 <Mp5shooter> i'll give you $5 if you package libreswan :O
2013-05-12 21:49:57 <kaniini> Mp5shooter: i'll do it for $25!
2013-05-12 21:50:33 <Mp5shooter> that's a tad much there chap :()
2013-05-12 22:41:47 <Mp5shooter> SIGH
2013-05-12 22:41:49 <Mp5shooter> kaniini: $15
2013-05-12 22:42:04 <Mp5shooter> and it has to work
2013-05-12 22:42:04 <Mp5shooter> lolz
2013-05-12 23:04:24 <kaniini> nope
2013-05-12 23:05:05 <kaniini> btw $25 is a good price, considering i usually charge a lot more for consulting-related services
2013-05-12 23:06:48 <Mp5shooter> :|
2013-05-12 23:13:20 <Mp5shooter> is it gonna be super perfect kaniini
2013-05-12 23:48:33 <Mp5shooter> man
2013-05-12 23:48:43 <Mp5shooter> i can't see someone paying to have something _packaged_
2013-05-12 23:56:01 <Mp5shooter> kaniini: FINE
2013-05-12 23:56:06 <Mp5shooter> but it needs to be perfect :3
2013-05-12 23:57:53 <Mp5shooter> actually
2013-05-12 23:57:53 <Mp5shooter> no
2013-05-12 23:58:01 <Mp5shooter> i'm not paying $25 for a package
2013-05-13 00:00:29 <kaniini> good, because my rate has increased to $50
2013-05-13 00:11:58 <Mp5shooter> lol
2013-05-13 00:12:08 <Mp5shooter> you're taking advantage of my ipsec desperation :(
2013-05-13 00:12:22 <kaniini> you don't say
2013-05-13 00:13:23 <Mp5shooter> do you know what libreswan has that openswan doesn't (aside from support for a bunch more things)? certificates and keys and stuff are all stored in an nss db <:~~~~~()
2013-05-13 00:13:28 <Mp5shooter> not in cute little folders
2013-05-13 00:13:34 <kaniini> sounds boring
2013-05-13 00:14:09 <Mp5shooter> i like it
2013-05-13 00:33:21 <Mp5shooter> DIE DOCBOOKS
2013-05-13 00:33:22 <Mp5shooter> DIE
2013-05-13 00:33:23 <Mp5shooter> DIEEEEE
2013-05-13 00:39:37 <Mp5shooter> ffs
2013-05-13 00:39:46 <Mp5shooter> http://forums.freebsd.org/showthread.php?t=16752
2013-05-13 00:39:49 <Mp5shooter> the solution is in ehre
2013-05-13 00:39:57 <Mp5shooter> but idk how to fucking put that in the apkbuild
2013-05-13 00:40:36 <Mp5shooter> this is too much work for something that probably only i will use
2013-05-13 00:41:27 <Mp5shooter> 
ACTION is just going to compile from source

2013-05-13 00:44:15 <Mp5shooter> grrrrr
2013-05-13 00:55:38 <Mp5shooter> http://puu.sh/2SlhF.png
2013-05-13 00:55:39 <Mp5shooter> wtf
2013-05-13 00:57:57 <Mp5shooter> man
2013-05-13 01:06:10 <Mp5shooter> how do i install non-busybox install
2013-05-13 01:06:11 <Mp5shooter> ??
2013-05-13 05:54:20 <fcolista> hi
2013-05-13 05:54:35 <fcolista> Barthalion, i think you worte something here that i lost :/
2013-05-13 05:54:39 <fcolista> *wrote
2013-05-13 05:54:45 <fcolista> i've to check the log
2013-05-13 05:54:58 <fcolista> Oh, btw...morning to all :)
2013-05-13 07:34:46 <clandmet1r> good morning
2013-05-13 07:36:00 <clandmeter> ncopa: do you mind if I upgrade linux-vanilla to 3.9.2?
2013-05-13 07:36:58 <ncopa> clandmeter: no that would be good
2013-05-13 07:37:01 <ncopa> i am working on it though
2013-05-13 07:37:08 <ncopa> and i have a config
2013-05-13 07:38:30 <Barthalion> Mp5shooter: Just install coreutils
2013-05-13 07:38:49 <Barthalion> Mp5shooter: It should take precedence
2013-05-13 07:39:23 <Barthalion> fcolista: Nothing important, I've seen your packages in AUR - in Arch pkgrel should start from 1, but in fact it doesn't matter
2013-05-13 07:39:45 <fcolista> Ok Barthalion
2013-05-13 07:39:50 <fcolista> thx for feedback
2013-05-13 07:40:02 <fcolista> I'll change it
2013-05-13 07:41:08 <ncopa> i have updated git.a.o to v2.6
2013-05-13 07:44:25 <ncopa> clandmeter: i am working on linux-vanilla 3.9.y now
2013-05-13 07:44:46 <ncopa> i did the config last week but it failed to build
2013-05-13 07:44:54 <ncopa> i just relized that it needs 'bc' in makedepends
2013-05-13 07:48:41 <Barthalion> Ouch, it's libre-swan, not lib-reswan
2013-05-13 07:48:49 <Barthalion> Misleading name, I really though it's a library
2013-05-13 07:49:13 <ncopa> heh
2013-05-13 07:49:14 <ncopa> same here
2013-05-13 07:53:10 <clandmeter> ncopa: yes it does
2013-05-13 07:53:14 <clandmeter> i have it done
2013-05-13 07:53:45 <clandmeter> ncopa: i would like to add a patch to support my dvb adapter. would that be ok?
2013-05-13 07:53:59 <ncopa> kernelconfig.x86 diff?
2013-05-13 07:54:06 <ncopa> yes, thats ok
2013-05-13 07:54:44 <clandmeter> i think my diff for 3.9 needs some cleanup
2013-05-13 07:55:01 <clandmeter> lots of new devices which will prob never use on alpine.
2013-05-13 07:55:10 <clandmeter> did you remove them?
2013-05-13 07:55:26 <clandmeter> most of them are sensors
2013-05-13 07:55:39 <ncopa> http://sprunge.us/gihZ
2013-05-13 07:56:32 <clandmeter> yours is much cleaner. I pressed M for each possible option...
2013-05-13 07:57:44 <ncopa> you could wait til i push the updated apkbuild and take it from there, or you could send me a link for the patch
2013-05-13 07:57:54 <ncopa> and I'll include it
2013-05-13 07:58:05 <clandmeter> is sprunge ok for patches?
2013-05-13 07:58:09 <ncopa> yeah
2013-05-13 07:58:12 <ncopa> well
2013-05-13 07:58:17 <ncopa> i'll wget it and add it think
2013-05-13 07:58:29 <clandmeter> i think the patch is too big
2013-05-13 07:58:39 <ncopa> big? hmm.
2013-05-13 07:58:50 <ncopa> how big is it?
2013-05-13 07:59:00 <ncopa> will it enable config options?
2013-05-13 07:59:09 <clandmeter> yes
2013-05-13 07:59:12 <clandmeter> 299.1K May 12 21:53 linux-3.9-dvbsky-130507.patch
2013-05-13 07:59:12 <ncopa> :-/
2013-05-13 07:59:23 <clandmeter> it adds 3 devices
2013-05-13 07:59:30 <clandmeter> or 4
2013-05-13 07:59:43 <ncopa> not too big patch
2013-05-13 07:59:44 <ncopa> its ok
2013-05-13 07:59:50 <ncopa> since its you ;)
2013-05-13 07:59:56 <clandmeter> hehe
2013-05-13 08:00:03 <clandmeter> but too big for sprunge i guess
2013-05-13 08:00:13 <ncopa> we could try..
2013-05-13 08:00:30 <ncopa> or you could gzip it and pipe to sprunge
2013-05-13 08:00:41 <clandmeter> ill give you a link
2013-05-13 08:02:34 <clandmeter> ncopa: https://dl.dropboxusercontent.com/u/18871200/linux-3.9-dvbsky-130507.patch
2013-05-13 08:03:53 <ncopa> do you mind if we add it to grsec kernel later instead?
2013-05-13 08:04:05 <ncopa> because otherwise 'vanilla' is no longer 'vanilla'
2013-05-13 08:04:29 <ncopa> can it be built as an out-of-tree module?
2013-05-13 08:05:04 <clandmeter> they have a media tree
2013-05-13 08:06:09 <clandmeter> didnt look if i can build oot
2013-05-13 08:06:46 <ncopa> do you think they are in progress to get it into mainline?
2013-05-13 08:06:49 <ncopa> looks so..
2013-05-13 08:06:52 <Barthalion> Mp5shooter: Take a look at testing/libreswan
2013-05-13 08:07:03 <kaniini> Barthalion: you should tell him your paypal
2013-05-13 08:07:07 <kaniini> Barthalion: he did offer $5!
2013-05-13 08:07:11 <Barthalion> Mp5shooter: I wasn't sure what to do with kernel module
2013-05-13 08:08:04 <Barthalion> huh
2013-05-13 08:08:18 <Barthalion> ach, checksums
2013-05-13 08:08:42 <clandmeter> ncopa: there have been some discussions about mainline, but there were some politics involved. not sure about the status now.
2013-05-13 08:08:55 <ncopa> ok, maybe i was too early to ACK it :-/
2013-05-13 08:08:58 <ncopa> so
2013-05-13 08:09:08 <ncopa> do we bogehr about grsec for xbmc?
2013-05-13 08:09:12 <ncopa> bother*
2013-05-13 08:09:55 <clandmeter> xbmc and this patch you mean?
2013-05-13 08:10:13 <ncopa> you want this for your xbmc?
2013-05-13 08:10:18 <clandmeter> no
2013-05-13 08:10:20 <clandmeter> tvheadend
2013-05-13 08:10:29 <clandmeter> dvb server
2013-05-13 08:10:31 <ncopa> ok
2013-05-13 08:11:03 <clandmeter> if its not inside vanilla, i can build it myself local and wait for grsec.
2013-05-13 08:11:41 <clandmeter> i didnt see any patch yet for grsec
2013-05-13 08:12:09 <clandmeter> my card arrives tomorrow i think :) so i *need* to test it :)
2013-05-13 08:13:03 <Barthalion> kaniini: $5 aren't worth making PayPal account :p
2013-05-13 08:15:19 <kaniini> :P
2013-05-13 08:15:38 <ncopa> clandmeter: ok i think we'll be better off to kepp vanilla kernel without patches
2013-05-13 08:16:30 <clandmeter> np
2013-05-13 08:17:35 <clandmeter> you can push it.
2013-05-13 08:18:30 <kaniini> http://grsecurity.net/news.php#smap
2013-05-13 08:18:41 <kaniini> man, spender is arrogant sometimes
2013-05-13 08:20:49 <Barthalion> \o/
2013-05-13 08:21:42 <clandmeter> Barthalion: now you spoiled that 5$ for me...
2013-05-13 08:22:57 <Barthalion> :D
2013-05-13 08:25:34 <Barthalion> Mp5shooter: Ah, I can't promise it works, because I'm not familiar with IPSec and how it should work
2013-05-13 11:51:44 <vkrishn> ncopa, http://dpaste.de/zjDZk/
2013-05-13 11:52:08 <vkrishn> http://dpaste.de/FytEa/
2013-05-13 11:52:22 <vkrishn> this will make dpaste output same as sprunge
2013-05-13 11:53:18 <vkrishn> correction: libpath=/usr/bin
2013-05-13 11:54:01 <ncopa> i think we can do it without using multiple files
2013-05-13 11:54:12 <vkrishn> if you prefer random, just uncomment it
2013-05-13 11:54:27 <ncopa> nah
2013-05-13 11:54:30 <vkrishn> yes, but sprunge BC needs be taken care of
2013-05-13 11:55:03 <vkrishn> all others can be in /usr/bin/pastie
2013-05-13 11:55:25 <vkrishn> just note 'paste' already exists
2013-05-13 11:55:56 <vkrishn> sprunge backward compat...
2013-05-13 11:57:36 <vkrishn> paste(cmd) - merge lines of files
2013-05-13 11:58:59 <vkrishn> some alternate name: bleep, scoop, pastebus, textshare
2013-05-13 11:59:06 <vkrishn> ;)
2013-05-13 12:01:14 <clandmeter> ncopa: 2 things about vanilla
2013-05-13 12:01:56 <clandmeter> first, you added -j1 (i guess to check the build error regarding missing bc)
2013-05-13 12:02:06 <ncopa> aw
2013-05-13 12:02:52 <clandmeter> and dev package could need some cleanup
2013-05-13 12:04:02 <clandmeter> i dont know about all the parts, but the drivers/media header files are incorrect.
2013-05-13 12:06:43 <clandmeter> it seems to link to an ubuntu build script but the link is broken.
2013-05-13 12:09:01 <ncopa> vkrishn: what do you think about this: http://dpaste.de/tVJTg/
2013-05-13 12:09:18 <ncopa> maybe with your sed lines for removing the " and append /raw
2013-05-13 12:09:43 <ncopa> clandmeter: yes i copied that from ubuntu makefile some time ago
2013-05-13 12:09:49 <ncopa> i suppose things has changed since then
2013-05-13 12:12:35 <vkrishn> the " issue would make output similar, useful to output somewhere
2013-05-13 12:14:33 <vkrishn> #L24 would take care of usage like: $ pastie < file 1 ?
2013-05-13 12:15:24 <ncopa> yes
2013-05-13 12:15:26 <ncopa> well
2013-05-13 12:15:37 <ncopa> it will check how the program was called
2013-05-13 12:15:52 <vkrishn> ok
2013-05-13 12:15:55 <ncopa> so we can just symlink dpaste to sprunge
2013-05-13 12:15:58 <ncopa> same prog
2013-05-13 12:16:00 <ncopa> 1 file
2013-05-13 12:16:15 <clandmeter> ncopa: looks like i looked in the wrong spot about -j1. sorry.
2013-05-13 12:16:37 <vkrishn> no 'exec' for dpaste?
2013-05-13 12:16:47 <ncopa> because it needs to be piped
2013-05-13 12:17:45 <vkrishn> ok
2013-05-13 12:17:46 <ncopa> hm
2013-05-13 12:17:51 <ncopa> looks like you can exec and pipe...
2013-05-13 12:17:52 <ncopa> nice
2013-05-13 12:18:44 <ncopa> but curl insists to print the download stats
2013-05-13 12:18:46 <ncopa> i dont want that
2013-05-13 12:18:56 <vkrishn> -s
2013-05-13 12:19:37 <ncopa> yeah
2013-05-13 12:19:39 <ncopa> http://dpaste.de/C0cHS/raw/
2013-05-13 12:19:42 <ncopa> that seems to work
2013-05-13 12:20:49 <ncopa> vkrishn: look sok? I push it like that?
2013-05-13 12:20:53 <vkrishn> now I would start to see if more similar pastebins can be added
2013-05-13 12:20:55 <vkrishn> thanks
2013-05-13 12:21:11 <ncopa> i can push it as is
2013-05-13 12:21:15 <ncopa> and you can work on it from there ;)
2013-05-13 12:21:21 <vkrishn> ok
2013-05-13 12:27:06 <ncopa> \o/
2013-05-13 12:27:12 <vkrishn> no /usr/bin/pastie ?
2013-05-13 12:27:14 <ncopa> 3.9.2 pax patch
2013-05-13 12:27:26 <ncopa> vkrishn: feel free to work on it
2013-05-13 12:27:33 <vkrishn> ok
2013-05-13 12:27:38 <ncopa> i have other priorities atm ;)
2013-05-13 12:27:53 <ncopa> i think we will have grsecurity patch for 3.9.2 any moment
2013-05-13 12:28:27 <vkrishn> yes
2013-05-13 13:25:44 <clandmeter> ncopa: what happend to ~/.cache directory?
2013-05-13 13:26:26 <clandmeter> its deprecated in favor of /var/lib/packages ?
2013-05-13 13:26:45 <ncopa> both are supported
2013-05-13 13:26:47 <ncopa> basically
2013-05-13 13:27:02 <ncopa> if you specify REPODEST, the ~/cache will not be used
2013-05-13 13:27:15 <ncopa> ~/.cache
2013-05-13 13:27:22 <clandmeter> ah ok
2013-05-13 14:19:38 <clandmeter> ncopa: is it possible to build an alpine iso with another repo as the system itself? like if the system is edge, but i want to build an iso for 2.6?
2013-05-13 14:22:30 <clandmeter> i see i can add a repo, but i dont see an option to ignore the system's repo.
2013-05-13 14:24:13 <ncopa> apk add --repositories /dev/null --repository http://somerepo-thats-not-in-sys-repos
2013-05-13 14:24:52 <clandmeter> in alpine iso?
2013-05-13 14:26:05 <ncopa> not sure havent looked at the makefile
2013-05-13 14:26:11 <ncopa> but thats the principle
2013-05-13 14:26:27 <ncopa> maybe make APK="apk --repositories /dev/null"
2013-05-13 14:26:29 <ncopa> or similar
2013-05-13 14:27:15 <Mp5shooter> <Barthalion> Mp5shooter: Just install coreutils
2013-05-13 14:27:19 <Mp5shooter> oh that's what it was
2013-05-13 14:27:34 <Mp5shooter> <Barthalion> Mp5shooter: Take a look at testing/libreswan
2013-05-13 14:27:36 <Mp5shooter> oh ok
2013-05-13 14:27:45 <Mp5shooter> <Barthalion> Mp5shooter: I wasn't sure what to do with kernel module
2013-05-13 14:27:52 <Mp5shooter> if you meant netkey or klips, netkey
2013-05-13 14:27:55 <Mp5shooter> cuz it's native
2013-05-13 14:28:25 <Mp5shooter> 
ACTION shall test it :O

2013-05-13 14:39:06 <Mp5shooter> hm weird
2013-05-13 14:39:13 <Mp5shooter> did an upgrade and terminal colors went missing
2013-05-13 14:41:19 <Mp5shooter> hey Barthalion, what did you name the init script for it
2013-05-13 16:07:31 <Barthalion> Mp5shooter: Err, I didn't
2013-05-13 16:07:37 <Barthalion> Mp5shooter: I'll add it later if it can wait
2013-05-13 16:07:42 <Mp5shooter> oh :p that's fine
2013-05-13 16:07:44 <Mp5shooter> thanks
2013-05-13 16:07:51 <Mp5shooter> also ipsec.conf wasn't put in /etc
2013-05-13 16:08:15 <Mp5shooter> did you come across that xml problem btw?
2013-05-13 16:08:23 <Mp5shooter> when you were building?
2013-05-13 16:11:33 <Barthalion> No, I didn't
2013-05-13 16:11:51 <Barthalion> I just added needed makedeps
2013-05-13 16:54:10 <Mp5shooter> can i see your APKBUILD Barthalion
2013-05-13 16:55:26 <Barthalion> http://git.alpinelinux.org/cgit/aports/tree/testing/libreswan
2013-05-13 16:56:14 <Mp5shooter> o
2013-05-13 16:56:24 <Mp5shooter> i thought coreutils would've been a makedepend
2013-05-13 17:02:51 <Barthalion> I probably read on their website about dependency
2013-05-13 18:17:44 <Mp5shooter> lmao
2013-05-13 18:17:48 <Mp5shooter> tiny ptoblem Barthalion
2013-05-13 18:17:50 <Mp5shooter> problem*
2013-05-13 18:17:58 <Mp5shooter> isolated:~# /usr/usr/local/sbin/ipsec
2013-05-13 18:18:03 <Mp5shooter> this is where it went
2013-05-13 18:18:03 <Mp5shooter> xD
2013-05-13 18:21:29 <Mp5shooter> sec
2013-05-13 18:24:18 <Mp5shooter> trying something real quick before i have to go to work
2013-05-13 18:26:45 <Barthalion> crap
2013-05-13 18:26:47 <Barthalion> Well :D
2013-05-13 18:27:07 <Barthalion> I'll fix it tommorow
2013-05-13 18:27:53 <Mp5shooter> Barthalion
2013-05-13 18:27:55 <Mp5shooter> http://puu.sh/2SL3G.txt
2013-05-13 18:27:59 <Mp5shooter> can you try that APKBUILD
2013-05-13 18:28:02 <Mp5shooter> i think i fixed it
2013-05-13 18:28:09 <Mp5shooter> borrowed some stuff from openswan's apkbuild
2013-05-13 18:28:10 <Mp5shooter> also
2013-05-13 18:28:22 <Mp5shooter> http://puu.sh/2SL4L.txt
2013-05-13 18:28:25 <Mp5shooter> ipsec.initd
2013-05-13 18:28:41 <Mp5shooter> ok i have to go to work but i'll get on my laptop there when i'm finished doing stuff
2013-05-13 18:28:41 <Mp5shooter> :D
2013-05-13 18:28:42 <Mp5shooter> ty
2013-05-13 18:28:43 <Mp5shooter> you're the best
2013-05-13 18:42:02 <build64-2-5> files from v2.5.4-72-g54a1652 uploaded
2013-05-13 19:10:38 <Mp5shooter> ok i'm back
2013-05-13 19:41:00 <Mp5shooter> huh
2013-05-13 19:42:12 <Mp5shooter> http://pastebin.com/Ys1SxFcm
2013-05-13 19:42:22 <Mp5shooter> Barthalion: ever seen that "redirecting to" thing?
2013-05-13 19:43:16 <Mp5shooter> also, they just threw up version 3.3
2013-05-13 19:52:09 <clandmeter> looks like the init script from openswan is not compatible
2013-05-13 19:53:06 <Mp5shooter> it's just the base init script + the three lines filled in
2013-05-13 19:53:18 <Mp5shooter> name, somethinsomethin, pidfile
2013-05-13 19:53:25 <Mp5shooter> :s
2013-05-13 19:53:51 <Mp5shooter> libreswan dev:
2013-05-13 19:53:52 <Mp5shooter> [15:21:50] LetoTo:	 Mp5shooter: your paths are broken
2013-05-13 19:53:53 <Mp5shooter> [15:21:54] LetoTo:	 your init system is prob very nonstandard
2013-05-13 20:02:56 <Barthalion> OpenRC is rather standard
2013-05-13 20:51:35 <vkrishn> Barthalion, is strongswan different than libreswan ?
2013-05-13 20:52:38 <Mp5shooter> libreswan is a fork of openswan which is a fork of strong swan vkrishn
2013-05-13 20:53:33 <vkrishn> libreswan<-openswan<-freeswan
2013-05-13 20:53:54 <Mp5shooter> o
2013-05-13 20:53:55 <Mp5shooter> rly?
2013-05-13 20:54:01 <Mp5shooter> thought it was strong
2013-05-13 20:54:48 <Mp5shooter> oh ok
2013-05-13 20:55:04 <Mp5shooter> strong swan and openswan both came from freeswan
2013-05-13 20:55:26 <vkrishn> freeswan common ancestor
2013-05-13 20:56:11 <Mp5shooter> i wish the libreswan package would work :(
2013-05-13 20:56:21 <vkrishn> strongswan has a good implementation in sophos
2013-05-13 20:56:54 <vkrishn> just is case if difficult, than we already have openswan
2013-05-13 20:57:12 <vkrishn> compare with strongswan
2013-05-13 20:57:20 <vkrishn> see what fits
2013-05-13 20:57:33 <Mp5shooter> libreswan is a lot better than openswan
2013-05-13 20:57:37 <Mp5shooter> also openswan will probably die
2013-05-13 20:57:49 <Mp5shooter> because all of the openswan devs left openswan to fork it
2013-05-13 20:58:02 <Mp5shooter> because of some legal battle with one of the other openswan devs
2013-05-13 20:58:33 <vkrishn> check protocol implementations, strongswan seems more modular
2013-05-13 21:22:47 <Mp5shooter> [17:21:57] LetoTo:	 Mp5shooter: never heard of it. you can look at packaging/utils/lswan_detect.sh to see how we detect initsystems
2013-05-13 21:22:47 <Mp5shooter> [17:22:13] Mp5shooter:	 ok
2013-05-13 21:22:47 <Mp5shooter> [17:22:24] LetoTo:	 then copy initsystems/XXX to initsystems/openrc and fix it up to do what's right for your init system
2013-05-13 21:22:48 <Mp5shooter> mmm
2013-05-13 22:13:51 <Mp5shooter> hey guys
2013-05-13 22:13:58 <Mp5shooter> does alpine use "sysvinit"?
2013-05-13 22:17:17 <Mp5shooter> kaniini do you know
2013-05-13 22:21:34 <kaniini> no
2013-05-13 22:21:41 <kaniini> it uses openrc
2013-05-13 22:22:28 <Mp5shooter> oh ok
2013-05-13 22:22:34 <Mp5shooter> i see
2013-05-13 22:26:33 <Mp5shooter> getting libreswan working is taking more work than i thought
2013-05-13 22:26:38 <Mp5shooter> i should've just given you the $25 lolz
2013-05-13 23:00:27 <Mp5shooter> [18:28:03] <LetoTo> so write a patch to let us know 1) how to recognise that init system, 2) how to use it
2013-05-13 23:00:34 <Mp5shooter> hell
2013-05-13 23:00:37 <Mp5shooter> idk how it works :|
2013-05-13 23:11:29 <Mp5shooter> ffs
2013-05-13 23:11:32 <Mp5shooter> this is so much of a hassle
2013-05-14 01:59:02 <Guest95457> can someone assist me with building a package?
2013-05-14 01:59:16 <Guest95457> WARNING: Ignoring /home/user/.cache/abuild/testing/x86_64/APKINDEX.tar.gz: No such file or directory
2013-05-14 01:59:16 <Guest95457> ERROR: You tried to add a non-repository package to system, but it would be lost on next reboot. Enable package caching (apk cache --help) or use --force if you know what you are doing.
2013-05-14 02:46:02 <kaniini> are you trying to run from ram?
2013-05-14 02:49:05 <super_duck> is there any other way to run on a diskless system?
2013-05-14 02:49:58 <super_duck> I'm trying to build some packages as proper Alpine packages.  I built them standalone, but I want apk to track them.
2013-05-14 02:50:26 <super_duck> Once I get the hang of the build system, I'll work on getting my ARM system ready for the bootstrapping
2013-05-14 02:51:41 <kaniini> you have to have a disk to build packages...
2013-05-14 02:51:54 <kaniini> the code is very insistant on that :P
2013-05-14 02:51:55 <super_duck> what? why? that sounds stupid
2013-05-14 02:52:13 <super_duck> does tmpfs make it happy, or just ext4?
2013-05-14 02:53:30 <kaniini> well, a filesystem that is permanent
2013-05-14 02:53:48 <kaniini> and, why is that stupid
2013-05-14 02:53:49 <super_duck> unfortunately, I don't think I have that at the moment
2013-05-14 02:53:58 <kaniini> you need a place to store the packages you build...
2013-05-14 02:54:03 <super_duck> I do, its called RAM
2013-05-14 02:54:20 <kaniini> and when you reboot...?
2013-05-14 02:54:33 <super_duck> I'll hopefully have it emailed to you guys to commit before then
2013-05-14 02:54:49 <super_duck> I'm running a stateless setup
2013-05-14 02:55:02 <kaniini> well building packages isn't stateless
2013-05-14 02:55:08 <kaniini> abuild has a repo
2013-05-14 02:55:30 <super_duck> so I have lots of RAM, and as long as the electricity flows, it will maintain state....so why does abuild care?
2013-05-14 02:55:42 <kaniini> it doesn't, but apk does
2013-05-14 02:55:54 <kaniini> because apk realizes you are running from ram
2013-05-14 02:56:14 <kaniini> and it cannot recreate the ruleset from scratch if it depends on outside packages on unavailable repos
2013-05-14 02:56:16 <super_duck> so how do I spank apk to not have that forced view of the world?
2013-05-14 02:56:30 <kaniini> ``Enable package caching (apk cache --help) or use --force if you know what you are doing.''
2013-05-14 02:56:40 <super_duck> --force didn't help
2013-05-14 02:56:52 <kaniini> you're going to have to modify abuild
2013-05-14 02:56:56 <kaniini> to use --force on apk
2013-05-14 02:57:13 <kaniini> those messages are from apk
2013-05-14 02:57:38 <super_duck> $#%#$
2013-05-14 02:57:39 <super_duck> ok thanks
2013-05-14 02:57:49 <kaniini> you're not really doing a supported thing here
2013-05-14 02:57:58 <kaniini> it is assumed that only finished appliances will run from ram
2013-05-14 02:58:16 <kaniini> whether or not this assumption is valid is up for debate, but it is the present situation
2013-05-14 02:59:10 <kaniini> basically... apk maintains something called the 'world ruleset'
2013-05-14 02:59:18 <super_duck> so I've got the prepare, build, package hooks, how does abuild call apk from those?  Is there a short tutorial showing abuild paths/file integration points with apk?
2013-05-14 02:59:23 <kaniini> which is a list of rules used to put the appliance back together again
2013-05-14 02:59:37 <kaniini> abuild only invokes apk from abuildindex()
2013-05-14 02:59:46 <kaniini> to recreate the repository index
2013-05-14 02:59:53 <kaniini> and also to install dependencies
2013-05-14 03:00:01 <kaniini> i forget what step that is called
2013-05-14 03:00:28 <kaniini> apk will bok at dependencies it won't be able to remember next time around
2013-05-14 03:00:35 <kaniini> another fix
2013-05-14 03:00:37 <kaniini> try uhm
2013-05-14 03:00:41 <kaniini> mkdir /etc/apk/cache
2013-05-14 03:00:48 <kaniini> mount -t tmpfs tmpfs /etc/apk/cache
2013-05-14 03:00:55 <kaniini> then apk update
2013-05-14 03:01:23 <kaniini> that should fool it sufficiently into doing what you want
2013-05-14 03:01:36 <super_duck> chown -R :abuild /etc/apk/cache ?
2013-05-14 03:01:39 <kaniini> no
2013-05-14 03:01:59 <kaniini> abuild ships an suid root wrapper
2013-05-14 03:02:06 <kaniini> for invoking a subset of apk commands
2013-05-14 03:03:24 <super_duck> so I'm still getting the warning and error, its a bit late for me, I'll try again tomorrow
2013-05-14 03:03:29 <super_duck> kaniini, thanks for the help!
2013-05-14 03:03:38 <kaniini> i'll fiddle with it in a vm
2013-05-14 03:04:25 <super_duck> WARNING: Ignoring /home/user/.cache/abuild/testing/x86_64/APKINDEX.tar.gz: No such file or directory
2013-05-14 03:04:33 <super_duck> why is it looking in my user home directory?
2013-05-14 03:04:47 <super_duck> should it look in ~/.cache or /etc/apk/cache ?
2013-05-14 03:08:16 <kaniini> ~/.cache/abuild/testing/... is your abuild-generated repo
2013-05-14 03:08:27 <kaniini> putting a tmpfs on /etc/apk/cache is to resolve the fatal error
2013-05-14 03:08:35 <kaniini> do you get the same exact error or?
2013-05-14 03:08:51 <super_duck> same exact error
2013-05-14 03:09:17 <kaniini> hmm
2013-05-14 03:09:20 <kaniini> not sure :s
2013-05-14 03:10:11 <super_duck> I hate to say this, but rpmbuild works better for me :-/
2013-05-14 03:11:49 <super_duck> kaniini, well I'm off to sleep, I'll worry about this tomorrow
2013-05-14 03:13:50 <kaniini> super_duck: i'd write to alpine-devel, ncopa will know what to do probably (he wrote abuild)
2013-05-14 05:55:34 <ncopa> morning
2013-05-14 05:55:41 <ncopa> fabled: <Guest95457> ERROR: You tried to add a non-repository package to system, but it would be lost on next reboot. Enable package caching (apk cache --help) or use --force if you know what you are doing.
2013-05-14 05:55:55 <ncopa> i think maybe apk-tools is a bit too over protective
2013-05-14 05:56:09 <ncopa> if users want shoot themselves in the foot we maybe should let them
2013-05-14 05:56:29 <kaniini> how... UNIX
2013-05-14 05:56:31 <kaniini> :D
2013-05-14 05:56:48 <ncopa> :)
2013-05-14 05:58:21 <ncopa> whee!! 3.9.2 grsecurity patch is out
2013-05-14 05:58:26 <ncopa> lets do v2.6
2013-05-14 05:58:40 <kaniini> firefox-20 has a fontconfig bug
2013-05-14 05:59:00 <kaniini> i'd like to do a late-breaking patch to fix it in xulrunner
2013-05-14 05:59:11 <ncopa> ok?
2013-05-14 05:59:22 <ncopa> i had to spank xulrunner for find libc
2013-05-14 05:59:35 <ncopa> they do some weird ffi calls to libc
2013-05-14 05:59:45 <ncopa> and had hardcoded it to look for libc.so.6
2013-05-14 06:00:15 <ncopa> should i enable KMS now, everywhere?
2013-05-14 06:01:02 <kaniini> https://admin.fedoraproject.org/updates/fontconfig-2.10.92-3.fc19
2013-05-14 06:01:13 <kaniini> fix is there, actually.
2013-05-14 06:02:06 <ncopa> how does the fontconfig bug manifest itself?
2013-05-14 06:02:45 <ncopa> webfonts
2013-05-14 06:03:03 <kaniini> yeah
2013-05-14 06:03:09 <kaniini> it doesn't use the webfonts :P
2013-05-14 06:03:17 <kaniini> going to google's webfonts gallery is a good testcase
2013-05-14 06:03:19 <kaniini> or github
2013-05-14 06:04:57 <ncopa> this? http://pkgs.fedoraproject.org/cgit/fontconfig.git/commit/?id=e669d0170be4492cd966114122ad5281ec2276de
2013-05-14 06:05:43 <kaniini> yes
2013-05-14 06:06:06 <ncopa> do you want me to add it?
2013-05-14 06:06:27 <Barthalion> actually firefox 21 is almost out
2013-05-14 06:06:29 <kaniini> i got it
2013-05-14 06:07:18 <Barthalion> it should be announced today
2013-05-14 06:07:39 <kaniini> well
2013-05-14 06:07:41 <kaniini> either way
2013-05-14 06:07:45 <kaniini> fontconfig has a regression :P
2013-05-14 06:08:14 <Barthalion> yup
2013-05-14 06:11:05 <kaniini> xulrunner-21 has hit ftp.mozilla.org
2013-05-14 06:11:15 <kaniini> i'll work on that while ncopa works on linux-grsec
2013-05-14 06:11:24 <ncopa> thanks!
2013-05-14 06:13:58 <ncopa> kaniini, the fontconfig patch makes webfotns work
2013-05-14 06:14:01 <ncopa> thanks!!
2013-05-14 06:14:27 <kaniini> yep, i tested it first
2013-05-14 06:15:19 <kaniini> ncopa: btw, did you see that uclibc ldd was causing gobject-introspection to go insane?
2013-05-14 06:15:34 <ncopa> i saw your patch
2013-05-14 06:15:46 <ncopa> i never saw it in action though
2013-05-14 06:16:11 <kaniini> ncopa: it was the bug behind awesome being unbuildable on x86_64
2013-05-14 06:16:31 <ncopa> oh ok..
2013-05-14 06:16:50 <ncopa> i am generaly a bit sceptic to generated code
2013-05-14 06:20:20 <kaniini> >>> xulrunner: getchar.patch
2013-05-14 06:20:22 <kaniini> patching file media/webrtc/trunk/webrtc/system_wrappers/source/spreadsortlib/spreadsort.hpp
2013-05-14 06:20:24 <kaniini> Hunk #1 FAILED at 21 (different line endings).
2013-05-14 06:20:26 <kaniini> are you kidding me
2013-05-14 06:21:20 <ncopa> you mane the patch or the algitbot
2013-05-14 06:21:43 <kaniini> the patch
2013-05-14 06:21:47 <kaniini> 
ACTION grumbles

2013-05-14 06:22:08 <ncopa> the stat patch too
2013-05-14 06:22:40 <ncopa> and http://git.alpinelinux.org/cgit/aports/tree/main/xulrunner/alpine-libc.patch
2013-05-14 06:22:56 <ncopa> someone should tell them how elf is supposed to work...
2013-05-14 06:23:36 <ncopa> it caused really weird issues
2013-05-14 06:23:39 <ncopa> https://bugzilla.mozilla.org/show_bug.cgi?id=843564
2013-05-14 06:23:56 <ncopa> and https://bugzilla.mozilla.org/show_bug.cgi?id=843513
2013-05-14 06:31:37 <kaniini> gonna need to rebase some of these patches soon i think
2013-05-14 06:32:23 <kaniini> running a test build of xulrunner-21 now
2013-05-14 07:05:49 <build64-2-5> files from v2.5.4-74-g8a93cf7 uploaded
2013-05-14 07:09:47 <ncopa> # uname -a
2013-05-14 07:09:47 <ncopa> Linux localhost 3.9.2-0-grsec #1-Alpine SMP Tue May 14 06:28:40 UTC 2013 i686 Linux
2013-05-14 07:11:57 <ncopa> now, lets fight the 3rd party mods...
2013-05-14 07:53:53 <kaniini> /home/kaniini/aports/main/xulrunner/src/mozilla-release/toolkit/mozapps/installer/precompile_cache.js:78: ReferenceError: Services is not defined
2013-05-14 07:53:55 <kaniini> .
2013-05-14 08:15:00 <ncopa> i am reverting the vserver kernel to 3.4.y
2013-05-14 08:15:12 <ncopa> and will ship the vserver iso image with that kernel
2013-05-14 08:15:15 <ncopa> its stable
2013-05-14 08:15:16 <ncopa> it works
2013-05-14 08:15:26 <ncopa> i will also disable losts of stuff in vserver kernel
2013-05-14 08:15:31 <ncopa> like multimedia support
2013-05-14 08:15:44 <ncopa> no webcam support in vserver kernel
2013-05-14 09:12:52 <ncopa> algitbot: build master
2013-05-14 09:14:28 <ncopa> algitbot: build master
2013-05-14 11:30:44 <Barthalion> Damn, I just rebooted my thinclient to 3.8.12
2013-05-14 11:30:48 <Barthalion> And 3.9.2 is uploaded
2013-05-14 11:39:50 <StarWarsFan> hi
2013-05-14 11:40:37 <StarWarsFan> question: what is the best place to put a binary to the system, which should be started via an init script?
2013-05-14 11:40:49 <StarWarsFan> /usr/lib/?
2013-05-14 11:40:59 <StarWarsFan>  /usr/bin/?
2013-05-14 11:41:26 <clandmeter> StarWarsFan: http://en.wikipedia.org/wiki/Filesystem_Hierarchy_Standard
2013-05-14 11:42:17 <ncopa> StarWarsFan, i'd say either /usr/bin or /usr/sbin
2013-05-14 11:43:05 <Barthalion> StarWarsFan: Well, if it won't come from package, /usr/local/{bin,sbin}
2013-05-14 11:44:02 <StarWarsFan> i think, it will be from a package and i'm asking because abuild stops as i tried to use /usr/local/...
2013-05-14 11:44:22 <ncopa> yes
2013-05-14 11:44:23 <Barthalion> As expected :p
2013-05-14 11:44:30 <StarWarsFan> :)
2013-05-14 11:45:02 <ncopa> check the FHS link that clandmeter gave you
2013-05-14 11:45:08 <ncopa> i'll reboot now
2013-05-14 11:45:08 <ncopa> brb
2013-05-14 11:45:16 <StarWarsFan> yeah, i'm on it
2013-05-14 11:45:20 <StarWarsFan> thx a lot!
2013-05-14 11:47:45 <ncopa> Linux ncopa-desktop 3.9.2-0-grsec #1-Alpine SMP Tue May 14 09:13:43 UTC 2013 x86_64 Linux
2013-05-14 11:47:49 <ncopa> so far so good
2013-05-14 11:58:03 <super_duck> ncopa, how do you propose I modify apk-tools to allow me to shoot myself in my foot?
2013-05-14 11:58:56 <super_duck> ncopa, I believe this is similar to the lxc issue of not working on rootfs due to their retarded use of pivot_root to avoid cleaning up their chroot properly
2013-05-14 11:59:25 <super_duck> ncopa, similar in that someone tried doing something smart, but it was overly smart and required a lot of extra work to let the smarts function properly
2013-05-14 12:03:14 <ncopa> this check here: http://git.alpinelinux.org/cgit/apk-tools/tree/src/add.c#n44
2013-05-14 12:03:53 <ncopa> it checks if the db is on tmpfs
2013-05-14 12:04:06 <ncopa> and if not it gives you the error
2013-05-14 12:04:11 <ncopa> hm
2013-05-14 12:04:14 <ncopa> looking at the code
2013-05-14 12:04:36 <ncopa> what happens if you do: mkdir /var/cache/apk; ln -s /var/cache/apk /etc/apk/cache
2013-05-14 12:04:40 <ncopa> does it work then?
2013-05-14 12:04:54 <ncopa> it checks if you have set up an apk cache
2013-05-14 12:05:02 <ncopa> but it does not check if the cache also is on tmpfs
2013-05-14 12:21:15 <ncopa> fabled: super_duck runs abuild on tmpfs. he getc bitten by the tmpfs protection
2013-05-14 12:21:30 <ncopa> only way to disable it is using --force
2013-05-14 12:21:38 <super_duck> what is fabled?
2013-05-14 12:21:54 <ncopa> but there might be other reasons to not use --force
2013-05-14 12:22:08 <ncopa> so i wonder if we should have a "config" option, like: touch /etc/apk/allow_tmpfs
2013-05-14 12:22:25 <ncopa> or: touch /etc/apk/no_tmpfs_check
2013-05-14 12:22:42 <ncopa> fabled is the author/maintainer of apk-tools
2013-05-14 12:26:18 <super_duck> can someone help me with understanding apk.  I have an updated libotr in aports/main/libotr, and am trying to build aports/testing/pidgin-otr which depends on libotr.  When building pidgin-otr it cannot find the library for libotr (path not found).  Does abuild not build the dependencies and put them in a locatable place for other packages?
2013-05-14 12:26:57 <ncopa> correct
2013-05-14 12:27:10 <ncopa> it only finds packages in same repos (eg 'testing')
2013-05-14 12:27:10 <Barthalion> you may need libotr-dev
2013-05-14 12:27:36 <ncopa> does it fail when it tries to install the deps
2013-05-14 12:27:43 <ncopa> or when it compiles it?
2013-05-14 12:28:18 <super_duck> >>> pidgin-otr*: Tracing dependencies...
2013-05-14 12:28:18 <super_duck> >>> pidgin-otr*:   added libgcc (due to libpthread)
2013-05-14 12:28:18 <super_duck> >>> ERROR: pidgin-otr*: libotr.so.5: path not found
2013-05-14 12:28:42 <ncopa> hm
2013-05-14 12:28:47 <ncopa> so it actually builds it
2013-05-14 12:28:53 <super_duck> yay
2013-05-14 12:29:02 <super_duck> the issue was not having depends="libotr-dev"
2013-05-14 12:29:07 <super_duck> I had it depends="libotr"
2013-05-14 12:29:16 <Barthalion> Weird
2013-05-14 12:29:21 <ncopa> you should not need to add it in depends
2013-05-14 12:29:33 <ncopa> it should be enough to add libotr-dev in makedepends
2013-05-14 12:29:34 <super_duck> makedepends sorry
2013-05-14 12:29:38 <ncopa> and leave depends alone
2013-05-14 12:29:39 <Barthalion> :)
2013-05-14 12:29:41 <Barthalion> As I said
2013-05-14 12:29:59 <super_duck> depends="libotr pidgin"
2013-05-14 12:29:59 <super_duck> makedepends="libotr-dev gtk+-dev intltool"
2013-05-14 12:30:15 <ncopa> the error above happens when abuild tries to find out what package provides libotr.so.5
2013-05-14 12:30:25 <super_duck> Barthalion, The ID10T error was identified when transcribing from APKBUILD to xchat....
2013-05-14 12:30:25 <ncopa> remove libotr from depends
2013-05-14 12:30:47 <super_duck> ncopa, why? should I remove pidgin too?
2013-05-14 12:30:58 <ncopa> abuild depends a bit
2013-05-14 12:31:20 <ncopa> if your package links anything directly to any lib in pidgin, then it should autodetect it
2013-05-14 12:31:42 <ncopa> if you are building a plugin or someething for pidgin, then it might make sense to keep it
2013-05-14 12:31:43 <super_duck> no, it just installs to the plugins directory that pidgin should provide
2013-05-14 12:32:13 <super_duck> how should I submit these APKBUILDs?
2013-05-14 12:32:22 <ncopa> git send email
2013-05-14 12:32:26 <ncopa> are you familiar with git?
2013-05-14 12:32:34 <super_duck> so-so
2013-05-14 12:32:44 <super_duck> never used it except as a local svn replacement
2013-05-14 12:32:59 <ncopa> http://wiki.alpinelinux.org/wiki/Creating_patches
2013-05-14 12:33:01 <ncopa> aw ok
2013-05-14 12:33:19 <ncopa> beeing used tiwh svn is not a good thing when switching to git :)
2013-05-14 12:33:27 <ncopa> its a completely different animal
2013-05-14 12:33:52 <super_duck> oh, I know how to use git, but only as a local i.e. non-distributed system
2013-05-14 12:34:22 <ncopa> do you have many patches/files in addition to the APKBUILD?
2013-05-14 12:34:29 <ncopa> or is it only the single APKBUILD
2013-05-14 12:34:51 <ncopa> if you want a quick review you can just sprunge it
2013-05-14 12:35:01 <ncopa> apk add sprunge; sprunge < APKBUILD
2013-05-14 12:35:04 <ncopa> and paste the url in here
2013-05-14 12:37:08 <super_duck> git: 'send-email' is not a git command. See 'git --help'.
2013-05-14 12:37:14 <ncopa> apk add git-perl
2013-05-14 12:37:24 <ncopa> we splitted out the perl deps to a separate pkg
2013-05-14 12:37:31 <super_duck> yes, perl should die
2013-05-14 12:37:54 <ncopa> i wish somebody rewrote git send-email in lua or C
2013-05-14 12:38:21 <super_duck> http://sprunge.us/BcQI
2013-05-14 12:38:35 <super_duck> http://sprunge.us/HaXc
2013-05-14 12:38:59 <Barthalion> Wait
2013-05-14 12:39:18 <Barthalion> libotr4 breaks otr support in bitlbee and mcabber
2013-05-14 12:39:18 <super_duck> ?
2013-05-14 12:39:23 <Barthalion> That's why I didn't upgrade it
2013-05-14 12:39:43 <ncopa> hm
2013-05-14 12:39:57 <super_duck> libotr4 fixes the issue of multiple disparate Jabber clients connected simultaneously using OTR
2013-05-14 12:40:16 <super_duck> It used to be a race condition between all the Jabber clients as to who got the encrypted conversation
2013-05-14 12:40:16 <Barthalion> We should upgrade libotr to 4 and add new package - libotr3
2013-05-14 12:40:33 <Barthalion> Didn't have time to do it, as I don't use Alpine on my desktop
2013-05-14 12:40:40 <ncopa> understand
2013-05-14 12:40:41 <super_duck> why not?
2013-05-14 12:40:56 <ncopa> hm
2013-05-14 12:40:57 <super_duck> I only have a little trouble getting Alpine + Google Chrome playing nice.
2013-05-14 12:41:22 <ncopa> other alternative is to name the new pk fo libotr4
2013-05-14 12:41:33 <ncopa> and keep the existing as is
2013-05-14 12:41:50 <Barthalion> I prefer keeping project name for newest upstream release
2013-05-14 12:41:57 <ncopa> yes
2013-05-14 12:42:10 <super_duck> And by a little trouble, I have to build a linux container of an Ubuntu flavor that installs Google Chrome, bind mount lots of stuff so that udev and pulseaudio play nice between host and chroot, and cry that libc.so.6 isn't on uClibc.
2013-05-14 12:42:12 <Barthalion> And with branch/version suffix for obsolete, but required for other packages
2013-05-14 12:42:27 <super_duck> I like Barthalion's thinking
2013-05-14 12:42:38 <ncopa> yeah
2013-05-14 12:42:40 <Barthalion> super_duck: You can try to compile Chromium, but it may be pain in ass
2013-05-14 12:42:54 <ncopa> chromium... we had a testing pkg for it
2013-05-14 12:42:58 <ncopa> i think it even built once
2013-05-14 12:43:01 <ncopa> but...
2013-05-14 12:43:04 <ncopa> nightmare...
2013-05-14 12:43:04 <super_duck> Barthalion, does Chromium support Google branding and integration with a buildtime flag?
2013-05-14 12:43:16 <Barthalion> Don't know about the latter
2013-05-14 12:43:24 <Barthalion> But Google branding - no
2013-05-14 12:43:30 <super_duck> I really want Google Chrome, not chromium.
2013-05-14 12:44:36 <ncopa> super_duck: do you think you can paste me the diff?
2013-05-14 12:44:42 <ncopa> git diff | sprunge
2013-05-14 12:44:43 <super_duck> ncopa, diff of what?
2013-05-14 12:44:49 <ncopa> libotr
2013-05-14 12:44:54 <ncopa> git diff . | sprunge
2013-05-14 12:45:17 <ncopa> or if you have committed it in git, git diff origin/master
2013-05-14 12:45:55 <super_duck> http://sprunge.us/AObJ
2013-05-14 12:46:22 <ncopa> ok
2013-05-14 12:46:23 <ncopa> thanks
2013-05-14 12:46:35 <ncopa> it was only a version number shnage
2013-05-14 12:46:37 <ncopa> change
2013-05-14 12:46:48 <super_duck> so I didn't update the Contributor/Maintainer in pidgin-otr/APKBUILD
2013-05-14 12:46:54 <super_duck> and md5sums
2013-05-14 12:48:33 <ncopa> do you think you could add contributor?
2013-05-14 12:48:42 <ncopa> so you get credit for your work
2013-05-14 12:48:55 <ncopa> (and we know who we will hang when things break...)
2013-05-14 12:49:03 <super_duck> +# Contributor: User <user@void>
2013-05-14 12:49:20 <ncopa> :)
2013-05-14 12:49:36 <ncopa> do you mind if i add contributor: super_duck
2013-05-14 12:49:45 <ncopa> np
2013-05-14 12:49:48 <ncopa> we can leave it
2013-05-14 12:49:53 <super_duck> Kinda, someone else might choose to steal this super special nic.
2013-05-14 12:50:02 <ncopa> ok
2013-05-14 12:50:04 <ncopa> we leave it
2013-05-14 12:50:04 <ncopa> np
2013-05-14 12:50:16 <super_duck> I'm indifferent, I'll pipe up that I contributed it, and I'm pretty sure nobody will care to claim ownership other than me.
2013-05-14 12:50:37 <super_duck> If I contribute x2go
2013-05-14 12:50:40 <super_duck> I'll claim credit
2013-05-14 12:50:44 <super_duck> that will be a beast
2013-05-14 12:57:50 <ncopa> should we remove Kiyoshi from maintener/contributor?
2013-05-14 12:58:05 <ncopa> might be he dont want maintain it
2013-05-14 12:59:18 <ncopa> pidgin-dev needs to be in makedepends
2013-05-14 12:59:23 <Barthalion> Hm, btw, what kind of magic I need to see myself as verified maintainer in package search on al.org?
2013-05-14 12:59:46 <ncopa> i think that is not supported :-/
2013-05-14 12:59:54 <Barthalion> ah
2013-05-14 13:00:09 <Barthalion> Just asking, I don't know how it works :p
2013-05-14 13:00:19 <ncopa> i think clandmeter did the packages browser
2013-05-14 13:04:57 <clandmeter> Barthalion: sign up with your email address
2013-05-14 13:05:15 <clandmeter> and when you sign up let me know
2013-05-14 13:05:18 <clandmeter> so i can activate you.
2013-05-14 13:05:27 <Barthalion> clandmeter: Great, will do tommorow
2013-05-14 13:06:00 <clandmeter> Barthalion: http://alpinelinux.org/user/register
2013-05-14 13:07:01 <super_duck> ncopa, why pidgin-dev?  It just needs the /usr/lib/pidgin/plugins directory, or some nonsense
2013-05-14 13:07:14 <ncopa> it needs the headers too
2013-05-14 13:07:23 <ncopa> and some .pc files
2013-05-14 13:07:30 <super_duck> oh, ya, I remember that
2013-05-14 13:07:42 <super_duck> ncopa, thank you for the lessons
2013-05-14 13:07:45 <ncopa> http://tinyurl.com/cvcge28
2013-05-14 13:07:50 <ncopa> thats how it ended up
2013-05-14 13:08:15 <ncopa> thanks for your contribution
2013-05-14 13:09:49 <super_duck> what's the process for it to end up in edge?
2013-05-14 13:10:42 <ncopa> it is in edge/testing now
2013-05-14 13:10:56 <ncopa> now, what i need you to do is install it from http repo, and test it
2013-05-14 13:11:02 <super_duck> I did an apk update, and apk search pidgin doesn't show it
2013-05-14 13:11:21 <ncopa> do you have edge/testing in your /etc/apk/repositories ?
2013-05-14 13:11:44 <ncopa> what middor do you have?
2013-05-14 13:11:47 <super_duck> I do now :)
2013-05-14 13:11:48 <ncopa> mirror
2013-05-14 13:12:03 <ncopa> it might take an hour or so for the mirror to rsync
2013-05-14 13:12:25 <ncopa> you can use http://nl.alpinelinux.org as mirror if you want it faster
2013-05-14 13:12:38 <super_duck> midori-0.5.0-r0
2013-05-14 13:12:54 <ncopa> i meant mirror, i just cannot type
2013-05-14 13:12:56 <super_duck> fetch http://dl-3.alpinelinux.org/alpine/edge/testing/x86_64/APKINDEX.tar.gz
2013-05-14 13:13:22 <ncopa> change dl-3.alpinelinux.org in /etc/apk/repositories to nl.alpinelinux.org
2013-05-14 13:13:28 <super_duck> yay!
2013-05-14 13:13:34 <super_duck> I see it now
2013-05-14 13:14:06 <ncopa> test if it work
2013-05-14 13:14:13 <ncopa> if it does, I'll move it to main
2013-05-14 13:14:24 <super_duck> oh
2013-05-14 13:14:31 <super_duck> um, probably not the best idea, I manually installed it
2013-05-14 13:14:43 <super_duck> I'd have to reboot and I'd lose a lot of work...
2013-05-14 13:14:49 <super_duck> probably with a stateless system
2013-05-14 13:14:53 <ncopa> lol
2013-05-14 13:15:53 <ncopa> you can 'lbu include /home/user/dir/with/work' and 'lbu package' to generate a backup ;)
2013-05-14 13:15:59 <ncopa> it will take all the diffs in /etc
2013-05-14 13:16:17 <ncopa> so you can get it back to same state afterwards
2013-05-14 13:16:18 <super_duck> vostro:~# apk add pidgin-otr
2013-05-14 13:16:18 <super_duck> ERROR: 1 unsatisfiable dependencies:
2013-05-14 13:16:18 <super_duck>   pidgin-otr-4.0.0-r0: so:libotr.so.5
2013-05-14 13:16:50 <ncopa> do you have edge/main in repositories too?
2013-05-14 13:16:57 <super_duck> yes
2013-05-14 13:17:03 <super_duck> homefs on /home type tmpfs (rw,relatime)
2013-05-14 13:17:05 <ncopa> hm
2013-05-14 13:17:12 <ncopa> try apk upgrade -U
2013-05-14 13:17:19 <super_duck> not sure I want to trust lbu on tmpfs
2013-05-14 13:17:46 <super_duck> not sure I want to trust lbu on tmpfs
2013-05-14 13:17:48 <super_duck> Bus error
2013-05-14 13:18:01 <super_duck> apk upgrade -U ends with a Bus error on 7/40 Upgrading dbus-libs
2013-05-14 13:18:18 <super_duck> ncopa, I'll test this later
2013-05-14 13:18:20 <super_duck> I've gotta scram
2013-05-14 13:18:41 <ncopa> sounds like out of "disk" space
2013-05-14 13:19:13 <super_duck> yes yes it is
2013-05-14 13:19:19 <super_duck> my build box has 32GB of ram
2013-05-14 13:19:23 <super_duck> my laptop has 2GB of ram
2013-05-14 13:19:44 <super_duck> laptop runs full GUI, build box runs bash
2013-05-14 13:19:54 <super_duck> so ya, I'm outta ram
2013-05-14 13:20:07 <ncopa> abuild clean
2013-05-14 13:20:19 <super_duck> ncopa, not the issue
2013-05-14 13:20:26 <ncopa> ok
2013-05-14 13:20:28 <super_duck> ncopa, the laptop is really out of tmpfs space
2013-05-14 13:22:54 <ncopa> ok
2013-05-14 13:22:57 <ncopa> i tested it ehre
2013-05-14 13:22:59 <ncopa> it works
2013-05-14 13:30:36 <ncopa> i wonder if i dare upgrade alpine-vserver host
2013-05-14 13:30:39 <ncopa> and restart it
2013-05-14 13:46:43 <vkrishn> ncopa, will it be any issue if kms.cfg has `options cirrus modeset=0` ?
2013-05-14 13:47:53 <vkrishn> kms.conf
2013-05-14 13:49:02 <vkrishn> I thinks this was in one of the list-thread during 2.5 release
2013-05-14 13:54:56 <vkrishn> http://bugs.alpinelinux.org/issues/1848 can be closed. I get any similar issues with other mod, I would create new b.a.o
2013-05-14 13:55:09 <vkrishn> if I get*
2013-05-14 14:00:45 <ncopa> i think we want rename the .ini
2013-05-14 14:00:54 <ncopa> will keep it open meanwhile
2013-05-14 14:01:04 <ncopa> no sure about cirrus option
2013-05-14 14:07:54 <vkrishn> yes xml_wddx.ini
2013-05-14 14:08:20 <ncopa> here comes rc4
2013-05-14 14:09:20 <ncopa> algitbot: build master
2013-05-14 14:12:23 <vkrishn> http://lists.alpinelinux.org/alpine-devel/2311.html
2013-05-14 14:15:15 <ncopa> do you think we want/need the nomodeset for cirrus?
2013-05-14 14:23:38 <vkrishn> if fonts can be a little niicer in normal kvm boot I guess not
2013-05-14 14:23:40 <vkrishn> also was thinking there were some nifty alias set by default
2013-05-14 14:23:41 <vkrishn> like alias ll=ls -l
2013-05-14 22:01:58 <kaniini> super_duck: we talked about the problem you encountered
2013-05-14 22:02:13 <super_duck> and I'm a nut case?
2013-05-14 22:02:40 <kaniini> well
2013-05-14 22:02:43 <kaniini> of course
2013-05-14 22:02:45 <kaniini> but
2013-05-14 22:03:04 <kaniini> 05:58:27      @ncopa | if users want shoot themselves in the foot we maybe should let them
2013-05-14 22:03:07 <kaniini> viewpoint is that
2013-05-14 22:03:18 <super_duck> ok
2013-05-14 22:03:41 <super_duck> out of curiosity, why didn't Alpine steal the Arch PKGBUILD scripts to bootstrap their package repository?
2013-05-14 22:03:44 <kaniini> so, i am going to push a patch to apk-tools momentarily.
2013-05-14 22:03:54 <kaniini> well...
2013-05-14 22:04:09 <kaniini> because arch's pkgbuild scripts are usually poorly done
2013-05-14 22:04:12 <kaniini> to be honest
2013-05-14 22:04:26 <kaniini> we did not wish to debug their PKGBUILDs
2013-05-14 22:05:18 <super_duck> I think that statement applies to anything that resembles code
2013-05-14 22:05:30 <super_duck> things are usually poorly done
2013-05-14 22:05:47 <super_duck> so, before you publish the patch
2013-05-14 22:06:01 <super_duck> how exactly am I shooting myself in the foot building on tmpfs?
2013-05-14 22:06:24 <super_duck> I know the environment is tmpfs, is it normally considered static storage when abuild is used?
2013-05-14 22:07:17 <Mp5shooter> can't wait for libreswan's people to add alpine in their little list
2013-05-14 22:07:23 <Mp5shooter> so it'll actually work on alpine
2013-05-14 22:07:23 <Mp5shooter> lo
2013-05-14 22:09:55 <super_duck> kaniini, so there's like 5 or 6 packages that Arch has that Alpine does not.  If I steal their PKGBUILD's and fix them into APKBUILD's and the build works, will that be accepted by the devs, or should I do mine from scratch?
2013-05-14 22:14:02 <kaniini> super_duck: it is fine
2013-05-14 22:14:35 <super_duck> how stickler are folks for getting the licenses 100% correct?
2013-05-14 22:14:49 <kaniini> it is important to us
2013-05-14 22:15:01 <super_duck> I'm looking at one PKGBUILD that has 2 licenses, GPL and PerlArtistic
2013-05-14 22:15:16 <kaniini> yes, that's fine
2013-05-14 22:16:49 <super_duck> okay, well I got the license wrong with pidgin-otr
2013-05-14 22:16:55 <super_duck> so I'll need to make sure that gets corrected
2013-05-14 22:17:10 <super_duck> GPLv2 not LGPL
2013-05-15 03:41:40 <super_duck> I'm trying to build a package with SUDO_APK="abuild-apk --force" abuild -r
2013-05-15 03:41:40 <super_duck>  and I get no output
2013-05-15 03:42:04 <super_duck> http://sprunge.us/QQLJ
2013-05-15 03:43:07 <super_duck> nvm, shot myself in foot with bad arch
2013-05-15 03:54:31 <super_duck> so I do abuild -r, where does the package actual reside and how to I configure apk to know about the new package?
2013-05-15 06:43:59 <ncopa> super_duck: depends if you have set REPODEST in /etc/abuild.conf or not
2013-05-15 06:44:08 <ncopa> if its unset, the package will end up in current dir
2013-05-15 06:44:41 <ncopa> and in ~/.cache/abuild/* you'll have a temp repo with symlinks
2013-05-15 06:45:21 <ncopa> if you set REPODEST=$HOME/packages  in /etc/abuild.conf then you packages will end up in ~/packages/<repo>/<arch>/
2013-05-15 06:45:43 <ncopa> and you can add /home/<user>/<repo> to your /etc/apk/repositories
2013-05-15 06:45:51 <ncopa> and apk will fidn all the locally built packages
2013-05-15 06:45:57 <ncopa> thats how i do here
2013-05-15 06:46:15 <ncopa> *and you can add /home/<user>/packages/<repo> to your /etc/apk/repositories
2013-05-15 08:03:13 <build-2-5> files from v2.5.4-75-gf8fd1f1 uploaded
2013-05-15 08:03:39 <build64-2-5> files from v2.5.4-75-gf8fd1f1 uploaded
2013-05-15 08:40:41 <ncopa> we have another kernel bug
2013-05-15 08:41:05 <kaniini> semtex.c works but with modifications
2013-05-15 08:41:13 <ncopa> somehow my lxc named dev32-edge ended up in a locked state
2013-05-15 08:41:15 <ncopa> it just hanged
2013-05-15 08:41:20 <ncopa> when trying to shut it down
2013-05-15 08:41:26 <kaniini> oh, wonderful
2013-05-15 08:41:33 <ncopa> the process does not die
2013-05-15 08:41:38 <ncopa> and cannot be killed with -9
2013-05-15 08:41:46 <ncopa> so i did some digging
2013-05-15 08:41:51 <ncopa> State:	D (disk sleep)
2013-05-15 08:42:01 <barthalion> kaniini: Hm, on 2.5 or edge?
2013-05-15 08:42:02 <ncopa> which means it waits for IO
2013-05-15 08:42:21 <ncopa> log is showing this
2013-05-15 08:42:31 <ncopa> May 15 10:39:20 ncopa-desktop kern.emerg kernel: [75215.168462] unregister_netdevice: waiting for vethzMtePm to become free. Usage count = 1
2013-05-15 08:42:31 <ncopa> May 15 10:39:30 ncopa-desktop kern.emerg kernel: [75225.326378] unregister_netdevice: waiting for vethzMtePm to become free. Usage count = 1
2013-05-15 08:42:32 <kaniini> 2.5
2013-05-15 08:42:57 <kaniini> it reliably causes an oops
2013-05-15 08:43:03 <kaniini> i am... concerned about it
2013-05-15 08:43:35 <ncopa> so kernel is waiting to release the net interface vethzMtePm
2013-05-15 08:43:46 <ncopa> which has one 'user'
2013-05-15 08:43:53 <ncopa> but the net interface is gone
2013-05-15 08:44:22 <ncopa> it does not appear in /sys/class/net/
2013-05-15 08:44:34 <ncopa> so i wonder what is holding it
2013-05-15 08:44:51 <ncopa> i suspect that dhcpcd used to hold it
2013-05-15 08:45:01 <ncopa> or something
2013-05-15 08:45:15 <ncopa> but those lockup are not supposed to be able to be triggered from userspace
2013-05-15 08:45:18 <ncopa> so i think its a kernel bug
2013-05-15 10:40:37 <build64-2-5> files from v2.5.4-83-gf216910 uploaded
2013-05-15 12:15:29 <super_duck> ncopa, so I run  SUDO_APK="abuild-apk --force" abuild -r
2013-05-15 12:15:29 <super_duck> and I get no packages
2013-05-15 12:16:02 <super_duck> /home/user/packages/testing/x86_64/APKINDEX.tar.gz.unsigned has a description that's filled in and an empty APKINDEX file
2013-05-15 12:17:18 <ncopa> have you built any packages in testing?
2013-05-15 12:17:37 <ncopa> is your REPODEST set to /home/user/packages in /etc/abuild.conf?
2013-05-15 12:20:45 <super_duck> REPODEST=/home/user/packages
2013-05-15 12:20:58 <super_duck> and I just build testing/empty and testing/zim (my package)
2013-05-15 12:20:58 <ncopa> do you have any packages in there?
2013-05-15 12:21:04 <super_duck> there's nothing in /home/user/packages
2013-05-15 12:21:19 <ncopa> what happens if you try rebuild zim?
2013-05-15 12:21:23 <super_duck> just -rw-r--r-- 1 user user 157 May 15 12:21 /home/user/packages/testing/x86_64/APKINDEX.tar.gz.unsigned
2013-05-15 12:21:43 <ncopa> hm
2013-05-15 12:21:49 <super_duck> says something about packager key
2013-05-15 12:21:51 <ncopa> it should have generated a signed
2013-05-15 12:21:52 <super_duck> do I need a key?
2013-05-15 12:21:56 <ncopa> ah yes
2013-05-15 12:22:01 <super_duck> wtf?@##$#@
2013-05-15 12:22:07 <super_duck> I have to have a key?
2013-05-15 12:22:09 <ncopa> you need a keypair
2013-05-15 12:22:21 <super_duck> I'd like to file this under shooting oneself in the foot category
2013-05-15 12:22:29 <barthalion> There is a command to generate and add it automatically
2013-05-15 12:22:29 <ncopa> yes, abuild will sign the apk
2013-05-15 12:22:34 <super_duck> Can I update the wiki with wiki/Super_Duck_Foot_Shooting?
2013-05-15 12:23:04 <super_duck> I know I can generate one, but it seems like its only necessary if you want a crypto signed trusted infrastructure
2013-05-15 12:23:46 <ncopa> http://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Setup_your_system_and_account
2013-05-15 12:23:50 <ncopa> last setp there
2013-05-15 12:23:53 <ncopa> step*
2013-05-15 12:23:58 <super_duck> yay, that made packages
2013-05-15 12:24:03 <ncopa> abuild-keygen -a -i
2013-05-15 12:24:09 <barthalion> It's not big deal to just add key
2013-05-15 12:24:17 <barthalion> It will make abuild and apk happy
2013-05-15 12:24:31 <super_duck> I *strongly* recommend that not having a key produce abuild errors, rather than some notice/warning text
2013-05-15 12:25:12 <ncopa> well apk will not install anything thats not signed
2013-05-15 12:25:21 <ncopa> so if its not signed the apk is broke
2013-05-15 12:25:28 <ncopa> and abuild will not build a broken apk
2013-05-15 12:26:09 <ncopa> but sure... we could support it, it was just not worth the extra time to implement it
2013-05-15 12:26:31 <ncopa> instead we made it really simple to make a keypair
2013-05-15 12:27:18 <super_duck> well if you have to sign, and you have no key, abuild should throw an error, not the nice warning/notice text indicating you should generate a keypair
2013-05-15 12:27:28 <super_duck> I'd opt for the latter personally over supporting no package signing
2013-05-15 12:27:38 <ncopa> i think you get error if you dont have the --force
2013-05-15 12:27:40 <super_duck> ncopa, thanks for all the assistance
2013-05-15 12:28:01 <super_duck> ncopa, interesting
2013-05-15 12:28:16 <super_duck> ncopa, I'm using the force to bypass the tmpfs restriction
2013-05-15 12:28:19 <ncopa> with --force you are basically telling apk, well, i know you think this is wrong, but please just go ahead and do whatever i have asked you to
2013-05-15 12:28:25 <ncopa> yes
2013-05-15 12:28:33 <ncopa> i think that might be worth fixing though
2013-05-15 12:28:53 <super_duck> ncopa, alright, I'll just add this to the list of super_duck foot shooting aka DUCK HUNT!
2013-05-15 12:29:06 <ncopa> :)
2013-05-15 12:32:40 <build-2-4> files from v2.4.10-48-gd792e89 uploaded
2013-05-15 12:35:12 <build64-2-3> files from v2.3.6-203-ged634db uploaded
2013-05-15 13:10:19 <kaniini> oh geeze
2013-05-15 13:11:41 <kaniini> i had an uncle which did nothing but play duck hunt all day (well, he had some sort of syndrome that stunted his neurodevelopment)
2013-05-15 13:13:12 <barthalion> http://www.gotoandplay.it/_games/_nesDuckHunt/nesDuckHunt.jpg
2013-05-15 13:13:43 <kaniini> yes
2013-05-15 13:13:47 <ncopa> lol
2013-05-15 13:13:49 <kaniini> that game, with the light gun
2013-05-15 13:14:04 <jbilyk__> that brings back memories
2013-05-15 13:14:07 <barthalion> yeah
2013-05-15 13:14:48 <barthalion> NES was pretty cool, my parents said that I lost a lot of time playing when I had 4 years
2013-05-15 13:14:59 <barthalion> I still can't believe I was able to finish Contra
2013-05-15 13:15:34 <kaniini> when i was growing up, it was SNES
2013-05-15 13:16:16 <ncopa> ncopa-desktop:~$ apk search snes
2013-05-15 13:16:16 <ncopa> bsnes-0.91-r1
2013-05-15 13:16:43 <barthalion> I'll be brutal -- I'm from central Europe
2013-05-15 13:16:49 <barthalion> There was no such thing as SNES
2013-05-15 13:17:00 <barthalion> at least easily accessible in Poland
2013-05-15 13:17:34 <ncopa> i was not fortunate enought to have NES
2013-05-15 13:17:41 <ncopa> or SNES
2013-05-15 13:18:59 <ncopa> but i remember playing price of persia on my 8088
2013-05-15 13:19:11 <ncopa> or was it 8086
2013-05-15 13:19:33 <barthalion> I wasn't fortunate enough to have old PC, my first one was AMD Duron :p
2013-05-15 13:21:11 <ncopa> but i think i played more with turbo pascal and turbo asm back then :)
2013-05-15 13:21:30 <kaniini> yeah, i played prince of persia on my dad's 8088
2013-05-15 13:23:06 <ncopa> kaniini: i pushed fixed kernels for v2.5, v2.4, v2.3 and are about to push v2.2
2013-05-15 13:23:45 <ncopa> well, its only fix for CVE-2013-2094
2013-05-15 13:24:06 <ncopa> i saw some other nasty things when working on the 3.4.y kernel
2013-05-15 13:25:07 <ncopa> something about be able to bypass net acl on suid root binaries
2013-05-15 13:27:18 <ncopa> https://github.com/ncopa/linux-3.4-grsec/commit/ca4bf7c6970aee586c4b6b642e011e3847ac5f93
2013-05-15 13:27:55 <kaniini> but, ncopa, grsecurity is not vulnerable!  because people won't run paxctl on the exploit binary ;)
2013-05-15 13:28:52 <ncopa> like java and firefox and miroi and all qt quick stuff... (or any other binary who uses JIT)
2013-05-15 13:29:18 <ncopa> i havent digged into how to expoit it
2013-05-15 13:29:42 <kaniini> i got an oops with it
2013-05-15 13:29:58 <kaniini> not a "PAX killed this because it looks bad"
2013-05-15 13:30:03 <ncopa> thats bad
2013-05-15 13:30:41 <ncopa> did test any of the kernels i pushed today?
2013-05-15 13:30:54 <kaniini> about to
2013-05-15 13:30:58 <kaniini> i just woke up a bit ago
2013-05-15 13:31:04 <ncopa> good morning!
2013-05-15 13:31:08 <ncopa> reminds me
2013-05-15 13:31:11 <kaniini> took a nap
2013-05-15 13:31:16 <ncopa> im gonna look for coffe
2013-05-15 13:31:31 <ncopa> amazing what sleep can do with you, isnt it :)
2013-05-15 13:56:12 <kaniini> i'm not seeing an oops anymore
2013-05-15 13:56:19 <kaniini> on the new 2.5 kernel
2013-05-15 13:56:46 <kaniini> some gentoo person of course had this amusing wisdom: why would we want to ship with support for event tracing
2013-05-15 13:56:58 <kaniini> gee, no idea why we'd ever want to do that
2013-05-15 13:56:58 <ncopa> nice!
2013-05-15 13:57:08 <ncopa> hm
2013-05-15 13:57:10 <ncopa> true
2013-05-15 13:57:20 <ncopa> we disable that in our config?
2013-05-15 13:57:27 <ncopa> at least for v2.6
2013-05-15 13:57:31 <kaniini> well, no
2013-05-15 13:57:34 <kaniini> we should keep it
2013-05-15 13:57:50 <kaniini> i think the more preferable option is to have a kernel image variant that has a smaller attack surface
2013-05-15 13:57:57 <kaniini> linux-paranoid or something
2013-05-15 13:58:01 <ncopa> heh
2013-05-15 13:58:06 <ncopa> thats a home built kernel
2013-05-15 13:58:26 <kaniini> CONFIG_PERF_EVENTS is needed for things like iotop i believe
2013-05-15 13:58:38 <kaniini> so, we definitely want to keep that
2013-05-15 13:58:45 <fabled> (or honor paxctl only on signed binaries)
2013-05-15 13:58:58 <ncopa> ok
2013-05-15 14:00:00 <ncopa> kaniini: how did it wo with ff-21?
2013-05-15 14:00:58 <kaniini> i need to take another look at it
2013-05-15 14:06:48 <kaniini> const Cu = Components.utils;
2013-05-15 14:06:57 <kaniini> this is what bombs in make install
2013-05-15 14:29:14 <kaniini> petrie:~/aports/main/xulrunner$ sha512sum /var/cache/distfiles/firefox-21.0.source.tar.bz2
2013-05-15 14:29:16 <kaniini> 0607e587a03c7402b186bc6e254f79a6cc29dfdd1836d3857233a0175e0d7e9f03ab69261e98b503ae6b5b47deac8ebbb982649729e3ed4d9e646f13073e0822  /var/cache/distfiles/firefox-21.0.source.tar.bz2
2013-05-15 14:29:18 <kaniini> petrie:~/aports/main/xulrunner$ sha512sum /var/cache/distfiles/xulrunner-21.0.source.tar.bz2
2013-05-15 14:29:20 <kaniini> 5851fb3215ecb05ee4a9161326890b3157c6ce45d3e635c7c11fd00529f6e3fbaba7d4ceccea749267999b9751861c7ad14f598a39a051cc3fc8046724c55a5a  /var/cache/distfiles/xulrunner-21.0.source.tar.bz2
2013-05-15 14:29:22 <kaniini> hmm...
2013-05-15 14:31:56 <ncopa> kaniini, so ff and xulrunner sources does not match?
2013-05-15 14:32:11 <kaniini> seems not
2013-05-15 14:32:26 <kaniini> my disapproval has been communicated to certain mozilla corporation employees
2013-05-15 14:33:04 <kaniini> ncopa: this is annoying because, they do not always put up a new xulrunner
2013-05-15 14:33:18 <kaniini> so, we just normally pull it out of firefox sources
2013-05-15 14:33:20 <kaniini> haha
2013-05-15 14:33:41 <ncopa> heh yes
2013-05-15 14:51:11 <kaniini> running another build, please wait
2013-05-15 15:07:46 <kaniini> also, something i noticed...
2013-05-15 15:07:58 <kaniini> lists.alpinelinux.org URLs are broken
2013-05-15 15:08:10 <kaniini> every time a message is posted to it
2013-05-15 15:20:33 <kaniini> mozpack.errors.ErrorMessage: Error: Error while running startup cache precompilation
2013-05-15 15:20:35 <kaniini> same error
2013-05-15 15:22:12 <kaniini> http://paste2.org/Xc0OJftN
2013-05-16 06:39:22 <ncopa> mornin
2013-05-16 06:39:36 <kaniini> upgrading to that grsecurity patch seems kinda pointless :P
2013-05-16 06:39:40 <kaniini> it just renames a variable
2013-05-16 06:39:45 <ncopa> oh :)
2013-05-16 06:39:59 <ncopa> i just saw that there was a new
2013-05-16 06:40:11 <ncopa> kaniini: i have some ideas for buildlab
2013-05-16 06:40:20 <kaniini> i want to start checking the grsec patches by making interdiffs
2013-05-16 06:40:37 <ncopa> kaniini, yes thats a good idea
2013-05-16 06:40:44 <ncopa> i'd like to track them in git
2013-05-16 06:40:47 <ncopa> but its not easy
2013-05-16 06:40:54 <ncopa> would be awseome to be able to git bisec
2013-05-16 06:40:55 <ncopa> git log
2013-05-16 06:41:02 <ncopa> to have the history visible
2013-05-16 06:41:54 <ncopa> kaniini: i collect the patches
2013-05-16 06:42:00 <ncopa> both the pax and the grsecurity
2013-05-16 06:43:02 <ncopa> i think i have since grsecurity-2.9.1-3.7.1-201212171734.patch
2013-05-16 06:44:14 <ncopa> http://ld-linux.so/~minipli/tools/grsec-get-patch.sh
2013-05-16 06:45:00 <ncopa> cool domain :)
2013-05-16 06:46:06 <ncopa> i'd like to register http://libc.so :)
2013-05-16 06:55:42 <ncopa> kaniini, so the ideas for buildlab
2013-05-16 06:55:53 <ncopa> i wonder if we could set up tem lxc's
2013-05-16 06:56:00 <ncopa> dunno if that makes any sense
2013-05-16 06:56:08 <ncopa> but the main idea is:
2013-05-16 06:56:26 <ncopa> 1) install a tmp root (apk add build-base --root $tmp)
2013-05-16 06:56:29 <kaniini> lxc's sound awesome yeah
2013-05-16 06:57:01 <ncopa> 2) copy the current user (grep $USER /etc/passwd >> $tmp/etc/passwd)
2013-05-16 06:57:17 <kaniini> this sounds alright with me so far
2013-05-16 06:57:32 <ncopa> 3) bind mount $HOME (mount --bind $HOME $tmp/$HOME)
2013-05-16 06:57:46 <ncopa> now when you enter, the container
2013-05-16 06:57:52 <ncopa> you have it all there
2013-05-16 06:57:59 <ncopa> you abuild keys,
2013-05-16 06:58:02 <ncopa> your abuild cache
2013-05-16 06:58:15 <ncopa> your temp build dir will still be in your homedir
2013-05-16 06:58:57 <ncopa> you just get your /usr (installed packages) different
2013-05-16 06:59:17 <ncopa> second idea is to make it work with both archs
2013-05-16 06:59:43 <ncopa> so you can use same APKBUILD to build x86_64 apk and x86
2013-05-16 07:00:32 <ncopa> as it is now, i have to test kernels in different boxes, x86 and x86_64
2013-05-16 07:00:45 <ncopa> and copy the kernelconfig.x86/x86_64 around between them
2013-05-16 07:01:00 <ncopa> woudl be nice to use same workdir to test both
2013-05-16 07:01:25 <ncopa> if now REPODEST=$HOME/packages, and we bind mount $HOME
2013-05-16 07:01:50 <ncopa> then it'd just work
2013-05-16 07:02:16 <ncopa> i saw that the ubuntu lxc template supports bind mount of $HOME
2013-05-16 07:02:28 <ncopa> that where i got the idea
2013-05-16 07:31:53 <vkrishn> ncopa, I got some thought on git tracking of grsec patches, if you are still planning ?
2013-05-16 07:32:30 <ncopa> planning and planning... its a question of time to actually do it
2013-05-16 07:32:52 <ncopa> *dedicated time
2013-05-16 07:33:00 <vkrishn> should not be that difficult, if you already have git clone of linux kernel
2013-05-16 07:33:27 <ncopa> it is not so simple as it sounds like
2013-05-16 07:33:43 <vkrishn> creating git branch is cheap and disposable
2013-05-16 07:33:54 <ncopa> well, you could always create branches for each dot release
2013-05-16 07:33:57 <ncopa> and apply a patch
2013-05-16 07:34:19 <vkrishn> just create a branch grsectracking
2013-05-16 07:34:26 <ncopa> you would basically end up with a branch for each grsec/pax patch
2013-05-16 07:34:44 <vkrishn> not branch the kernel version that has patch available
2013-05-16 07:34:50 <vkrishn> now branch*
2013-05-16 07:35:06 <vkrishn> apply patch to that temp branch
2013-05-16 07:35:22 <vkrishn> merge to tracking branch..... repeat process
2013-05-16 07:35:23 <vkrishn> voila
2013-05-16 07:35:39 <vkrishn> yes, delete the temp branch when done
2013-05-16 07:35:42 <vkrishn> voila2
2013-05-16 07:35:58 <ncopa> and when you jump from 3.8 to 3.9?
2013-05-16 07:36:12 <ncopa> start over?
2013-05-16 07:36:25 <ncopa> the main problem is that it is manual work
2013-05-16 07:36:30 <ncopa> and once you start with it
2013-05-16 07:36:39 <vkrishn> yes to some extent
2013-05-16 07:36:42 <ncopa> you kinda have to continue do it til you die
2013-05-16 07:36:46 <vkrishn> possible to script it
2013-05-16 07:37:41 <vkrishn> since the process is repetitive
2013-05-16 07:37:46 <ncopa> but as i said, its a question of taking the time to actually do it
2013-05-16 07:37:56 <vkrishn> hmm
2013-05-16 07:38:45 <ncopa> also
2013-05-16 07:38:58 <ncopa> if you have v3.9.2 vanilla
2013-05-16 07:39:09 <ncopa> you apply the pax-3.9.2-test1
2013-05-16 07:39:13 <ncopa> then comes test2 out
2013-05-16 07:39:46 <ncopa> you branch a new tmp from 3.9.2-vanilla , and apply test2 to that?
2013-05-16 07:39:53 <ncopa> and then merge it in?
2013-05-16 07:40:29 <ncopa> then comes v3.9.3
2013-05-16 07:40:49 <ncopa> hm
2013-05-16 07:41:09 <ncopa> i think you'll end up with manual work regardless
2013-05-16 07:41:24 <ncopa> now, i know for sure that spender uses git
2013-05-16 07:41:29 <ncopa> i think pipacs does not
2013-05-16 07:41:57 <ncopa> would been great to have access to spenders git tree
2013-05-16 08:06:36 <vkrishn> if they do have a public accessible git than ok
2013-05-16 08:07:04 <vkrishn> I did not say it would be easy
2013-05-16 08:07:13 <vkrishn>  but the process seems repetitive
2013-05-16 08:08:15 <vkrishn> ncopa, thanks again for the irc log
2013-05-16 08:09:07 <ncopa> np :)
2013-05-16 08:09:11 <vkrishn> I had send last three msgs , they appeared in XChat box but not on irc..
2013-05-16 08:09:16 <vkrishn> log helped me
2013-05-16 08:09:18 <ncopa> i didnt do the irclogger_com though
2013-05-16 08:09:20 <vkrishn> to track
2013-05-16 08:09:49 <vkrishn> ;) they are useful
2013-05-16 08:15:04 <vkrishn> getlog.sh - http://dpaste.de/ETZVe/
2013-05-16 08:15:05 <vkrishn> alpine-irclog2html.sh - http://dpaste.de/ae0KC/
2013-05-16 08:15:17 <vkrishn> for anyone who has a similar problem
2013-05-16 10:04:01 <build64-2-5> files from v2.5.4-86-g2666936 uploaded
2013-05-16 10:05:42 <build64-2-5> files from v2.5.4-87-ge515946 uploaded
2013-05-16 10:06:13 <build-2-4> files from v2.4.10-52-g48428d9 uploaded
2013-05-16 10:10:56 <build-2-2> files from v2.2.3-228-g9d7bb41 uploaded
2013-05-16 10:12:40 <build-2-2> files from v2.2.3-229-g265b297 uploaded
2013-05-16 10:19:30 <build64-2-3> files from v2.3.6-205-g2bf234d uploaded
2013-05-16 10:21:20 <build64-2-3> files from v2.3.6-206-g6fc12c2 uploaded
2013-05-16 11:48:34 <ncopa> fabled: any suggestion what wed to with uclibc getchar?
2013-05-16 11:48:54 <ncopa> i wonder if we should just make it an inline in uclibc
2013-05-16 11:49:05 <ncopa> and recompile lvm2
2013-05-16 11:49:11 <ncopa> should not break anything
2013-05-16 11:55:45 <build-2-4> files from v2.4.10-53-gead36ed uploaded
2013-05-16 12:10:03 <fabled> ncopa, yes, i'd probably suggest that
2013-05-16 12:10:09 <fabled> make inline, and recompile broken apps
2013-05-16 12:28:44 <ncopa> royger: are you ok with the diff from http://bugs.alpinelinux.org/issues/1850
2013-05-16 12:30:20 <ncopa> royger: i'm thinking of this: http://sprunge.us/PJZL
2013-05-16 12:45:16 <build-2-2> files from v2.2.3-230-g38fb81b uploaded
2013-05-16 12:46:47 <build-2-4> files from v2.4.10-54-g3791168 uploaded
2013-05-16 12:46:58 <build-2-5> files from v2.5.4-88-ge8b58a0 uploaded
2013-05-16 12:53:52 <ncopa> that patch will make it sleep even if it is only 1 domain starting up
2013-05-16 14:56:55 <ncopa> i need help with testing xendomains init.d script
2013-05-16 14:57:16 <ncopa> would be great if somebody could test it til tm
2013-05-16 14:59:54 <ncopa> maybe should do rc6
2013-05-16 19:06:23 <iilluzion2> Systems in 2012 year!
2013-05-16 19:59:37 <barthalion> Ech, I can't register an account
2013-05-16 20:00:16 <barthalion> While creating website says that "Barthalion" is taken, when I want to request new password I get message that "Barthalion" is not recognized as username
2013-05-17 02:15:24 <super_duck> I'm getting an error in my postcheck function for a package.  How do I override that function?
2013-05-17 02:19:55 <super_duck> >>> ERROR: gtkhotkey*: Libtool archives (*.la) files found and $options has no 'libtool' flag:
2013-05-17 02:20:55 <jbilyk_> super_duck: is the package supposed to have libtool archives? most don't need them in the package so they could be deleted
2013-05-17 02:21:36 <jbilyk_> super_duck: http://lists.alpinelinux.org/alpine-devel/0679.html
2013-05-17 02:23:04 <super_duck> I have no idea, its a dependnecy for another package I'm trying to build
2013-05-17 02:23:34 <jbilyk_> super_duck: try removing them then
2013-05-17 02:41:19 <super_duck> jbilyk_, seems to have resolved my issue, thanks
2013-05-17 02:58:23 <jbilyk_> cool
2013-05-17 06:20:10 <ncopa> morning
2013-05-17 06:20:16 <ncopa> *.la files should be removed
2013-05-17 06:20:29 <ncopa> we should have jumped to samba4...
2013-05-17 06:20:32 <ncopa> 08 May 2013
2013-05-17 06:20:32 <ncopa> Samba 3.6.15 Available for Download
2013-05-17 06:20:32 <ncopa> This is the latest stable release of the Samba 3.6 series.
2013-05-17 06:39:27 <barthalion> And new util-linux is out
2013-05-17 06:43:04 <ncopa> hmpf
2013-05-17 06:43:16 <ncopa> i think we will have to replace heimdal with mit krb :/
2013-05-17 06:43:27 <ncopa> they conflict
2013-05-17 06:43:43 <ncopa> so currently you cannot have both samba and nfs server on same box
2013-05-17 06:56:53 <ncopa> rnalrd: what do you think about switching to mit krb5?
2013-05-17 06:57:25 <ncopa> all packages that is built against heimdal just builds fine against mit krb (except the mail prog alpine - but i'm on it)
2013-05-17 06:57:31 <rnalrd> ncopa, i thought we choose heimdal since it suffered less security bugs
2013-05-17 06:57:37 <ncopa> hm
2013-05-17 06:57:58 <ncopa> there is a conflict
2013-05-17 06:58:06 <ncopa> heimdal and krb5
2013-05-17 06:58:29 <ncopa> well naturally, you cannot have both heimdal server and krb5 server installed in parallel
2013-05-17 07:00:40 <ncopa> i dont know if there are less sec bugs there...
2013-05-17 07:00:44 <ncopa> in heimdal
2013-05-17 07:01:17 <rnalrd> from what I could see, i noticed less sec bugs in heimdal, than in MIT
2013-05-17 07:01:28 <rnalrd> but I don't have exact count
2013-05-17 07:01:47 <ncopa> and openbsd picked heimdal
2013-05-17 07:01:50 <ncopa> probably for a reason
2013-05-17 07:01:56 <rnalrd> yeah
2013-05-17 07:02:26 <rnalrd> they are security-minded above all
2013-05-17 07:03:12 <ncopa> but i cannot see any reference on internet that explains why heimdal is more secure
2013-05-17 07:03:46 <barthalion> Maybe just ask OpenBSD devs? Afaik they're helpful
2013-05-17 07:03:55 <ncopa> number of sec bugs might also be related on how many people are searching
2013-05-17 07:03:57 <ncopa> good idea
2013-05-17 07:05:14 <rnalrd> bah, forgot pkgname
2013-05-17 07:06:09 <ncopa> they picked himdal long time ago
2013-05-17 07:06:16 <ncopa> and probably just stick to it
2013-05-17 07:06:31 <ncopa> i dont think its related number of sec bugs.. but i dunno
2013-05-17 07:06:33 <barthalion> So no real reason
2013-05-17 07:06:44 <ncopa> i do have a real reason to swithc to mit though
2013-05-17 07:06:51 <ncopa> nfs does not work with heimdal
2013-05-17 07:06:58 <ncopa> mit seems to be more mainstream
2013-05-17 07:07:04 <ncopa> that said
2013-05-17 07:07:20 <ncopa> i am not comortable to switch the day we do the release...
2013-05-17 07:07:29 <ncopa> but we have a real probelm we need to solve
2013-05-17 07:07:53 <ncopa> atm you cannot have nfs and samba on same box
2013-05-17 07:08:12 <ncopa> i think we can workaround it by not shipping the /etc/krb5.conf file in krb5-libs
2013-05-17 07:08:25 <ncopa> or let krb5-libs replace the heimdal pkg
2013-05-17 07:08:43 <ncopa> i think the libs' names does not clash
2013-05-17 07:08:58 <ncopa> we have those though: ERROR: heimdal: Trying to overwrite usr/bin/kdestroy owned by krb5.
2013-05-17 07:09:12 <ncopa> so basically
2013-05-17 07:09:18 <ncopa> we have 2 options
2013-05-17 07:09:28 <ncopa> 1) pull a switch to mit-krb5 in last minute stunt
2013-05-17 07:10:06 <ncopa> 2) try acoomodate mit krb5 pkgs so they dont conflict (i think gentoo does this) and keep things as they are
2013-05-17 07:10:54 <ncopa> since krb5 is the newcomer, option 2 is probably the most reasonable option
2013-05-17 07:16:16 <ncopa> here is what openbsd ppl said
2013-05-17 07:16:18 <ncopa> <IcePic> at the time, it was better. It might still be, but obsd hasn't upgraded in a while.
2013-05-17 07:16:18 <ncopa> <IcePic> also, mit equals the US, which means "we rape you in customs if you bring single-DES out of our borders"
2013-05-17 07:16:21 <ncopa> makes sense
2013-05-17 07:26:55 <kaniini> ncopa: have you successfully used bsnes lately?
2013-05-17 07:27:10 <ncopa> nope
2013-05-17 07:27:21 <kaniini> ouch
2013-05-17 07:27:30 <kaniini> because it broke in 0.91 it seems
2013-05-17 07:28:04 <ncopa> hmpf
2013-05-17 07:28:18 <ncopa> i havent used bsnes since the likely
2013-05-17 07:28:26 <ncopa> been plaing with other things lately
2013-05-17 07:28:33 <ncopa> like javascript...
2013-05-17 07:28:49 <kaniini> file selection box was bugged, so i fixed that with a patch, but it seems the emulator does not actually start
2013-05-17 07:28:58 <kaniini> i might just do a revert
2013-05-17 07:29:25 <kaniini> i got annoyed with it and wound up just installing alpine-x86 in a chroot and using zsnes
2013-05-17 07:29:53 <ncopa> hm
2013-05-17 07:29:54 <ncopa> ok
2013-05-17 07:30:02 <ncopa> i get "illegal instruction" on my laptop
2013-05-17 07:30:06 <ncopa> yeah, you can revert it
2013-05-17 07:37:25 <kaniini> are heimdal and krb5 abi compatible?
2013-05-17 07:41:15 <ncopa> dont think so
2013-05-17 07:41:43 <ncopa> currently i'm thinking if /etc/krb5.conf is compatible
2013-05-17 07:42:08 <ncopa> i think I'd like to have a krb5-conf package
2013-05-17 07:42:24 <ncopa> or, krb5-conf-example package
2013-05-17 07:42:35 <ncopa> its only the /etc/krb5.conf that conflicts really
2013-05-17 07:42:54 <ncopa> well you cannot run the kerberos servers in paralell, but why would you
2013-05-17 07:43:22 <ncopa> StarWarsFan's problem is due to he upgraded from v2.4, here the heimdal-libs was not split out
2013-05-17 07:43:31 <ncopa> so he got conflict
2013-05-17 08:04:56 <ncopa> larena: are you ok that we remove the krb5.conf from heimdal pkg and move it to a separate krb5-conf package?
2013-05-17 08:05:09 <ncopa> then the file can be sharde by heimdal and krb5
2013-05-17 08:05:23 <ncopa> i also have some minor cleanups in the hedimal apkbuild
2013-05-17 08:05:44 <ncopa> larena: are you ok with this? http://sprunge.us/bLUd
2013-05-17 08:06:17 <ncopa> the main diff is that both heimdal and heimdal-libs depends on krb5-conf, (which only provides /etc/krb5.conf)
2013-05-17 08:06:42 <ncopa> the krb-5conf is only a convenience pkg
2013-05-17 08:26:12 <ncopa> i wonder if i should sneak in samba-4.0.x in there...
2013-05-17 08:27:06 <barthalion> I think so
2013-05-17 08:27:17 <ncopa> i think i got it compile
2013-05-17 08:27:23 <ncopa> but might need some cleanup
2013-05-17 08:27:41 <barthalion> 3.6.15 is last release of 3.6 branch, right?
2013-05-17 08:27:49 <ncopa> thats what they said yes
2013-05-17 08:28:01 <ncopa> which is why i'd like to have samba4
2013-05-17 08:28:19 <barthalion> Hm, but I see this sentence every release
2013-05-17 08:28:25 <ncopa> :)
2013-05-17 08:28:35 <barthalion> https://www.samba.org/samba/latest_news.html
2013-05-17 08:28:43 <barthalion> 3.6.15, 14, 13, 11
2013-05-17 08:28:58 <barthalion> So I don't think they mean 3.6 is discontinued
2013-05-17 08:28:58 <ncopa> ok
2013-05-17 08:29:07 <ncopa> i read it as 'last'
2013-05-17 08:29:12 <ncopa> meh
2013-05-17 08:29:13 <ncopa> ok
2013-05-17 08:29:16 <barthalion> :D
2013-05-17 08:29:20 <ncopa> then i think we're set
2013-05-17 08:32:53 <StarWarsFan> ncopa: to the utf-8 things we talked some days ago:
2013-05-17 08:32:57 <StarWarsFan> take a look here: http://pastebin.com/SHCz4cuA
2013-05-17 08:33:05 <StarWarsFan> with this it seems to work :-)
2013-05-17 08:36:12 <StarWarsFan> so it seems that the unicode switch via env-variables is not working properly...
2013-05-17 08:49:51 <ncopa> StarWarsFan: then i'd expect it to work if we find the correct magic env var...
2013-05-17 08:51:12 <ncopa> ha
2013-05-17 08:51:35 <ncopa> try set LANG=en.UTF
2013-05-17 09:03:06 <ncopa> yup
2013-05-17 09:03:26 <ncopa> setting LANG=en.utf8 solves utf8 in shell
2013-05-17 09:03:42 <ncopa> StarWarsFan: thanks a bunch for that!
2013-05-17 09:03:46 <ncopa> lont time annoyance..
2013-05-17 09:03:51 <ncopa> long*
2013-05-17 09:05:05 <StarWarsFan> you're welcome :-)
2013-05-17 09:05:50 <build64-2-5> files from v2.5.4-89-g832dda9 uploaded
2013-05-17 09:56:12 <build64-2-5> files from v2.5.4-91-g745c41c uploaded
2013-05-17 09:56:48 <build-2-4> files from v2.4.10-56-gdd89521 uploaded
2013-05-17 11:26:27 <ncopa> argh
2013-05-17 11:26:33 <ncopa> setup-bootable is broke :-(
2013-05-17 11:34:18 <ncopa> no its not
2013-05-17 11:34:21 <ncopa> i think it works
2013-05-17 11:50:30 <ncopa> virtio does not work with xen
2013-05-17 11:50:45 <ncopa> that is, running xen in kvm
2013-05-17 12:22:30 <ncopa> ok
2013-05-17 12:22:44 <ncopa> i need a bit help with some hacking/coding
2013-05-17 12:22:59 <ncopa> perl + awk or C or Lua wizards
2013-05-17 12:23:04 <ncopa> http://bugs.alpinelinux.org/issues/1524
2013-05-17 12:23:20 <ncopa> its 5 lines of perl that costs us 46MB
2013-05-17 12:24:01 <ncopa> the full file http://sprunge.us/JDCb
2013-05-17 12:33:11 <ncopa> xendomains init.d script is broke
2013-05-17 12:33:27 <ncopa> at least with screen
2013-05-17 12:53:37 <ncopa> xen is pretty nice
2013-05-17 14:15:30 <ncopa> algitbot: build 2.6-stable
2013-05-17 14:17:01 <ncopa> algitbot: build 2.6-stable
2013-05-17 14:17:34 <ncopa> algitbot: hello
2013-05-17 14:17:37 <ncopa> algitbot: hi
2013-05-17 14:17:45 <ncopa> algitbot: build 2.6-stable
2013-05-17 14:18:50 <ncopa> ok its a bug in algitbot
2013-05-17 14:20:23 <ncopa> algitbot: build 2.6-stable
2013-05-17 14:20:33 <ncopa> here we go
2013-05-17 14:27:46 <ncopa> what should i mention in the release notes?
2013-05-17 19:04:06 <barthalion> ncopa: Don't forget to update downloads page
2013-05-18 04:09:07 <super_duck> is package done inside of a chroot or fakeroot or on my real filesystem?
2013-05-18 04:09:16 <super_duck> nevermind...missed a variable
2013-05-18 12:35:47 <super_duck> the Alpine webpage has http://wiki.alpinelinux.org/wiki/Include:Apkbuild-pypi but I can't find the package, does it still exist?
2013-05-18 13:24:44 <super_duck> question on package naming conventions
2013-05-18 13:24:57 <super_duck> I'm packaging zeitgeist and down the chain it wants isodate.
2013-05-18 13:25:07 <super_duck> Should I package them up as py-<name> or as <name> if they'
2013-05-18 13:25:14 <super_duck> they're fairly recognizable standalone projects?
2013-05-18 18:48:14 <barthalion> super_duck: I would name it py-isodate
2013-05-19 05:26:15 <hl> https://www.youtube.com/watch?v=edflm7Hh3hs
2013-05-19 05:26:20 <hl> whoops.
2013-05-20 06:10:20 <build64-2-5> files from v2.5.4-92-gbf75571 uploaded
2013-05-20 09:41:55 <build-2-4> files from v2.4.10-57-g13e7303 uploaded
2013-05-20 09:43:24 <ncopa> build64-2-4: retry
2013-05-20 11:28:53 <build-2-4> files from v2.4.10-61-ga2b39d6 uploaded
2013-05-20 11:56:27 <build-2-4> files from v2.4.10-64-g641df89 uploaded
2013-05-20 12:19:32 <build64-2-4> files from v2.4.10-64-g641df89 uploaded
2013-05-20 12:57:59 <build64-2-4> files from v2.4.10-65-g4294da6 uploaded
2013-05-20 13:15:02 <build64-2-4> files from v2.4.10-66-g8cdc30c uploaded
2013-05-20 13:15:12 <build-2-4> files from v2.4.10-66-g8cdc30c uploaded
2013-05-20 13:16:17 <build64-2-4> files from v2.4.11 uploaded
2013-05-20 13:16:36 <build-2-4> files from v2.4.11 uploaded
2013-05-20 13:24:23 <build-2-4> release v2.4.11 uploaded
2013-05-20 13:27:39 <build64-2-4> release v2.4.11 uploaded
2013-05-21 08:38:50 <clandmeter> ncopa: here?
2013-05-21 08:39:01 <ncopa> hi
2013-05-21 08:40:19 <clandmeter> morning
2013-05-21 08:41:19 <clandmeter> ncopa: what is the current idea about alpine ARM?
2013-05-21 08:41:57 <kaniini> some guy offered to run builders
2013-05-21 08:42:06 <kaniini> the new architecture should make it easier to run builders.
2013-05-21 08:42:26 <barthalion> And I guess that armv7 will be supported?
2013-05-21 08:42:34 <clandmeter> hardware is not really an issue i think
2013-05-21 08:42:44 <kaniini> i would assume we want armv4m+
2013-05-21 08:42:45 <clandmeter> i am able to support it
2013-05-21 08:42:51 <clandmeter> for hardware that is
2013-05-21 08:42:59 <ncopa> hm
2013-05-21 08:43:07 <kaniini> i will be on pseudo-holiday june 7 through 23rd
2013-05-21 08:43:12 <kaniini> we can try then?
2013-05-21 08:43:26 <ncopa> i think the bottleneck is finding someone with enough tech skills and enough time to actually do it
2013-05-21 08:43:39 <kaniini> (well, on holiday from wonderous startup, to look into other employment opportunities)
2013-05-21 08:44:04 <kaniini> time isn't the issue, lack of hardware is my problem
2013-05-21 08:44:11 <clandmeter> i can buy a ODROID-U2
2013-05-21 08:44:24 <clandmeter> put it in our server room with hd attached.
2013-05-21 08:44:25 <kaniini> the main thing is we want a builder that can keep up
2013-05-21 08:44:27 <kaniini> but...
2013-05-21 08:44:28 <barthalion> Beaglebone Black is pretty cheap
2013-05-21 08:44:33 <kaniini> it shouldn't be a problem these days
2013-05-21 08:44:41 <kaniini> because you can get like 2.0ghz arm chip
2013-05-21 08:44:43 <ncopa> hw is not the problem
2013-05-21 08:44:45 <clandmeter> U2 is quad core with 2GB
2013-05-21 08:44:53 <ncopa> the problem is to do the porting
2013-05-21 08:45:05 <clandmeter> ncopa: thats why i asked you :)
2013-05-21 08:45:06 <kaniini> well, i would assume uclibc/arm is decent
2013-05-21 08:45:17 <ncopa> we talked about try musl...
2013-05-21 08:45:47 <kaniini> i think we should stick with one libc implementation in mainline
2013-05-21 08:46:00 <ncopa> uclibc code is a mess
2013-05-21 08:46:05 <kaniini> if someone wants a musl variant, then we can integrate the code and they can build with ALPINE_LIBC=musl
2013-05-21 08:46:09 <kaniini> yeah, it is.
2013-05-21 08:46:20 <ncopa> if musl had been where it is now, when we started alpine...
2013-05-21 08:46:28 <kaniini> indeed
2013-05-21 08:46:28 <ncopa> then we'd only had musl :)
2013-05-21 08:46:45 <kaniini> and musl is abi-compatible with glibc
2013-05-21 08:46:48 <kaniini> which is interesting
2013-05-21 08:47:00 <ncopa> tries to be
2013-05-21 08:47:36 <kaniini> uclibc is in a bad state of affairs though for sure
2013-05-21 08:47:40 <kaniini> we're carrying a lot of patches again
2013-05-21 08:47:46 <ncopa> if it shows up that we need to patch uclibc to make arm port work
2013-05-21 08:47:48 <kaniini> and upstream seems to be very asleep
2013-05-21 08:47:52 <barthalion> We can work on 2.7 and 2.8 simultaneously, 2.8 would be stable musl release
2013-05-21 08:47:58 <kaniini> musl would be 3.0
2013-05-21 08:48:15 <kaniini> alpine2 was when we added NPTL for xen etc
2013-05-21 08:48:26 <ncopa> switch to musl is a huge job
2013-05-21 08:48:34 <ncopa> yeah it would be 3.0
2013-05-21 08:48:35 <barthalion> versioning issue, you know what I mean
2013-05-21 08:48:50 <ncopa> i thought that wil working on arm
2013-05-21 08:48:53 <kaniini> i think a better approach is start doing tinderbox builds with musl
2013-05-21 08:48:56 <ncopa> you'd need to crosscompile anyways
2013-05-21 08:48:59 <ncopa> and bootstrapping
2013-05-21 08:49:09 <ncopa> might be worth take a look at musl at the same time
2013-05-21 08:49:11 <kaniini> nah, we can bootstrap like we did x86_64
2013-05-21 08:49:27 <kaniini> boot the porting machine into debian/arm
2013-05-21 08:49:31 <kaniini> bootstrap from that
2013-05-21 08:50:40 <ncopa> other things i'd like to do for v2.7
2013-05-21 08:50:42 <ncopa> python3
2013-05-21 08:50:45 <ncopa> lua-5.2
2013-05-21 08:50:54 <ncopa> gnome3?
2013-05-21 08:51:01 <ncopa> qt5
2013-05-21 08:51:03 <kaniini> gnome3 requires systemd :/
2013-05-21 08:51:10 <ncopa> he???
2013-05-21 08:51:14 <barthalion> yeah, don't touch gnome
2013-05-21 08:51:17 <clandmeter> :)
2013-05-21 08:51:26 <barthalion> eventually drop it all
2013-05-21 08:51:34 <kaniini> python3 might be a no-go
2013-05-21 08:51:50 <kaniini> there's a few major libs that are still stuck on python2
2013-05-21 08:51:57 <ncopa> sigh
2013-05-21 08:52:08 <ncopa> we might need py2-<lib>
2013-05-21 08:52:18 <barthalion> probably
2013-05-21 08:52:39 <kaniini> https://python3wos.appspot.com/
2013-05-21 08:53:14 <kaniini> boto, paste and flask are hard blockers there
2013-05-21 08:53:34 <kaniini> way i see it is simple
2013-05-21 08:53:43 <kaniini> if we push py3 and these libs are broken
2013-05-21 08:54:21 <kaniini> we'll shake a lot of good feelings towards using alpine in *shudders* 'the cloud' for *shudders* 'apps'
2013-05-21 08:54:36 <kaniini> where it is starting to gain a lot of attention due to it's low footprint
2013-05-21 08:54:52 <barthalion> we can provide both
2013-05-21 08:54:55 <kaniini> but something that has a low footprint but doesn't support the frameworks people are using... :P
2013-05-21 08:54:58 <ncopa> and we will end up needing both py2-* and py3-* which i'd like to avoid if possible
2013-05-21 08:55:19 <barthalion> but it's not very comfortable to mess with shebangs
2013-05-21 08:55:27 <kaniini> we can't do what arch did
2013-05-21 08:55:34 <kaniini> that will burn serious bridges
2013-05-21 08:56:02 <kaniini> on the other hand, arch is not something one should deploy on server anyway imho
2013-05-21 08:56:14 <ncopa> so
2013-05-21 08:56:21 <ncopa> our only way forward is
2013-05-21 08:56:27 <ncopa> add a python3 package
2013-05-21 08:56:37 <ncopa> "python" == python2
2013-05-21 08:56:43 <kaniini> yes, it seems so for now.
2013-05-21 08:56:47 <ncopa> and start adding py3-*
2013-05-21 08:57:18 <ncopa> i wonder if we should do the same for lua
2013-05-21 08:57:23 <ncopa> lua52
2013-05-21 08:57:51 <kaniini> i am not familiar with the situation there, i just know that people like alpine for spawning up lightweight python environments
2013-05-21 08:58:22 <ncopa> lightweight and python on same sentence :)
2013-05-21 08:58:28 <ncopa> is a contradiction :)
2013-05-21 08:58:34 <kaniini> well
2013-05-21 08:58:37 <ncopa> i know what you mean
2013-05-21 08:58:45 <kaniini> it is lightweight compared to say: RedHat Enterprise Linux 6
2013-05-21 08:59:01 <ncopa> and i think we want give them that
2013-05-21 08:59:33 <kaniini> honestly, if we just continue on a cadence of 'shipping software that works in the real world'
2013-05-21 09:00:00 <kaniini> i think we may gain against RHEL in enterprise segment
2013-05-21 09:00:33 <kaniini> py3 doesn't yet work in 'the real world'
2013-05-21 09:00:52 <ncopa> we already get emails about "alpine vs RHEL vs freebsd netbsd"
2013-05-21 09:00:56 <kaniini> it is getting there slowly :)
2013-05-21 09:01:13 <hl> python 3 works if you pick the right packages
2013-05-21 09:01:23 <kaniini> yeah but people want to use flask
2013-05-21 09:01:27 <hl> i don't see the harm in a python3 package, in fact I would welcome it.
2013-05-21 09:01:27 <barthalion> everything works if you pick the right *
2013-05-21 09:01:37 <ncopa> :)
2013-05-21 09:01:45 <kaniini> hl: nothing wrong with shipping a python3, no
2013-05-21 09:01:47 <ncopa> i think we should work towards py3
2013-05-21 09:02:03 <barthalion> same with lua 5.2, but provide 5.1 as default
2013-05-21 09:02:07 <kaniini> hl: but /usr/bin/env python has to remain python2 until there is a lot less red on that site
2013-05-21 09:02:09 <kaniini> ;)
2013-05-21 09:02:19 <barthalion> ncopa: and maybe switch to luajit someday
2013-05-21 09:02:29 <ncopa> im sceptig to *jit
2013-05-21 09:02:35 <ncopa> due to PaX
2013-05-21 09:02:38 <hl> kaniini: well, yes, if you're not doing /usr/bin/env python3 for py3k that's a bit nuts
2013-05-21 09:02:40 <barthalion> ah, true
2013-05-21 09:02:54 <barthalion> well, but 5.1 won't get any updates I guess
2013-05-21 09:02:57 <kaniini> hl: yes, arch made 'python' python3
2013-05-21 09:03:03 <kaniini> hl: we certainly don't want to do that
2013-05-21 09:03:07 <hl> hmmm, yes
2013-05-21 09:03:10 <kaniini> it will scare many people
2013-05-21 09:03:35 <kaniini> i mean, we should do it eventually, but it's probably a year out before we can make it default python
2013-05-21 09:03:45 <kaniini> and we'll still need to provide 2.x environment for backwards compat.
2013-05-21 09:03:53 <hl> by the way, i noticed a zfs APKBUILD in testing. is that what you use, kaniini?
2013-05-21 09:04:00 <kaniini> yes
2013-05-21 09:04:05 <hl> neato.
2013-05-21 09:04:14 <ncopa> thats pretty cool
2013-05-21 09:04:19 <ncopa> can i move it to main?
2013-05-21 09:04:36 <kaniini> go for it
2013-05-21 09:04:48 <ncopa> good!
2013-05-21 09:04:50 <ncopa> will do
2013-05-21 09:05:21 <clandmeter> zfs the filesystem?
2013-05-21 09:05:21 <kaniini> maybe systemd will eventually not require things like dbus too
2013-05-21 09:05:36 <hl> clandmeter: yes
2013-05-21 09:05:36 <ncopa> i think when kdbus is in mainline
2013-05-21 09:05:39 <kaniini> that would sure be swell
2013-05-21 09:05:45 <kaniini> hahah
2013-05-21 09:05:53 <hl> tbh, I think dbus is the least of the reasons to avoid systemd
2013-05-21 09:05:54 <kaniini> kdbus/kbus/kevent
2013-05-21 09:05:57 <clandmeter> zfs has some issues i think with uclibc
2013-05-21 09:06:14 <kaniini> oh man, don't get me started on that :P
2013-05-21 09:06:18 <ncopa> the zfs tools
2013-05-21 09:06:23 <clandmeter> yes
2013-05-21 09:06:24 <kaniini> zfs tools work fine for me
2013-05-21 09:06:32 <ncopa> yes please lets drop systemd discuss for now
2013-05-21 09:06:36 <kaniini> i'm only using the zpools now
2013-05-21 09:06:40 <kaniini> though
2013-05-21 09:06:45 <kaniini> i don't trust zfs itself
2013-05-21 09:06:48 <clandmeter> there are some tools that dont work afaik
2013-05-21 09:06:52 <ncopa> systemd its not gonna happen anytime soon
2013-05-21 09:06:52 <hl> why not just use LVM in that case?
2013-05-21 09:06:58 <clandmeter> to analyze zfs pools
2013-05-21 09:07:09 <clandmeter> i dont remember the cmd
2013-05-21 09:07:31 <kaniini> hl: better control over I/O latencies and SLAs
2013-05-21 09:07:38 <hl> hmm, interesting.
2013-05-21 09:07:42 <hl> anyway
2013-05-21 09:07:44 <clandmeter> kaniini: zdb does it work
2013-05-21 09:07:46 <hl> is alpine planning on staging a move to musl?
2013-05-21 09:07:55 <kaniini> hl: not for 2.7
2013-05-21 09:08:06 <ncopa> hl: we want musl, when it is practically doable
2013-05-21 09:08:11 <kaniini> ncopa: the other thing i would like to do is look at replacing busybox with rob landley's toybox
2013-05-21 09:08:16 <ncopa> i dont think it is yet
2013-05-21 09:08:26 <ncopa> i havent looked at toybox for a while
2013-05-21 09:08:41 <kaniini> it's getting there.  i might send him patches to get it up to speed.
2013-05-21 09:08:49 <kaniini> busybox guys make me nervous.
2013-05-21 09:08:54 <ncopa> but bb has become pretty susv compliant
2013-05-21 09:09:09 <kaniini> they like that gpl-violations.org guy
2013-05-21 09:09:23 <kaniini> so, this makes toybox attractive ;)
2013-05-21 09:09:35 <hl> busybox appears to be some sort of favourite honeypot for embedded systems gpl crusaders
2013-05-21 09:09:57 <kaniini> musl has a more attractive license too
2013-05-21 09:09:59 <kaniini> :P
2013-05-21 09:10:10 <hl> MIT, nice.
2013-05-21 09:10:21 <ncopa> i'd be more interested in tech reasons...
2013-05-21 09:10:28 <kaniini> well, yes
2013-05-21 09:10:32 <kaniini> that too
2013-05-21 09:10:36 <kaniini> obviously.
2013-05-21 09:10:48 <hl> http://www.etalabs.net/compare_libcs.html
2013-05-21 09:10:52 <hl> this is probably dated, though
2013-05-21 09:11:02 <kaniini> one big win for musl is
2013-05-21 09:11:10 <kaniini> utf-8 is correctly isolated from locales
2013-05-21 09:11:19 <ncopa> and the code is soooo much cleaner
2013-05-21 09:11:25 <hl> come to think of it
2013-05-21 09:11:38 <hl> are there any libcs which actually separate "system calls" from "ANSI C"? or is that just a pipe dream
2013-05-21 09:12:39 <kaniini> dalias is also a lot easier to work with
2013-05-21 09:12:55 <kaniini> as in, you can just poke him on IRC with a pull request and he'll merge it
2013-05-21 09:12:57 <clandmeter> alpine port should be easy according to musle wiki :) http://wiki.musl-libc.org/wiki/Main_Page
2013-05-21 09:13:14 <ncopa> clandmeter: i saw that yes
2013-05-21 09:13:15 <kaniini> clandmeter: yes, in theory, we just need to write an APKBUILD for it
2013-05-21 09:13:21 <kaniini> clandmeter: and then switch ALPINE_LIBC to use it
2013-05-21 09:13:31 <kaniini> in a few weeks i'll look into it.
2013-05-21 09:13:34 <ncopa> and build gcc
2013-05-21 09:13:49 <kaniini> i'll have access to some very heavy compute resources for a little bit
2013-05-21 09:13:54 <clandmeter> but does it bring anything functional?
2013-05-21 09:14:04 <clandmeter> i still really like an ARM port
2013-05-21 09:14:06 <ncopa> utf-8?
2013-05-21 09:14:07 <kaniini> well... do you consider 'readable code' functional?
2013-05-21 09:14:10 <kaniini> and utf-8?
2013-05-21 09:14:20 <kaniini> what concerns me on that graph is 'gmon profiling'
2013-05-21 09:14:25 <kaniini> -fprofile does work on alpine
2013-05-21 09:14:35 <hl> are you telling me uclibc does not handle utf-8 properly?
2013-05-21 09:14:44 <kaniini> yes
2013-05-21 09:14:47 <kaniini> it does not
2013-05-21 09:14:48 <hl> what in the hell
2013-05-21 09:14:49 <kaniini> well
2013-05-21 09:14:52 <kaniini> it "can"
2013-05-21 09:14:57 <kaniini> we have some hacks
2013-05-21 09:15:08 <ncopa> export LANG=en.utf8
2013-05-21 09:15:08 <kaniini> to try to make it work better, but...
2013-05-21 09:15:10 <ncopa> helps
2013-05-21 09:15:41 <kaniini> musl's goals are more in line with what we want than uclibc's too
2013-05-21 09:16:01 <kaniini> uclibc guys are afraid of features
2013-05-21 09:16:22 <barthalion> I don't understand Richard Johnson messages, he doesn't report bug and doesn't ask either
2013-05-21 09:16:23 <kaniini> dalias is okay with features as long as they aren't too bloaty
2013-05-21 09:17:05 <kaniini> barthalion: qq.com is a chinese mail provider
2013-05-21 09:17:27 <kaniini> my guess is he is complaining that we do not support CJK very well (likely because our utf-8 support is an entire hackjob right now ;))
2013-05-21 09:17:55 <barthalion> :D
2013-05-21 09:18:20 <kaniini> we might be able to do better if we can figure out what they are talking about re: "locale files"
2013-05-21 09:18:47 <kaniini> but then that creates a new question of
2013-05-21 09:18:53 <kaniini> does this change our ABI :/
2013-05-21 09:21:52 <ncopa> thats my main concern with locales
2013-05-21 09:23:37 <kaniini> re: http://lists.alpinelinux.org/alpine-devel/1139.html
2013-05-21 09:23:57 <kaniini> --disable-fixed-point is so we build packages that don't use the FPU, so we can work on i486 without an FPU emulation hit
2013-05-21 09:24:17 <kaniini> if we don't care about performance on i486 anymore, we could remove it.
2013-05-21 09:24:31 <kaniini> we should probably remove it on x86_64, but i think it is just ignored there anyway.
2013-05-21 09:25:17 <ncopa> i486 has fpu
2013-05-21 09:25:38 <ncopa> i386 has not
2013-05-21 09:25:49 <ncopa> maybe i468 sx had not
2013-05-21 09:26:21 <ncopa> i think alix boards are our current lowest end
2013-05-21 09:26:35 <ncopa> via c7
2013-05-21 09:26:40 <royger> regarding that, are we going to ship a 32bit kernel with PAE enabled?
2013-05-21 09:26:50 <ncopa> oh yes
2013-05-21 09:26:55 <ncopa> that was one point for v2.7
2013-05-21 09:27:04 <ncopa> i was thinking enable PAE for 32bit
2013-05-21 09:27:22 <ncopa> and ship a nopae kernel in addition
2013-05-21 09:27:28 <royger> I used to have a C7, but it no longer works :(
2013-05-21 09:27:39 <ncopa> my laptop has c7
2013-05-21 09:27:58 <ncopa> the c7 padlock crypto engine rocks
2013-05-21 09:28:40 <kaniini> another interesting target: x32-abi
2013-05-21 09:28:51 <royger> I guess mine burned out due to bad refrigeration
2013-05-21 09:29:03 <ncopa> i doubt x32-abi is worth the effort
2013-05-21 09:29:22 <kaniini> hmm, yes
2013-05-21 09:29:34 <kaniini> the current footprint on x86-64 is pretty acceptable
2013-05-21 09:35:53 <hl> I like the toybox news page.
2013-05-21 09:35:58 <hl> "The major difference between a thing that might go wrong and a thing that cannot possibly go wrong is that when a thing that cannot possibly go wrong goes wrong it usually turns out to be impossible to get at or repair."
2013-05-21 09:37:35 <ncopa> lol
2013-05-21 09:37:39 <ncopa> true true
2013-05-21 09:50:12 <ncopa> kaniini, btw do you think you could pastebin/sprunge me the wip you have for ff-21?
2013-05-21 10:00:18 <kaniini> i'll get it to you in few hours
2013-05-21 10:14:57 <build-2-2> files from v2.2.3-231-gc624044 uploaded
2013-05-21 10:14:58 <build64-2-3> files from v2.3.6-208-g43633e5 uploaded
2013-05-21 10:15:45 <build-2-4> files from v2.4.11-1-ge0bbe76 uploaded
2013-05-21 10:16:19 <build64-2-5> files from v2.5.4-93-gc6addec uploaded
2013-05-21 10:55:44 <barthalion> I'm going to update poppler today
2013-05-21 10:55:54 <barthalion> http://sprunge.us/fbJM is it everything to rebuild?
2013-05-21 11:06:40 <ncopa> probably but...
2013-05-21 11:06:52 <ncopa> iirc there was some issue with gtk2
2013-05-21 11:07:03 <ncopa> i think gtk2 was dropped
2013-05-21 11:07:14 <ncopa> which i think will open a can of worms
2013-05-21 11:07:39 <ncopa> i recommend you to work on it a separate local branch
2013-05-21 11:08:04 <barthalion> Sure
2013-05-21 11:15:52 <ncopa> clandmeter
2013-05-21 11:15:58 <ncopa> we have an issue with redmine
2013-05-21 11:16:16 <ncopa> try filter on all 'closed' issues
2013-05-21 11:16:59 <clandmeter> ok i see
2013-05-21 11:20:11 <ncopa> clandmeter: do you think we dare to upgrade alpine-mysql to v2.6?
2013-05-21 11:21:11 <clandmeter> sure
2013-05-21 11:21:19 <clandmeter> dump it first
2013-05-21 11:23:42 <clandmeter> ncopa: did you see the actual error?
2013-05-21 11:23:57 <ncopa> short look yes
2013-05-21 11:31:18 <ncopa> mysql updated
2013-05-21 11:31:28 <ncopa> i kinda like alpine linux :)
2013-05-21 11:31:38 <ncopa> took less than 10 sec
2013-05-21 11:31:49 <ncopa> mysqldump -p --all-databases > dumpall.mysql && /etc/init.d/mysq
2013-05-21 11:31:49 <ncopa> l stop && apk upgrade -U && /etc/init.d/mysql start
2013-05-21 11:33:44 <clandmeter> ncopa: did you recently update redmine?
2013-05-21 11:33:57 <ncopa> i think maybe yes
2013-05-21 11:34:02 <ncopa> i am thinking the same i think
2013-05-21 11:34:09 <ncopa> should probalby check the schema
2013-05-21 11:34:16 <clandmeter> upgrade the schema yes
2013-05-21 11:35:07 <clandmeter> rake db:migrate RAILS_ENV=production
2013-05-21 11:35:12 <ncopa> just did
2013-05-21 11:35:51 <ncopa> error still happens
2013-05-21 11:36:22 <clandmeter> ah
2013-05-21 11:36:26 <clandmeter> i think i know
2013-05-21 11:36:32 <clandmeter> tmp too small
2013-05-21 11:36:37 <ncopa> i think #1924 is related
2013-05-21 11:36:41 <ncopa> ok?
2013-05-21 11:36:45 <clandmeter> vserver....
2013-05-21 11:36:47 <ncopa> aha
2013-05-21 11:36:50 <ncopa> which of them?
2013-05-21 11:36:54 <clandmeter> mysql
2013-05-21 11:37:02 <ncopa> I'll increase it
2013-05-21 11:37:03 <clandmeter> tmp gets full so mysql has issues
2013-05-21 11:38:14 <ncopa> ok
2013-05-21 11:38:16 <ncopa> good catch
2013-05-21 11:38:19 <ncopa> works now
2013-05-21 11:45:08 <ncopa> i think I'll upgrade alpine-proxy and alpine-www too
2013-05-21 11:45:31 <clandmeter> i still need to check redmine deps
2013-05-21 11:45:32 <ncopa> clandmeter: do you know any issues with php 5.4 and drupal?
2013-05-21 11:45:50 <clandmeter> not that i know
2013-05-21 11:46:23 <ncopa> i really really like what we did in vservers
2013-05-21 11:46:28 <ncopa> its so convenient
2013-05-21 11:46:36 <ncopa> we can upgrade parts
2013-05-21 11:46:44 <ncopa> eg, upgrade mysql separately
2013-05-21 11:46:51 <ncopa> then nginx separately
2013-05-21 11:47:02 <clandmeter> i disabled ngnix caching recently
2013-05-21 11:47:10 <clandmeter> it gave some issues
2013-05-21 11:47:15 <ncopa> ok
2013-05-21 11:47:15 <clandmeter> with repo's
2013-05-21 11:47:22 <ncopa> nginx is updated now
2013-05-21 11:47:28 <clandmeter> and i think i had it misconfigured
2013-05-21 11:48:00 <clandmeter> i saw some feature was added to ngnix?
2013-05-21 11:48:13 <clandmeter> does that need any configuration?
2013-05-21 11:48:21 <ncopa> apparently not
2013-05-21 11:48:27 <ncopa> seems like it just worked
2013-05-21 11:48:31 <ncopa> i already upgraded it
2013-05-21 11:48:36 <ncopa> cowboystyle
2013-05-21 11:49:11 <ncopa> will do alpine-www now
2013-05-21 11:49:18 <ncopa> the only one i dont dare is alpine-bugs
2013-05-21 11:50:27 <clandmeter> let us first fix the rb packages
2013-05-21 11:50:31 <clandmeter> then the update should be easy
2013-05-21 11:51:30 <clandmeter> what i mean about ngnix is spdy
2013-05-21 11:51:38 <ncopa> i broke www...
2013-05-21 11:53:38 <clandmeter> did you restart php?
2013-05-21 11:53:59 <ncopa> that helped :)
2013-05-21 11:54:02 <ncopa> thanks!
2013-05-21 11:54:10 <ncopa> i started to sweat :)
2013-05-21 11:54:48 <ncopa> so, only alpine-bugs need to upgrade to 2.6
2013-05-21 11:54:53 <ncopa> and alpine-vserver itself
2013-05-21 11:54:53 <clandmeter> its the one i always miss
2013-05-21 12:28:28 <build64-2-5> files from v2.5.4-99-g84b9026 uploaded
2013-05-21 12:32:52 <build64-2-3> files from v2.3.6-211-g7338a7f uploaded
2013-05-21 12:57:04 <vkrishn> ncopa, there is some discussion on adding some msgs while upgrading major services
2013-05-21 12:57:20 <vkrishn> or a kernel upgrade
2013-05-21 12:57:54 <vkrishn> I think this was before v2.5 release, I forgot to bring it up
2013-05-21 13:01:38 <vkrishn> idea was a trigger watching /etc/init.d
2013-05-21 13:07:45 <ncopa> any idea for a pkgname?
2013-05-21 13:07:51 <ncopa> upgrade-restart-notify?
2013-05-21 13:08:24 <vkrishn> upgrade-notify
2013-05-21 13:09:25 <vkrishn> I was reading about monit, dnotify, imon ... are they related
2013-05-21 13:09:32 <vkrishn> fam...
2013-05-21 13:14:24 <clandmeter> ncopa: is there an option (to use in a script) in apk to check if there are package updates?
2013-05-21 13:14:28 <vkrishn> I mean when implementing 'upgrade-notify' will fam or monit be needed
2013-05-21 13:16:30 <ncopa> clandmeter: apk version -U
2013-05-21 13:17:11 <vkrishn> alt name, apk-notify
2013-05-21 13:19:31 <vkrishn> ncopa, there was some report of sound not working in laptop, does it have hdmi out?
2013-05-21 13:21:00 <vkrishn>  I recalled while reading http://lists.debian.org/debian-knoppix/2013/05/msg00006.html,
2013-05-21 13:31:04 <ncopa> i think we are affected of it too
2013-05-21 13:31:13 <ncopa> i dont have that kind of hw though
2013-05-21 13:33:17 <vkrishn> should I add a bugs.a.o for it ?
2013-05-21 13:39:43 <ncopa> is there anything we can do to fix it?
2013-05-21 14:16:40 <vkrishn> the posts says ... The fix is (probably),....
2013-05-21 14:18:59 <vkrishn>  would try to search
2013-05-21 14:22:57 <vkrishn> can `apk audit` do the checking?
2013-05-21 14:23:01 <vkrishn> I think wiki.a.o is slooow again
2013-05-21 14:24:44 <rnalrd> vkrishn, could you try to access wiki.a.o via these static mappings: 141.101.116.200 and 141.101.117.200 ?
2013-05-21 14:25:03 <vkrishn> ok
2013-05-21 14:25:06 <rnalrd> and see if you can notice any difference
2013-05-21 14:31:08 <vkrishn> there seems a difference but I need to check later again when I have a stable net connection
2013-05-21 14:43:02 <vkrishn> rnalrd, possibility of updating ettercap ?
2013-05-21 14:43:12 <vkrishn> apk audit needs more documentation
2013-05-21 14:44:40 <barthalion> clandmeter: you mean spdy? you need to enable it explicitly
2013-05-21 15:00:23 <clandmeter> barthalion: yes thats what i was refereing to
2013-05-21 15:00:33 <clandmeter> barthalion: in the examples i see only https
2013-05-21 15:00:42 <clandmeter> its interesting to enable for http?
2013-05-21 15:02:21 <barthalion> I don't think it's possible
2013-05-21 15:03:12 <clandmeter> its only for https?
2013-05-21 15:03:12 <barthalion> Or it is
2013-05-21 15:03:33 <clandmeter> i saw that you are the maintainer. so you know everything about ngnix ;-)
2013-05-21 15:03:45 <barthalion> I wish I would :p Let me take a look at docs
2013-05-21 15:04:44 <barthalion> Yeah, I remember correctly, spdy in nginx works only w/ https
2013-05-21 15:05:01 <clandmeter> ok
2013-05-21 15:05:10 <clandmeter> but it doesnt clearly state this anywhere
2013-05-21 15:05:11 <rnalrd> vkrishn, they switched to cmake
2013-05-21 15:05:12 <barthalion> Anyway, you need to add 'spdy' to listen directive
2013-05-21 15:05:30 <rnalrd> i'll update the apkbuild tm
2013-05-21 15:05:34 <rnalrd> gotta go
2013-05-21 15:05:35 <clandmeter> I dont think its interesting enough for https only
2013-05-21 15:05:47 <barthalion> it should be faster anyway
2013-05-21 15:06:35 <clandmeter> pagespeed seems interesting
2013-05-21 15:07:07 <barthalion> requires recompiling nginx, but yes, can be useful
2013-05-21 15:07:31 <clandmeter> there is some fork which allows external modules
2013-05-21 15:09:28 <barthalion> tanglu
2013-05-21 15:09:37 <barthalion> Er, no
2013-05-21 15:09:43 <barthalion> It's Debian fork
2013-05-21 15:09:49 <barthalion> Hrm, I've seen it in AUR
2013-05-21 15:10:23 <vkrishn> rnalrd, thanks
2013-05-21 15:10:26 <barthalion> tengine
2013-05-21 15:10:29 <barthalion> I was close :p
2013-05-21 15:10:37 <barthalion> I need to test it someday
2013-05-21 15:11:18 <clandmeter> yes, it would be a nice alternative and easier for the packager.
2013-05-21 15:14:06 <vkrishn> tengine say "Multiple CSS or JavaScript requests can be combined into one request to reduce downloading time"
2013-05-21 15:14:15 <barthalion> They merge security fixes from nginx pretty quick
2013-05-21 15:14:15 <vkrishn> but does not specify the protocol
2013-05-21 15:14:21 <vkrishn> is it spdy?
2013-05-21 15:14:28 <barthalion> I'll package it later in the week
2013-05-21 15:14:38 <barthalion> vkrishn: I think they're combined server-side
2013-05-21 15:15:39 <barthalion> http://tengine.taobao.org/document/http_concat.html
2013-05-21 15:17:32 <vkrishn> cool, how does browser understand the concat ?
2013-05-21 15:18:08 <barthalion> No idea
2013-05-21 15:18:30 <vkrishn> ah, url based
2013-05-21 15:19:23 <vkrishn> modconcat
2013-05-21 15:20:29 <vkrishn> would be nice to see as alterntive pkg in testing
2013-05-21 15:22:21 <barthalion> I don't touch Apache, sorry :p
2013-05-21 15:24:13 <barthalion> As far I see, tengine in AUR is just modified PKGBUILD for nginx…
2013-05-21 15:24:18 <barthalion> Won't be hard to package
2013-05-21 18:26:47 <vkrishn> barthalion, whats with maintainer/contributor of hiawatha ?
2013-05-21 18:27:00 <vkrishn> I thought you were maintainer
2013-05-21 18:27:32 <vkrishn> btw, thanx for the pkg, good to have something new to test
2013-05-21 18:27:49 <barthalion> Nothing, just didn't add myself
2013-05-21 18:41:57 <clandmeter> barthalion: didnt know polish tld was only .p :)
2013-05-21 18:42:31 <barthalion> pl, I must made a typo
2013-05-21 18:42:46 <clandmeter> i corrected it
2013-05-21 18:42:57 <barthalion> thanks
2013-05-21 18:43:05 <clandmeter> btw, it will only list when you are maintainer
2013-05-21 18:43:12 <clandmeter> and only for new pkgs
2013-05-21 18:43:16 <clandmeter> or updated
2013-05-21 18:43:29 <barthalion> Yeah, I expected that
2013-05-21 18:43:45 <clandmeter> when you go into your profile you can list your pkgs and see stats about views
2013-05-21 18:44:19 <barthalion> I'm not adding my mail to APKBUILDs yet because I'm getting hundreds of spam, and I wait for my mail hoster to react
2013-05-21 18:44:38 <clandmeter> hmm, i thought i saw a few?
2013-05-21 18:44:53 <barthalion> Yeah, but not everywhere
2013-05-21 18:45:08 <clandmeter> request an alpine address
2013-05-21 18:45:56 <barthalion> who should i ask?
2013-05-21 18:46:05 <clandmeter> ping jbilyk
2013-05-21 18:46:09 <jbilyk> hey
2013-05-21 18:46:27 <clandmeter> hi, can you set one up for him
2013-05-21 18:46:29 <barthalion> :D
2013-05-21 18:46:43 <jbilyk> barthalion@alpinelinux.org?
2013-05-21 18:46:49 <barthalion> bpiotrowski would be better
2013-05-21 18:47:11 <jbilyk> barthalion: they're just forwarders to an existing address - where do you want the mail to be forwarded to?
2013-05-21 18:47:45 <barthalion> hm, maybe alpine@bpiotrowski.pl
2013-05-21 19:33:33 <kaniini> hmm
2013-05-21 19:33:38 <kaniini> we do not build ccc-analyzer?
2013-05-21 19:47:31 <jbilyk> barthalion: ok, I'll add that in a bit - can you send me an email reminder to jbilyk@alpinelinux.org so that I don't forget?
2013-05-21 19:53:11 <iilluzion2> some progress in regard of installation/recovery procedures for stream infra...
2013-05-21 19:53:13 <iilluzion2> http://trac.cde.wtbts.net/projects/cde-usnet/wiki/AwsCloudStream
2013-05-21 19:59:41 <kaniini> not sure what that is, but it doesn't load here.
2013-05-21 20:01:56 <kaniini> ncopa: it appears that my xulrunner-21 stuff didn't get git-stash'd
2013-05-21 20:07:24 <kaniini> i am working right now to build clang-analyzer
2013-05-21 20:07:28 <kaniini> as part of clang package
2013-05-21 20:07:30 <kaniini> :)
2013-05-21 20:07:53 <kaniini> it shouldn't take too long
2013-05-21 20:08:01 <kaniini> petrie has 48 threads available to it
2013-05-21 20:08:03 <kaniini> ;)
2013-05-22 06:09:05 <ncopa> morning
2013-05-22 06:09:09 <ncopa> build64-2-2: retry
2013-05-22 06:12:14 <StarWarsFan> moinmoin
2013-05-22 07:10:38 <ncopa> does this look ok?  http://dpaste.org/ySpvK/
2013-05-22 07:14:31 <kaniini> i like it
2013-05-22 07:15:46 <ncopa> im not sure it works as expected
2013-05-22 07:16:04 <ncopa> i compare the /run/openrc/started symlink date
2013-05-22 07:16:12 <ncopa> with the date of /etc/init.d/* script
2013-05-22 07:19:14 <ncopa> it wil use notify-send if found
2013-05-22 07:22:25 <build64-2-2> files from v2.2.3-231-gc624044 uploaded
2013-05-22 07:26:09 <ncopa> wow
2013-05-22 07:26:21 <ncopa> that krb5 issue is really something
2013-05-22 07:26:29 <ncopa> cve 1999!!!
2013-05-22 09:05:22 <clandmeter> ncopa: is it ok to update sqlite? i dont see abi changes, but that apkbuild has some version logic in it which is not completely clear to me.
2013-05-22 09:06:01 <ncopa> i think i updated sqlite alreaady
2013-05-22 09:06:16 <clandmeter> ah yes
2013-05-22 09:06:18 <clandmeter> sorry
2013-05-22 09:18:18 <clandmeter> ncopa: why did you switch zfs to x86?
2013-05-22 09:18:56 <ncopa> aw
2013-05-22 09:19:27 <ncopa> i think i means 'all'
2013-05-22 09:19:49 <ncopa> or 'all'
2013-05-22 09:20:47 <clandmeter> ok
2013-05-22 09:33:44 <vkrishn> clandmeter, sqlite 3.7.17 adds some new feature, can it reviewed to see they can be enabled
2013-05-22 09:35:12 <vkrishn> 1. enhanced extension loading (new ext are "amatch, closure, fuzzer, ieee754, nextchar, regexp, spellfix, and wholenumber.")
2013-05-22 09:35:30 <vkrishn> 2. FTS4 - though this may break BC
2013-05-22 09:53:41 <clandmeter> vkrishn: sure, just cehck it doesnt break anything.
2013-05-22 09:56:26 <vkrishn> (1) - should not, 2 likely
2013-05-22 09:57:55 <vkrishn> FTS4 is recommended, and if its everybody thinks its ok then switching to fts4 would be nice
2013-05-22 09:58:11 <vkrishn> if everybody*
2013-05-22 09:59:42 <clandmeter> what i read fts4 is already added in 2010?
2013-05-22 10:03:38 <vkrishn> -DSQLITE_ENABLE_FTS3=1
2013-05-22 10:03:58 <vkrishn> does FTS4 require such flag
2013-05-22 10:04:08 <vkrishn> to be enabled
2013-05-22 10:12:46 <vkrishn> can both be enabled?
2013-05-22 10:13:06 <clandmeter> When this option is defined in the amalgamation, versions 3 and 4 of the full-text search engine is added to the build automatically.
2013-05-22 10:14:10 <vkrishn> ok
2013-05-22 10:14:57 <clandmeter> so i guess if you need it we can add it.
2013-05-22 10:15:11 <clandmeter> i dont see any other dist using it (the onces i checked)
2013-05-22 10:15:34 <clandmeter> adv ext loading, how to enable that? or is that by default?
2013-05-22 10:16:28 <vkrishn> ok just checked CREATE VIRTUAL TABLE pages USING fts4(title, keywords, body); in alpine v2.5
2013-05-22 10:16:31 <vkrishn> works :)
2013-05-22 10:25:06 <clandmeter> so no need to change anything?
2013-05-22 10:25:47 <vkrishn> would have to install 3.7.17 to see if extra extension are available
2013-05-22 11:12:40 <vkrishn> clandmeter, edge sqlite seems does not have them enabled,
2013-05-22 11:12:50 <vkrishn> I tried command .load spellfix
2013-05-22 11:39:17 <clandmeter> vkrishn: which option do i need to enable.
2013-05-22 11:46:15 <vkrishn> I am still checking, seems amalgamation does not have those misc .c files in them
2013-05-22 11:47:26 <vkrishn> something like gcc -shared <modname>.c -o libSqlite<modname>.so
2013-05-22 11:47:32 <vkrishn> needs to be done
2013-05-22 11:48:18 <vkrishn> or maybe add them to sqlite3.c(amalgamated)
2013-05-22 11:49:19 <vkrishn> pls review the misc mods to see if they are interesting'
2013-05-22 11:50:51 <vkrishn> I think they should not break anything, they just add new feature
2013-05-22 11:51:06 <vkrishn> it may possible to add them statically
2013-05-22 12:03:49 <rnalrd> vkrishn, ettercap updated
2013-05-22 12:05:19 <vkrishn> thanks, rnalrd
2013-05-22 12:05:53 <vkrishn> rnalrd, is cmake an issue in alpine?
2013-05-22 12:06:05 <rnalrd> no
2013-05-22 12:06:14 <vkrishn> ok
2013-05-22 12:06:41 <rnalrd> i just needed to partially rewrite the APKBUILD
2013-05-22 12:06:49 <rnalrd> it wasn't a simple version bump
2013-05-22 12:07:17 <vkrishn> I had checked their changes, seems a lot
2013-05-22 12:09:05 <vkrishn> I think these were the reason, the freeze in debian made the package being dropped from Wheezy by some fraction
2013-05-22 12:13:44 <vkrishn> clandmeter, http://www.sqlite.org/spellfix1.html seems nice
2013-05-22 14:05:08 <vkrishn> clandmeter, It may require sqlite-autoconf + sqlite-src-3071700 to get these extra mods packaged,
2013-05-22 14:05:40 <vkrishn> if too much a hassle, ok to drop
2013-05-22 14:07:30 <vkrishn> or sqlite-amalgamation-3071700
2013-05-22 14:08:29 <vkrishn> I copied spellfix.c from sqlite-src, then
2013-05-22 14:08:42 <vkrishn> pasted in sqlite-amalgamation dir
2013-05-22 14:08:51 <vkrishn> then, gcc -shared spellfix.c -o spellfix.so
2013-05-22 14:08:54 <vkrishn> voila
2013-05-22 14:10:10 <vkrishn> I still don't know how to get that in APKBUILD
2013-05-22 15:00:47 <clandmeter> perl dependency hell....
2013-05-22 21:06:53 <barthalion> sounds like fun
2013-05-22 21:07:13 <clandmeter> im not sure it was :D
2013-05-22 21:07:21 <barthalion> especially considering new Perl and possibly required rebuild :p
2013-05-22 21:07:59 <clandmeter> actually i tagged the first one wrong. should also be testing.
2013-05-22 21:08:11 <clandmeter> I drink too much white beer
2013-05-22 21:08:14 <clandmeter> :)
2013-05-22 21:11:56 <clandmeter> barthalion: are you going to push new perl?
2013-05-22 21:12:25 <barthalion> not yet
2013-05-22 21:12:45 <barthalion> I'm going to fight with poppler tommorow or day after
2013-05-22 21:14:38 <barthalion> About Perl, I'll prepare updated APKBUILD and rebuild list, then ask here for opinions
2013-05-22 21:16:15 <clandmeter> are you expecting issues with new perl?
2013-05-22 21:17:33 <barthalion> Went fine in Arch, but I want to do more testing before I push anything
2013-05-22 21:19:49 <barthalion> I have seen true sentence years ago, better safe than sorry :p
2013-05-22 21:20:34 <clandmeter> are you ever truely safe in a digital world?
2013-05-22 21:23:20 <barthalion> Probably no, but in more mundane meaning, at least I should't break anything
2013-05-22 21:39:13 <clandmeter> i just pushed all those pkgs to find out i most probably dont need them :(
2013-05-23 06:57:54 <ncopa> aw!
2013-05-23 06:57:59 <ncopa> perl breaks things...
2013-05-23 06:58:14 <barthalion> yup
2013-05-23 06:58:39 <ncopa> how do we fix it?
2013-05-23 06:58:40 <barthalion> That's why I was going to do more testing first :p
2013-05-23 06:58:45 <ncopa> i revert it?
2013-05-23 06:59:03 <barthalion> We need to rebuild all modules which aren't written in pure Perl
2013-05-23 06:59:10 <ncopa> thats all?
2013-05-23 06:59:16 <barthalion> I think so.
2013-05-23 06:59:24 <ncopa> i'll fix that then
2013-05-23 06:59:28 <ncopa> while the kernel builds..
2013-05-23 06:59:34 <barthalion> good, thanks
2013-05-23 07:00:59 <clandmeter> :)
2013-05-23 07:10:25 <ncopa> so, i run a for i in perl-*; do (cd $i; . ./APKBUILD; [ "$arch" != "noarch" ] && echo $i); done
2013-05-23 07:11:25 <ncopa> gives me a list of all the perl modules that has not pure perl modules (pure perl should have "noarch")
2013-05-23 07:11:45 <ncopa> then i have this nice little script: "rebuild-since"
2013-05-23 07:11:56 <clandmeter> what about perl apps?
2013-05-23 07:12:05 <ncopa> rebuild-since <commitid> 'rebuild against 5.18'
2013-05-23 07:12:22 <ncopa> i'll deal with those later
2013-05-23 07:12:38 <ncopa> i'm gonna scan for all perl-dev
2013-05-23 07:12:44 <ncopa> grep perl-dev */APKBUILD
2013-05-23 07:12:53 <ncopa> filter out all the perl-* pkgs
2013-05-23 07:12:59 <ncopa> and then have a closer look at the rest
2013-05-23 07:13:51 <clandmeter> barthalion said he was already working on it. not sure how far he got.
2013-05-23 07:15:58 <ncopa> barthalion: i'm sorry about that...
2013-05-23 07:16:09 <barthalion> Nowhere yet, poppler is on top of my list now
2013-05-23 07:16:22 <ncopa> did you figure out if it can do gtk2?
2013-05-23 07:16:32 <ncopa> or is it gtk3 only?
2013-05-23 07:16:47 <barthalion> I think it will work, I will test it locally first
2013-05-23 07:18:03 <barthalion> At least I haven't found anything relevant in changelog
2013-05-23 07:18:42 <barthalion> nor mailing lists
2013-05-23 07:20:39 <ncopa> the lis of aports that has perl-dev but is not a perl-*: http://sprunge.us/dDBH
2013-05-23 07:22:59 <barthalion> surely irssi requires rebuild, don't know about rest
2013-05-23 07:23:15 <ncopa> i think they all do
2013-05-23 07:23:29 <barthalion> Not sure about squid
2013-05-23 07:23:38 <ncopa> i think it does
2013-05-23 07:23:41 <ncopa> but ack does not
2013-05-23 07:23:53 <ncopa> i filter out the 'noarch's
2013-05-23 07:24:16 <barthalion> xscreensaver doesn't
2013-05-23 07:24:25 <ncopa> are you sure?
2013-05-23 07:24:36 <barthalion> yes
2013-05-23 07:24:40 <ncopa> why it has perl-dev in there then?
2013-05-23 07:24:56 <barthalion> It doesn't mean it require rebuild
2013-05-23 07:26:25 <ncopa> what does it use perl for?
2013-05-23 07:28:25 <ncopa> maybe should have 'perl' in makedepends instead of 'perl-dev'
2013-05-23 07:28:44 <barthalion> configure even doesn't check if perl is installed
2013-05-23 07:29:27 <barthalion> Or it does
2013-05-23 07:29:29 <barthalion> Hm
2013-05-23 07:29:36 <ncopa> it checks for perl
2013-05-23 07:29:49 <ncopa> but i think you might be right
2013-05-23 07:29:55 <ncopa> might not need perl-dev
2013-05-23 07:30:05 <barthalion> There are some perl scripts in sources
2013-05-23 07:30:43 <ncopa> i was thinking it was used for making perl plugins or something
2013-05-23 07:30:55 <ncopa> bit it seems like nothing is linking to libperl.so
2013-05-23 07:31:00 <barthalion> yup
2013-05-23 07:31:24 <ncopa> probably same with squid
2013-05-23 07:39:03 <ncopa> then we have all the stuff in testing :-/
2013-05-23 07:39:49 <kaniini> iirc the check it has depends on perl-dev
2013-05-23 07:45:53 <ncopa> kaniini: i just tried to build it with perl in makedepens instead of perl-dev, i think it works.
2013-05-23 07:46:02 <ncopa> they might have fixed it upstream
2013-05-23 07:47:09 <ncopa> we have no less than 206 perl-* aports in testing
2013-05-23 07:47:31 <kaniini> worksforme
2013-05-23 07:48:42 <ncopa> heh
2013-05-23 07:48:56 <ncopa> 7 of those perl-* also exists in main
2013-05-23 08:57:20 <ncopa> there are some perl modules that break with perl 5.18
2013-05-23 11:11:48 <vkrishn> why does http://en.wikipedia.org/wiki/Alpine_Linux  still have two callout boxes at the top
2013-05-23 11:12:34 <vkrishn> I think there are sufficient articles and couple of papers already there to add as citations...etc
2013-05-23 11:13:53 <kaniini> http://en.wikipedia.org/wiki/User:Kephir
2013-05-23 11:14:12 <kaniini> because that guy likely has asperger's syndrome (as basically all other wikipedians do)
2013-05-23 11:14:43 <kaniini> on a more serious note, i don't really see why we care about whether wikipedia thinks we're notable or not
2013-05-23 11:15:01 <kaniini> i think the fact that people are actually doing things with alpine is far more interesting than wikipedia's opinion on us
2013-05-23 11:15:06 <vkrishn> no its not about wikipedia in general
2013-05-23 11:15:50 <vkrishn> it has become more a defacto place to start to read something
2013-05-23 11:16:25 <kaniini> this is true, and if you feel it is a battle worth fighting, go for it
2013-05-23 11:16:37 <kaniini> however, you are dealing with a 'self-proclaimed deletionist'
2013-05-23 11:16:37 <clandmeter> vkrishn: fixtifxitfixitfixit
2013-05-23 11:16:39 <vkrishn> even if it does require we do make some effort to rectify it
2013-05-23 11:16:57 <kaniini> (which is code for 'aspergers syndrome' if i ever heard it to be honest)
2013-05-23 11:17:17 <kaniini> so, best of luck, don't die or something :P
2013-05-23 11:18:31 <kaniini> ``I am a deletionist (and immediatist at that), though I do give some benefit of doubt in favour of keeping articles about general, abstract subjects (as opposed to specific companies, people, software, etc.). I believe that articles better be deleted than stay unsourced, and that deletion should not be considered a prohibition from recreating the article, when sources are brought or circumstances
2013-05-23 11:18:32 <kaniini> change. I am also a moderately mergist (in that I believe it is better to have fewer articles unless there is a strong need for a split), mesopedian (though I sympathise with exopedians more), mediawikianist (as you can tell by my scripts), individualist sceptic.''
2013-05-23 11:18:36 <kaniini> that is from this guy's user page
2013-05-23 11:18:38 <kaniini> i mean
2013-05-23 11:18:43 <kaniini> if you wanna argue with this guy
2013-05-23 11:18:45 <kaniini> go for it
2013-05-23 11:18:50 <kaniini> but, it seems like a really bad time
2013-05-23 11:19:00 <vkrishn> those boxes just needs some reliable secondary sources, don't think that should be hard
2013-05-23 11:19:54 <vkrishn> atleast trying to know what is needed to get them removed, may help
2013-05-23 11:20:09 <kaniini> they want citations from books
2013-05-23 11:20:22 <kaniini> notable news sources
2013-05-23 11:20:33 <kaniini> academic journals (think JSTOR, ACM)
2013-05-23 11:20:45 <kaniini> i'm drawing blanks on this
2013-05-23 11:21:11 <hl> you know what would be amusing, if someone started requesting deletion of User: pages citing notability
2013-05-23 11:22:51 <kaniini> that guy's userpage is at the height of 'full of himself'
2013-05-23 11:23:18 <hl> i mean, wikipedia's obsession with deleting non-notable stuff is bizarre considering the crap they allow in meta namespaces
2013-05-23 11:23:29 <hl> https://en.wikipedia.org/wiki/Wikipedia:Userboxes/Colour
2013-05-23 11:24:00 <kaniini> hl: the problem is wikipedians are very much like Shiro
2013-05-23 11:24:09 <vkrishn> v2.6 is quit stable and feature rich, I think a review could be published easily
2013-05-23 11:24:22 <kaniini> vkrishn: a review from who, about what
2013-05-23 11:24:43 <vkrishn> on some news site
2013-05-23 11:24:47 <vkrishn> by some tech
2013-05-23 11:24:58 <kaniini> but why does anyone care
2013-05-23 11:25:18 <vkrishn> to grow userbase for alpine
2013-05-23 11:25:18 <kaniini> the big point of alpine is it's this space where technical work is being done
2013-05-23 11:25:39 <kaniini> okay, now answer this question: do we want the userbase that would use wikipedia to discover us?
2013-05-23 11:26:08 <kaniini> i am not convinced that type of user would be one that would get much out of alpine
2013-05-23 11:26:42 <kaniini> marketing is a more complex problem than just getting the word out
2013-05-23 11:26:50 <kaniini> it's getting the word out to the right people
2013-05-23 11:26:54 <vkrishn> its a good place to publish
2013-05-23 11:27:07 <vkrishn> they might have quirks
2013-05-23 11:27:16 <vkrishn> but still good,
2013-05-23 11:27:35 <kaniini> sure, but there's probably more valuable ways to grow alpine community than wikipedia and reviews from tech sites :P
2013-05-23 11:27:41 <vkrishn> their database is even being distributed in olpc projects
2013-05-23 11:28:14 <kaniini> anyway...
2013-05-23 11:28:25 <kaniini> one thing that might be usable here is the Xen blog about alpine
2013-05-23 11:28:45 <vkrishn> it should not be very hard to refine the page
2013-05-23 11:28:50 <kaniini> and, Cisco DM-VPN support in linux was written by us
2013-05-23 11:28:53 <vkrishn> yes
2013-05-23 11:29:01 <kaniini> so, there may actually be papers on that.
2013-05-23 11:29:06 <vkrishn> sounds good
2013-05-23 11:29:27 <kaniini> that was a while ago
2013-05-23 11:29:38 <vkrishn> then its should not be difficult to publish some white papers
2013-05-23 11:30:10 <vkrishn> they could even serve to documentation base
2013-05-23 11:30:32 <kaniini> self-published material is not allowed as wikipedia sources...
2013-05-23 11:30:36 <kaniini> http://blog.xen.org/index.php/2013/03/04/alpine-linux-and-xen/
2013-05-23 11:30:40 <kaniini> this may be usable
2013-05-23 11:30:46 <vkrishn> yes, I know
2013-05-23 11:31:15 <vkrishn> I had mentioned this and even put up a bugs.a.o for rss
2013-05-23 11:32:24 <vkrishn> they would help in building an env
2013-05-23 11:35:00 <vkrishn> after every release even a single review, 1 whitepaper, couple of blogs links should keep it going
2013-05-23 11:37:33 <vkrishn> I am not sure this is more of marketing, more like pushing facts more often
2013-05-23 11:38:05 <kaniini> well, there you have a simple problem
2013-05-23 11:38:20 <kaniini> none of us really care about this, as we are basically working on alpine for our own selfish reasons
2013-05-23 11:38:34 <kaniini> most work on alpine comes from people who specifically need a feature
2013-05-23 11:39:11 <kaniini> so, this means nobody really is interested in writing whitepapers
2013-05-23 11:40:01 <vkrishn> hmm
2013-05-23 11:40:10 <kaniini> for example, i have a large cluster of servers running alpine
2013-05-23 11:40:29 <kaniini> royger worked for a university which deployed alpine + xen on a supercomputing cluster
2013-05-23 11:41:13 <kaniini> alpine is a big player in all sorts of niche markets that few people care about
2013-05-23 11:41:50 <kaniini> scientific computing being one of them (not necessarily for running the scientific code, but for provisioning the compute resources)
2013-05-23 11:42:24 <kaniini> so, yes, there is probably papers on building clusters and things with alpine
2013-05-23 11:42:48 <kaniini> run from ram is a killer app when you are building something like a beowulf cluster or a cloud with xen
2013-05-23 11:43:01 <vkrishn> kaniini, you already gave some "Alpine in use" , they just need some webpage link or public published email link
2013-05-23 11:43:23 <vkrishn> to verify and the problem seems solved
2013-05-23 11:44:11 <kaniini> i am not sure what fabled is doing with alpine, but fabled, tdtrask and iilluzion seem to participate in operating a large enterprise WAN with alpine
2013-05-23 11:44:29 <vkrishn> cool
2013-05-23 11:44:57 <vkrishn> any mailing list, public discussion on it?
2013-05-23 11:45:11 <kaniini> there's some discussions about this stuff on alpine-devel over the years
2013-05-23 11:45:16 <kaniini> but
2013-05-23 11:45:26 <kaniini> the damn list archives don't generate reliable permanent urls
2013-05-23 11:45:28 <kaniini> :/
2013-05-23 11:45:30 <vkrishn> excluding alpinebase
2013-05-23 11:46:10 <kaniini> there's some on xen-users about building clusters with alpine
2013-05-23 11:46:18 <vkrishn> nice
2013-05-23 11:46:45 <kaniini> i mean, don't get me wrong
2013-05-23 11:46:52 <kaniini> it'd be cool if there were whitepapers and stuff
2013-05-23 11:47:21 <kaniini> but, i just don't see it happening because nobody either has time to do it, and if they have time to work on alpine, would rather work on technical things
2013-05-23 11:47:48 <vkrishn> nope, what you stated above are very good "usages" - they just need verify links
2013-05-23 11:48:36 <vkrishn> or atleast we now know what needs publishing
2013-05-23 11:52:39 <hl> hmm
2013-05-23 11:52:48 <hl> yes, alpine could do with a debian preseed-style system
2013-05-23 11:53:06 <hl> there are some bugs in the setup-alpine answer file system that mean it isn't fully non-interactive at this time
2013-05-23 11:54:35 <ncopa> i am actually pretty happy with the alpine linux user base
2013-05-23 11:54:51 <ncopa> i see bsd people
2013-05-23 11:55:06 <ncopa> people who likes alpine because "it gets out of the way"
2013-05-23 11:55:17 <kaniini> i see bsd people dreaming that BSD was doing things we are doing :P
2013-05-23 11:55:31 <ncopa> people who want build run from ram xen hosts
2013-05-23 11:55:33 <ncopa> etc
2013-05-23 11:55:51 <kaniini> i mean, debian and such are bigger
2013-05-23 11:56:15 <kaniini> but you can see real world examples of alpine running on everything from tiny computers to huge scientific clusters
2013-05-23 11:56:23 <ncopa> yes
2013-05-23 11:56:27 <kaniini> you can't say that with any other distro
2013-05-23 11:57:24 <kaniini> for example, in huge clusters, you have alpine, scientificlinux (which is rhel with hacks to make it less bloaty), and then proprietary stuff from oracle and cray
2013-05-23 11:58:24 <kaniini> and i think if alpine existed for as long as debian or redhat has, you would see things like the large hadron collider running on alpine instead of CERN's scientificlinux
2013-05-23 11:58:34 <kaniini> because CERN would never have bothered ;p
2013-05-23 11:58:44 <ncopa> :)
2013-05-23 11:59:08 <vkrishn> :)
2013-05-23 11:59:13 <ncopa> vkrishn: i think dmvpn is a good example of notable thing
2013-05-23 11:59:38 <ncopa> "first linux distro to support dmvpn"
2013-05-23 11:59:48 <vkrishn> so the rss request possible ?
2013-05-23 12:00:11 <vkrishn> good place to compile links
2013-05-23 12:00:11 <ncopa> rss request?
2013-05-23 12:00:15 <ncopa> ah
2013-05-23 12:00:18 <vkrishn> bugs.a.o
2013-05-23 12:00:24 <kaniini> point is...
2013-05-23 12:00:43 <kaniini> alpine has real-world deployments on everything from tiny little ALIX boards to TOP-500
2013-05-23 12:01:04 <kaniini> so that is a pretty diverse userbase
2013-05-23 12:02:12 <ncopa> vkrishn, what bug # was it?
2013-05-23 12:02:13 <kaniini> the ironic thing is that the people building TOP-500 installations want the same thing for their control plane that the people deploying on ALIX want
2013-05-23 12:02:22 <ncopa> lol
2013-05-23 12:02:23 <ncopa> yes
2013-05-23 12:02:59 <ncopa> i dont mind if somebody fix up the wikipedia article about al
2013-05-23 12:03:02 <ncopa> i think it woudl be nice
2013-05-23 12:03:11 <ncopa> but i dont think its worth my time ;)
2013-05-23 12:03:41 <ncopa> i think most of us feel similar
2013-05-23 12:03:57 <ncopa> we are more interested in hacking than tell the world how good our thing is
2013-05-23 12:04:01 <kaniini> that's kind of the problem alpine has though -- most of the things it is deployed for are boring
2013-05-23 12:04:18 <ncopa> yup, not fancy touch screen devices
2013-05-23 12:04:19 <kaniini> that doesn't mean it isn't an integral part of really neat things, but it is just a toolkit
2013-05-23 12:04:20 <ncopa> tablest
2013-05-23 12:04:50 <ncopa> what i would like to see is puppy ppl build a puppy variant based on alpine
2013-05-23 12:04:53 <ncopa> that would be cool
2013-05-23 12:05:07 <kaniini> for example in top-500, alpine is just used to provision redhat and debian environments either in container or as virtual machine
2013-05-23 12:05:22 <kaniini> people aren't doing actual science with alpine itself :P
2013-05-23 12:05:28 <ncopa> :)
2013-05-23 12:05:43 <kaniini> i mean, that'd be neat, but uclibc's math library is... not really up to the task of that ;)
2013-05-23 12:06:10 <ncopa> :)
2013-05-23 12:06:18 <kaniini> but that's fine, because we're really good at spawning those vserver containers and vpses
2013-05-23 12:06:33 <kaniini> and we're really good at routing, and filtering, and so on
2013-05-23 12:06:57 <ncopa> yes, i got a request from some ISP to enable multipath routing in quagga
2013-05-23 12:07:05 <ncopa> ISPs are using it
2013-05-23 12:10:15 <ncopa> vkrishn: re wikipedia article, maybe link to distrowatch's alpine page
2013-05-23 12:10:45 <ncopa> that xen blog is also a good source i think
2013-05-23 12:11:35 <ncopa> heh
2013-05-23 12:11:39 <ncopa> alpien is mentioned in a book
2013-05-23 12:11:46 <ncopa> http://books.google.no/books?id=hBlBYgEACAAJ&dq=%22Alpine+Linux%22&hl=no&sa=X&ei=TQeeUdXhOaKN4AS88ID4Dw&redir_esc=y
2013-05-23 12:12:28 <kaniini> hah
2013-05-23 12:12:37 <kaniini> a book that is basically a printed book of wikipedia articles
2013-05-23 12:12:43 <kaniini> for maximum irony, i assume ;)
2013-05-23 12:12:54 <ncopa> lol yes :)
2013-05-23 12:13:33 <ncopa> oh he has unreliable inet
2013-05-23 12:13:38 <kaniini> http://www.google.no/search?hl=en&tbo=p&tbm=bks&q=inauthor:%22Source+Wikipedia%22
2013-05-23 12:13:47 <kaniini> they have published lots of books
2013-05-23 12:13:57 <ncopa> indeed
2013-05-23 12:14:37 <ncopa> wb vkrishn :)
2013-05-23 12:15:35 <vkrishn>  ncopa, can the weblogs for a.o help
2013-05-23 12:18:40 <vkrishn> about distrowatch, not sure that would add weight, but still useful
2013-05-23 12:19:04 <vkrishn> for now compiling any external links is important, maybe filter them later
2013-05-23 12:23:31 <ncopa> vkrishn: thanks for working on it
2013-05-23 12:24:52 <vkrishn> I will do some checks on the links and see if they can be categorized
2013-05-23 12:25:01 <vkrishn> and maybe added to wiki.a.o
2013-05-23 12:25:29 <ncopa> vkrishn: what was the bug # for that rss request?
2013-05-23 12:25:48 <vkrishn> it should be in infrastructure category
2013-05-23 12:25:54 <vkrishn> just a moment
2013-05-23 12:25:54 <ncopa> found it
2013-05-23 12:25:56 <ncopa> https://bugs.alpinelinux.org/issues/1831
2013-05-23 12:25:57 <vkrishn> ok
2013-05-23 12:26:24 <ncopa> something like: http://planet.gentoo.org/ ?
2013-05-23 12:26:57 <vkrishn> yes
2013-05-23 12:27:11 <ncopa> and http://blog.xfce.org/
2013-05-23 12:28:09 <vkrishn> that is internal blog, that would be rss'd in combined rss published on planet
2013-05-23 12:29:00 <vkrishn> I think articles.xfce.org would be more generic, covering articles,whitepaper, blogs, survey... etc
2013-05-23 12:29:29 <ncopa> but it looks like an rss feed or something
2013-05-23 12:29:36 <ncopa> liks to other blogging sites
2013-05-23 12:30:04 <vkrishn> yes
2013-05-23 12:30:29 <vkrishn> even debian has planet.debian.org
2013-05-23 12:30:35 <kaniini> anyway, i think alpine has one of the most diverse userbases out there
2013-05-23 12:30:58 <vkrishn> :) soon to be published
2013-05-23 12:31:19 <kaniini> i can't think of any other distribution which has users ranging from tiny little singleboard x86 stuff to massive supercomputing installations
2013-05-23 12:31:29 <ncopa> :)
2013-05-23 12:31:39 <kaniini> i mean, *maybe* debian
2013-05-23 12:31:58 <kaniini> fedora's definitely out with systemd etc
2013-05-23 12:32:16 <kaniini> nobody wants systemd on big iron or a singleboard x86
2013-05-23 12:32:18 <kaniini> lol
2013-05-23 12:33:29 <kaniini> now, where we are not well represented in userbase is desktop
2013-05-23 12:34:06 <kaniini> but, i don't think we can make big gains there without compromising the features that make us interesting in server and embedded and big computing
2013-05-23 12:34:51 <kaniini> the people in the latter column simply want the most efficient compute environment possible so they have more resources available to their problem domain
2013-05-23 12:35:14 <kaniini> desktop users want udev and glibc for flash ;)
2013-05-23 12:35:55 <kaniini> though to be honest, i see alpine's desktop support mostly being there for the purpose of providing a development environment that has some creature comforts
2013-05-23 12:37:00 <ncopa> i use it on my desktop
2013-05-23 12:37:11 <ncopa> i eat my own dogfood
2013-05-23 12:37:23 <kaniini> well yes, but you also spend a lot of time maintaining and developing things ontop of alpine
2013-05-23 12:37:24 <kaniini> :P
2013-05-23 12:38:00 <ncopa> true :)
2013-05-23 12:39:42 <vkrishn> that when small nifty alpine based appliances come in
2013-05-23 12:42:18 <kaniini> well
2013-05-23 12:42:22 <kaniini> that's the point really
2013-05-23 12:42:59 <vkrishn> ncopa, weblogs have referer links
2013-05-23 12:43:20 <vkrishn> they are good starting point
2013-05-23 12:46:25 <kaniini> some kid installs alpine on a vm on youtube: http://www.youtube.com/watch?v=HJ_bDgV1aek
2013-05-23 12:47:40 <kaniini> "man, the installer is like OpenBSD"
2013-05-23 12:48:24 <hl> what
2013-05-23 12:49:07 <clandmeter> kaniini: you got a funny voice :p
2013-05-23 12:49:14 <kaniini> it is not my voice
2013-05-23 12:49:28 <kaniini> i do not think you have heard my voice
2013-05-23 12:49:52 <kaniini> wtf
2013-05-23 12:49:56 <kaniini> he doesn't actually install it
2013-05-23 12:50:08 <ncopa> http://www.youtube.com/watch?v=uSKbJPnQkjI
2013-05-23 12:50:52 <hl> #tuxhat on irc.freenode.net
2013-05-23 12:50:54 <hl> hmmmmmmm
2013-05-23 12:51:27 <kaniini> the distro is from "norway"
2013-05-23 12:51:47 <kaniini> oh, okay.
2013-05-23 12:51:51 <kaniini> so he does actually install it
2013-05-23 12:51:58 <ncopa> someone commented that youtube review: "alpine linux is so easy so you can even install it when you are high.."
2013-05-23 12:52:08 <kaniini> he is definitely high
2013-05-23 12:52:39 <clandmeter> kaniini: i was just kidding :)
2013-05-23 12:53:30 <kaniini> i think i could contain my composure better while stoned tbh
2013-05-23 12:56:21 <clandmeter> can we add these video's on wikipedia? :)
2013-05-23 12:57:51 <kaniini> ...do we want to?
2013-05-23 12:58:07 <kaniini> i mean, do we really want this person representing us? ;)
2013-05-23 12:58:09 <vkrishn> if the video link is on some distinct blog page maybe
2013-05-23 12:58:23 <vkrishn> making is verifyable
2013-05-23 12:58:34 <hl> [tuxhat] (~CYBERPUNX@unaffiliated/tuxhat): Live FreeBSD or Die!
2013-05-23 12:59:21 <kaniini> he mentions that ncopa is from norway like 300 times on this video as if it is some sort of relevant data
2013-05-23 13:00:08 <clandmeter> he is?
2013-05-23 13:00:17 <kaniini> yes, he is!
2013-05-23 13:00:27 <kaniini> hince the .no on his irc connection :P
2013-05-23 13:00:46 <kaniini> clandmeter: apparently we are all from norway
2013-05-23 13:00:47 <clandmeter> if only i had known that earlier!
2013-05-23 13:00:53 <hl> i never knew being norwegian was so fashionable.
2013-05-23 13:01:02 <kaniini> and in order to be an alpine contributor you must be norwegian
2013-05-23 13:01:09 <hl> 
ACTION looks for a norwegian bouncer provider.

2013-05-23 13:01:29 <kaniini> also, i hear an occasional use of a vaporiser on this install video
2013-05-23 13:01:38 <clandmeter> 
ACTION installs irssi on git.a.o

2013-05-23 13:02:19 <kaniini> "huh, why did vmware crash"
2013-05-23 13:02:47 <kaniini> hahahahahahahahahahah
2013-05-23 13:02:55 <kaniini> clandmeter: 10 minutes in, he googles shit in the wiki
2013-05-23 13:03:45 <jbilyk> "* clandmeter installs irssi on git.a.o" - clandmeter why do we need it?
2013-05-23 13:03:46 <clandmeter> maybe he is watching us in the channel right now. stay polite! :)
2013-05-23 13:04:21 <clandmeter> jbilyk: read up :)
2013-05-23 13:05:13 <jbilyk> clandmeter: i know what it does lol, but i thought we already had a bot from git.a.o in this channel
2013-05-23 13:05:33 <kaniini> clearly we need to install BitchX on git.alpinelinux.org
2013-05-23 13:05:39 <kaniini> it is the most fashionable irc client
2013-05-23 13:06:13 <clandmeter> jbilyk: we need to be fashionalble, so we need to pretent to come from .no
2013-05-23 13:06:13 <ncopa> kaniini, lol, and it only me that are in (not even from) norway
2013-05-23 13:06:54 <kaniini> i think he just got that from distrowatch
2013-05-23 13:07:54 <ncopa> cannot help it but i like this guy :)
2013-05-23 13:08:00 <ncopa> he is fun :)
2013-05-23 13:08:13 <kaniini> it reminds me of my youth
2013-05-23 13:08:21 <kaniini> not the UNIX errors, but the bong hits.
2013-05-23 13:08:46 <ncopa> i also think its interesting to se what he struggles with
2013-05-23 13:09:08 <kaniini> yes, it is a fascinating usability study
2013-05-23 13:09:14 <ncopa> we should probably add the stuff to bugs.a.o
2013-05-23 13:09:38 <kaniini> "can you install alpine with an X environment while wasted"
2013-05-23 13:11:28 <kaniini> oh dear
2013-05-23 13:11:32 <kaniini> he ran setup-alpine again
2013-05-23 13:11:46 <kaniini> perhaps we should make apk-tools be more helpful
2013-05-23 13:12:17 <kaniini> "HINT: No available repositories, add a repository URI to /etc/apk/repositories and run apk update to continue."
2013-05-23 13:12:40 <ncopa> how many minutes in the video?
2013-05-23 13:13:33 <kaniini> 17 minutes in
2013-05-23 13:14:36 <vkrishn> and: Available repository can be add later by running setup-apkrepos
2013-05-23 13:14:58 <ncopa> apk add update
2013-05-23 13:15:45 <kaniini> vkrishn: we don't want to necessarily do that, as it is possible that other distributions use apk-tools either now or in the future
2013-05-23 13:16:32 <ncopa> i think he gets confused by the error message from : apk add update
2013-05-23 13:16:45 <ncopa> ERROR: 1 unsatisfiable dependencies:
2013-05-23 13:16:45 <ncopa>   world: update
2013-05-23 13:17:11 <kaniini> yes, it is vague
2013-05-23 13:17:23 <ncopa> should be something like: "sorry but 'update' is not a valid package"
2013-05-23 13:17:35 <kaniini> 1: most people don't realize that 'world' is a set of rules that put the system together
2013-05-23 13:18:18 <ncopa> i wonder if we should add a setup-xfce-desktop script
2013-05-23 13:18:23 <ncopa> that pulls in what you need
2013-05-23 13:18:32 <ncopa> sets up udev
2013-05-23 13:18:33 <ncopa> etc
2013-05-23 13:18:34 <kaniini> yes, probably a good way
2013-05-23 13:18:58 <ncopa> i wonder how to detect what video driver to use, without install them all
2013-05-23 13:19:12 <ncopa> or if all are installed,
2013-05-23 13:19:21 <kaniini> lspci
2013-05-23 13:19:24 <ncopa> how do we detect which one is in use, so we can uninstall the rest
2013-05-23 13:19:41 <kaniini> but with kms, we could just look at the list of loaded kernel modules
2013-05-23 13:19:48 <ncopa> ah yes
2013-05-23 13:20:03 <kaniini> haha
2013-05-23 13:20:05 <kaniini> gnome
2013-05-23 13:20:07 <kaniini> gnome2
2013-05-23 13:20:13 <kaniini> oh, it's gnome-desktop-environment
2013-05-23 13:20:17 <kaniini> 
ACTION ducks

2013-05-23 13:20:22 <ncopa> lol
2013-05-23 13:20:36 <ncopa> gnome is ...
2013-05-23 13:20:38 <ncopa> hmm
2013-05-23 13:20:39 <ncopa> sad
2013-05-23 13:20:43 <ncopa> needs systemd
2013-05-23 13:20:54 <hl> ...gnome needs systemd now?
2013-05-23 13:20:54 <ncopa> i asked on the gnome3 chan
2013-05-23 13:21:07 <ncopa> it need components
2013-05-23 13:21:11 <ncopa> logind
2013-05-23 13:21:31 <kaniini> weren't the gnome guys just making their own distro now?
2013-05-23 13:21:46 <ncopa> yes
2013-05-23 13:21:48 <ncopa> gnome os
2013-05-23 13:22:06 <kaniini> with some funky git-based package manager since you're just getting a system image
2013-05-23 13:22:16 <ncopa> lol
2013-05-23 13:22:18 <ncopa> bah
2013-05-23 13:22:39 <ncopa> they clearly dont *want* gnome3 running on things like alpine linux
2013-05-23 13:22:40 <kaniini> like, literally, i read something about them using git to distribute binaries
2013-05-23 13:23:09 <ncopa> if a hammer is the only tool you have, every problem looks like a nail...
2013-05-23 13:23:28 <ncopa> someone is probably proud of have learned git
2013-05-23 13:23:50 <ncopa> i love git but i dont think i'd use it for distributing binaries...
2013-05-23 13:25:18 <clandmeter> that guy in the video is really confused...
2013-05-23 13:25:39 <vkrishn> are they using it for system rollback mechanism ?
2013-05-23 13:30:33 <kaniini> clandmeter: i think he is impared
2013-05-23 13:31:51 <kaniini> clearly we need to do CIA-like usability studies involving giving potential users LSD, waiting an hour, and seeing if they can use alpine
2013-05-23 13:35:07 <royger> re the alpine usage discussion, I think that we are missing presence at Linux/OSS events
2013-05-23 13:35:43 <royger> that's really important in order to get attention, and most big distros do it (suse, fedora, debian...)
2013-05-23 13:36:35 <kaniini> yes, i think we have a lot that we can say about alpine that would get attention of 'enterprises'
2013-05-23 13:37:14 <vkrishn> ncopa could do a small talk show, podcast
2013-05-23 13:37:25 <kaniini> that sounds amusing
2013-05-23 13:38:18 <clandmeter> in which language?
2013-05-23 13:38:39 <royger> having a small stand at LinuxCon and giving/selling some t-shirts could be a good start point
2013-05-23 13:38:52 <kaniini> "it's as usable as openbsd"
2013-05-23 13:39:06 <kaniini> "it brings the unix philosophy into the system as you can see"
2013-05-23 13:39:15 <kaniini> 
ACTION falls over laughing

2013-05-23 13:39:24 <royger> it's amusing how many ppl you can attract by giving away crappy t-shirts
2013-05-23 13:39:44 <royger> kaniini: at least it's better than saying "it has as much features as openbsd"
2013-05-23 13:40:07 <vkrishn> give free alpine stickers/other stuffs for those who buy alpine cd/usb
2013-05-23 13:40:13 <kaniini> royger: oh, i am talking about this guy who got wasted and tried to install alpine
2013-05-23 13:40:22 <kaniini> royger: and made a youtube about it
2013-05-23 13:40:38 <royger> I've missed that hahaha
2013-05-23 13:41:29 <kaniini> way i see it is -- suse has t-shirts, we have top-500 installs
2013-05-23 13:41:44 <kaniini> well
2013-05-23 13:41:56 <kaniini> technically so does suse, but it is some modified version done by cray
2013-05-23 13:42:57 <kaniini> i think there are things we can do to improve usability though
2013-05-23 13:43:15 <kaniini> Unsatisfied dependencies world: update
2013-05-23 13:43:18 <kaniini> is not very good
2013-05-23 13:47:03 <ncopa> i could do podcast in spanish
2013-05-23 13:47:07 <Mp5shooter> <<@kaniini> clearly we need to do CIA-like usability studies involving giving potential users LSD, waiting an hour, and seeing if they can use alpine
2013-05-23 13:47:07 <ncopa> or portugues
2013-05-23 13:47:11 <Mp5shooter> i'd like to participate
2013-05-23 13:47:27 <ncopa> or swedish (like the muppets chef)
2013-05-23 13:47:40 <ncopa> or norwegian if someone pays me well enough :)
2013-05-23 13:47:41 <fabled> kaniini, is your nick chosen intentionally based on Finnish language?
2013-05-23 13:48:45 <kaniini> Mp5shooter: i'm not buying
2013-05-23 13:48:56 <kaniini> (at least not for you)
2013-05-23 13:48:57 <Mp5shooter> :p
2013-05-23 13:49:01 <vkrishn> any language, but a transcript in english needs to be published later
2013-05-23 13:50:03 <kaniini> fabled: kanin is a rootword that means 'rabbit' in almost every indo-germanic language, it's kind of a hack ;)
2013-05-23 13:51:09 <fabled> 'kaniini' is a perfectly valid Finnish word, roughly meaning "bunny-rabbit" ;)
2013-05-23 13:51:10 <clandmeter> kaniini: you watch the matrix too many times
2013-05-23 13:52:43 <Mp5shooter> only $99 for 1mg kaniini
2013-05-23 13:52:43 <Mp5shooter> https://www.sigmaaldrich.com/catalog/product/sigma/l7007?lang=en&region=US
2013-05-23 13:53:04 <kaniini> fabled: yes, also swedish, norwegian, icelandic, danish, german, etcetera :)
2013-05-23 13:54:46 <Mp5shooter> kaniini: do you have an LSD license
2013-05-23 13:54:46 <Mp5shooter> :p
2013-05-23 13:55:14 <kaniini> i do not think that you can buy LSD off of a website
2013-05-23 13:55:20 <kaniini> well
2013-05-23 13:55:26 <kaniini> off of that website
2013-05-23 13:55:33 <Mp5shooter> apparently that's a legit site that chemists buy chemicals from
2013-05-23 13:55:37 <Mp5shooter> but
2013-05-23 13:55:45 <Mp5shooter> what kind of chemist would you have to be to be able to buy that
2013-05-23 13:55:45 <Mp5shooter> lo
2013-05-23 13:55:46 <kaniini> i mean, maybe 'silk road', but not that one :p
2013-05-23 13:57:11 <kaniini> fabled: but yes, i did like that -ini is a valid suffix in finnish
2013-05-23 13:57:24 <ncopa> no souch thing in swe or nor
2013-05-23 13:57:28 <ncopa> "kanin"
2013-05-23 13:59:06 <ncopa> i think it works on windows 3.11 and systemd too: kani.ini
2013-05-23 14:01:31 <Mp5shooter> lol
2013-05-23 14:01:52 <kaniini> kanin is rabbit in bokmal, wikitionary says so
2013-05-23 14:01:57 <kaniini> it couldn't be wrong
2013-05-23 14:02:17 <kaniini> http://en.wiktionary.org/wiki/kanin#Norwegian
2013-05-23 14:02:57 <kaniini> and... hey
2013-05-23 14:03:04 <kaniini> that means i can be optimized for parallel booting
2013-05-23 14:06:52 <kaniini> back in bit
2013-05-23 16:19:39 <StarWarsFan> What do you think regarding the mail with the syslog-ng package some days ago?
2013-05-23 16:22:17 <StarWarsFan> Ups, in fact it was yesterday on the mailinglist
2013-05-23 17:48:13 <barthalion> I'm going to pull it
2013-05-23 17:53:24 <barthalion> In fact I was interested in writing it on my own, but if someone did the job already…
2013-05-23 19:09:14 <StarWarsFan> re
2013-05-23 19:09:31 <StarWarsFan> barthalion: ah nice :-)
2013-05-23 21:45:49 <kaniini> Oh.
2013-05-23 21:45:56 <kaniini> hl: i talked to this tuxhat person
2013-05-23 21:46:03 <kaniini> hl: it turns out the problem is that he is quebecois.
2013-05-23 21:46:07 <kaniini> 
ACTION just nods.

2013-05-24 03:05:09 <kaniini> this guy is comical
2013-05-24 03:05:13 <kaniini> he listens to infowars
2013-05-24 03:05:16 <kaniini> runs freebsd
2013-05-24 03:05:31 <kaniini> and thinks X is actually good
2013-05-24 06:48:14 <ncopa> speaking of xorg
2013-05-24 06:48:16 <ncopa> http://www.phoronix.com/scan.php?page=news_item&px=MTM3ODA
2013-05-24 06:48:34 <ncopa> "
2013-05-24 06:48:34 <ncopa> While not all of the details were shared publicly, Ilja van Sprundel did recommend using dietlibc or uClibc over glibc, which he found to be "super bloated" yet the default for most Linux distributions."
2013-05-24 06:56:32 <ncopa> ugh
2013-05-24 06:56:43 <ncopa> all those security issues with xorg
2013-05-24 06:56:49 <ncopa> no releases yet
2013-05-24 07:31:53 <kaniini> X windows is the hallmark of inefficency and bad design
2013-05-24 07:32:00 <kaniini> but, we already know this ;)
2013-05-24 07:32:39 <kaniini> with the vesa-dri2-kms dri2d backend, i think we could ship wayland and run X rootless.
2013-05-24 07:32:43 <kaniini> i should work on it.
2013-05-24 07:33:00 <kaniini> i packaged kmscon
2013-05-24 07:42:59 <kaniini> i give up talking to this person
2013-05-24 07:43:11 <kaniini> he does not understand why 2 million X api calls to draw a window
2013-05-24 07:43:13 <kaniini> is a bad thing
2013-05-24 07:43:20 <kaniini> BUT BUT BUT it's UNIX
2013-05-24 07:51:40 <kaniini> incidentally, that reminds me
2013-05-24 07:52:40 <kaniini> there's a planned meeting of austin group about POSIX 2008-TCGr2
2013-05-24 07:52:52 <kaniini> any concerns anyone want raised from alpine direction ?
2013-05-24 07:55:33 <ncopa> hm
2013-05-24 07:55:35 <ncopa> i dunno
2013-05-24 07:55:38 <ncopa> what do you think?
2013-05-24 07:57:06 <ncopa> btw
2013-05-24 07:57:14 <ncopa> does anybody have a x11 with alpine v2.6?
2013-05-24 07:57:18 <kaniini> yes
2013-05-24 07:57:25 <kaniini> if by 2.6 you mean edge
2013-05-24 07:57:29 <ncopa> no
2013-05-24 07:57:33 <ncopa> i mean v2.6 repo
2013-05-24 07:57:38 <ncopa> i use edge myself
2013-05-24 07:57:49 <ncopa> for edge i pushed the new 1.5.99 release
2013-05-24 07:58:02 <ncopa> but for v2.6 i backported the cve patches
2013-05-24 07:58:28 <ncopa> with a for i in $(git rev-list ....); git cherry-pick -x $i || break;
2013-05-24 07:58:41 <kaniini> oh
2013-05-24 07:58:47 <kaniini> so, you want to see if it still works? :P
2013-05-24 07:58:50 <ncopa> http://cgit.freedesktop.org/xorg/lib/libX11/log/
2013-05-24 07:58:50 <ncopa> yes
2013-05-24 07:58:52 <ncopa> :)
2013-05-24 07:59:12 <ncopa> i picked everything from 2013-5-10 and later basically
2013-05-24 07:59:20 <kaniini> BSD enthusiasts are the worst
2013-05-24 07:59:29 <kaniini> it's like all logic and reasoning just goes out the window
2013-05-24 07:59:38 <kaniini> also, all OSes they like are BSD derived
2013-05-24 07:59:43 <ncopa> :)
2013-05-24 07:59:55 <kaniini> even if they are not
2013-05-24 08:00:07 <ncopa> i must admit that i do have deep respect for bsd
2013-05-24 08:00:26 <ncopa> specially openbsd
2013-05-24 08:01:05 <ncopa> i'm gonna push the libx11 patch
2013-05-24 08:01:30 <ncopa> would be nice if somebody wants to test it
2013-05-24 08:03:02 <kaniini> well, yes
2013-05-24 08:10:24 <ncopa> meh
2013-05-24 08:10:26 <ncopa> patch surgery
2013-05-24 08:11:18 <ncopa> incredibly how happy a single add whitespace can make gnu patch
2013-05-24 08:24:25 <kaniini>  cbnmmmmmmmmmmmmmmmmmmmmmmmmmmmmm~
2013-05-24 09:54:05 <kaniini> oops
2013-05-24 09:54:07 <kaniini> stupid cat
2013-05-24 09:56:19 <ncopa> finally
2013-05-24 09:57:34 <kaniini> X is the epitome of secure coding.
2013-05-24 10:03:19 <ncopa> interesting it broke on x86_64
2013-05-24 10:03:40 <ncopa> XpNotifyPdm.c:234:10: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
2013-05-24 10:03:43 <ncopa> smells fishy
2013-05-24 10:03:49 <ncopa> speaking of fish
2013-05-24 10:03:51 <ncopa> its lunch
2013-05-24 13:59:31 <ncopa> sigh
2013-05-24 13:59:39 <ncopa> only v2.4 and v2.3 to go now...
2013-05-24 14:01:41 <clandmeter> ncopa: i need to upgrade an older server (around 2-3 years). but it has edge repo, so i dont know which version its running
2013-05-24 14:02:03 <ncopa> check in /etc/allpine-release
2013-05-24 14:03:11 <ncopa> i'd recomend try upgrade to v2.6
2013-05-24 14:03:31 <ncopa> depending a bit on how old it is
2013-05-24 14:05:46 <clandmeter> ncopa: i need 2.4
2013-05-24 14:06:14 <clandmeter> i need to use scst
2013-05-24 14:06:18 <ncopa> ah ok
2013-05-24 14:06:35 <ncopa> what do you need from edge?
2013-05-24 14:06:35 <clandmeter> 2.4 is the only release i have tested (and most stable i think)
2013-05-24 14:07:02 <clandmeter> what do you mean what i need from edge?
2013-05-24 14:07:11 <ncopa> why you have edge repo
2013-05-24 14:07:24 <clandmeter> oh, i don't remember.
2013-05-24 14:07:30 <ncopa> if there is some package that does not exist in v2.4, then we can backport it
2013-05-24 14:07:41 <ncopa> i dont recommend you do run v2.4 + edge
2013-05-24 14:07:47 <ncopa> or anything pre v2.5
2013-05-24 14:08:07 <clandmeter> no i want just 2.4
2013-05-24 14:08:24 <clandmeter> there isnt much special in that setup
2013-05-24 14:08:43 <clandmeter> only using scst lvm raid combo
2013-05-24 14:09:01 <clandmeter> i remember it was short after we did x86_64
2013-05-24 14:09:11 <ncopa> so v2.2
2013-05-24 14:09:18 <clandmeter> maybe there was no release at that time
2013-05-24 14:09:34 <ncopa> iirc it should be fairly easy to upgrade from v2.2 to v2.4
2013-05-24 14:09:38 <clandmeter> 2.1.0_git20101221
2013-05-24 14:09:51 <ncopa> yeah.. you should upgrade :)
2013-05-24 14:10:15 <clandmeter> i should, but its a san
2013-05-24 14:10:24 <ncopa> does it run from ram or disk?
2013-05-24 14:10:27 <clandmeter> my esxi boxes run on top of it
2013-05-24 14:10:36 <clandmeter> ram
2013-05-24 14:10:43 <ncopa> boot from usb?
2013-05-24 14:10:47 <clandmeter> yes
2013-05-24 14:10:51 <ncopa> super
2013-05-24 14:11:00 <ncopa> i would use a new usb
2013-05-24 14:11:16 <ncopa> make a new v2.4 bootable
2013-05-24 14:11:26 <ncopa> copy over the apkovl
2013-05-24 14:11:36 <ncopa> test boot it on a box without any network connected
2013-05-24 14:11:39 <clandmeter> copy the config and run it on anther pc and see if it runs ok?
2013-05-24 14:11:49 <ncopa> (just to make sure the syslinux did its job)
2013-05-24 14:11:52 <ncopa> yes
2013-05-24 14:12:16 <ncopa> and finally, poweroff, replace usb and boot up
2013-05-24 14:12:30 <ncopa> i think it should "just work"
2013-05-24 14:12:42 <ncopa> hm
2013-05-24 14:12:44 <clandmeter> no need to upgrade?
2013-05-24 14:12:58 <ncopa> hm
2013-05-24 14:12:59 <clandmeter> i guess thats only on running sysytem
2013-05-24 14:13:08 <clandmeter> world doesnt have version info right?
2013-05-24 14:13:14 <ncopa> correct
2013-05-24 14:13:33 <ncopa> it should work unless you depend on things from network and apk cache
2013-05-24 14:13:58 <clandmeter> ill clean it to make sure i dont need anything
2013-05-24 14:14:12 <clandmeter> but all needed things are in scst iso
2013-05-24 14:14:48 <clandmeter> maybe i will give this a try on sunday.
2013-05-24 14:15:37 <clandmeter> i hope it solves some of my esxi issues.
2013-05-24 14:15:46 <ncopa> hm
2013-05-24 14:15:54 <ncopa> there was soemthing
2013-05-24 14:16:08 <ncopa> keep a backup of the old usb incase you get more issues
2013-05-24 14:16:31 <ncopa> http://bugs.alpinelinux.org/issues/645
2013-05-24 14:16:38 <ncopa> but its only when running it inesxi
2013-05-24 14:16:42 <clandmeter> already did a backup
2013-05-24 14:16:49 <ncopa> good
2013-05-24 14:16:52 <ncopa> i hope it works out
2013-05-24 14:17:31 <clandmeter> i have one that is running 2.4
2013-05-24 14:25:08 <clandmeter> its 2.3.5 to be exact, but i think that should be an issue to upgrade to 2.4
2013-05-24 14:25:17 <clandmeter> should not
2013-05-24 14:45:55 <build-2-4> files from v2.4.11-13-g0df792b uploaded
2013-05-24 14:49:58 <build64-2-4> files from v2.4.11-19-gf7aaccf uploaded
2013-05-24 14:50:12 <build-2-4> files from v2.4.11-19-gf7aaccf uploaded
2013-05-24 15:03:50 <build64-2-3> files from v2.3.6-213-gbb82712 uploaded
2013-05-24 15:20:13 <freedomrun> cen someone help?
2013-05-24 15:20:27 <freedomrun> *can
2013-05-24 15:39:09 <clandmeter> freedomrun: ^
2013-05-24 15:39:41 <freedomrun> clandmeter, thank you thank you thank you :):):):):)
2013-05-24 15:51:11 <clandmeter> ncopa: what does build-2-6 still react here?
2013-05-24 17:03:57 <alpinenoob1> Hi, I am trying to compile http://ezix.org/project/wiki/HardwareLiSter#Download lshw for alpine linux, and keep getting "undefined reference to `libintl_gettext'" error. I have all the apks I could think of installed, but no luck. Anybody else had issues with gettext libs?
2013-05-24 17:32:49 <barthalion> Well
2013-05-24 17:33:02 <barthalion> Have you tried installing gettext-dev?
2013-05-24 17:33:20 <barthalion> It should be good start
2013-05-24 17:33:23 <alpinenoob1> yes i did
2013-05-24 17:34:31 <alpinenoob1> did not help!
2013-05-24 17:36:37 <barthalion> I'm going to get some rest soon, so I'll take a look tommorow
2013-05-24 17:36:53 <barthalion> https://www.gnu.org/software/gettext/FAQ.html#integrating_undefined maybe here you will find something relevant
2013-05-24 17:37:27 <barthalion> Actually, if you don't find a solution, please leave me a message, I'll try to package it and push to testing
2013-05-24 17:37:55 <alpinenoob1> how do i leave you a message?
2013-05-24 17:38:10 <alpinenoob1> i was following those directions, no luck.
2013-05-24 17:38:34 <barthalion> Just write "barthalion" somewhere in message
2013-05-24 17:38:41 <alpinenoob1> oh ok
2013-05-24 17:39:01 <alpinenoob1> thank you for your help, would really appreciate if you could help package this up!
2013-05-24 17:39:22 <alpinenoob1> i am on day 2 of this, tried g++ and c++ compilers, no luck
2013-05-24 17:40:00 <freedomrun> can anyone tell me is this some kind of incopatibility : /usr/bin/Xorg: symbol 'shadowSetup': can't resolve symbol in lib '/usr/lib/xorg/modules/drivers/modesetting_drv.so'.
2013-05-24 17:40:22 <barthalion> freedomrun: Seems so. Does it occur after today upgrade of xorg packages in edge?
2013-05-24 17:41:22 <freedomrun> no from the start barthalion .. I know sis is sucky gfx but even using vesa .. I tried to generate xorg.conf with puppy linux and there it works
2013-05-24 17:41:36 <freedomrun> but same one here don`t
2013-05-24 17:41:58 <barthalion> You should wait for ncopa, I don't run Xorg here
2013-05-24 17:42:06 <barthalion> Or just report a bug
2013-05-24 17:43:33 <freedomrun> barthalion, clandmeter made missing modesettings package and I pull it from testing repo will that be ok for buf report or should I downgrade infact del this package as it was missing before
2013-05-24 17:43:34 <freedomrun> ??
2013-05-24 17:43:48 <freedomrun> *bug report
2013-05-24 17:44:18 <alpinenoob1> barthalion, someone else had a similar issue with gettext, http://www.spinics.net/lists/vdr/msg27690.html, I tried to solve it the same way and no luck. I just need lshw packaged.
2013-05-24 17:45:22 <barthalion> freedomrun: make a bug report
2013-05-24 17:45:28 <freedomrun> ok
2013-05-24 17:45:49 <barthalion> freedomrun: it can be closed anyway and you will get better explanation as i can't say anything helpful about xorg in alpine
2013-05-24 17:46:09 <barthalion> alpinenoob1: sure, i'll make a first try now, without packaging yet, just compiling
2013-05-24 17:46:18 <freedomrun> no worries barthalion and thanks
2013-05-24 17:49:58 <barthalion> alpinenoob1: it looks like missing uclibc stuff
2013-05-24 17:50:05 <barthalion> alpinenoob1: also known as "glibc sucks"
2013-05-24 17:50:28 <barthalion> ncopa, kaniini: more undefined references to `libintl_gettext' follow
2013-05-24 17:50:35 <barthalion> ncopa, kaniini: am i right?
2013-05-24 17:52:21 <alpinenoob1> barthalion, i had uclibc* packacges added, dev and util, is tehre more?
2013-05-24 17:52:50 <jbilyk> "<barthalion> ncopa, kaniini: more undefined references to `libintl_gettext' follow"...I just had same issue when trying again to setup cross-compile toolchain for ARM
2013-05-24 17:53:33 <barthalion> alpinenoob1: it won't help, it's a problem with code
2013-05-24 17:53:51 <barthalion> alpinenoob1: i'll try to fix it tommorow, i'm too tired to think
2013-05-24 17:54:06 <alpinenoob1> ok, would really appreciate that! thank you very much!
2013-05-24 17:54:16 <barthalion> alpinenoob1: and because i'm not programmer… fixing stuff like that require much from me :p
2013-05-24 17:55:48 <alpinenoob1> heh, sounds like you know what you are doing!
2013-05-24 17:57:17 <barthalion> i randomly hit my keyboard and then check if it made something working
2013-05-24 17:57:25 <barthalion> :p
2013-05-24 17:57:45 <alpinenoob1> at this point, sounds like it will be more productive than me randomly installing packages
2013-05-24 17:58:01 <alpinenoob1> is that issue documented anywhere?
2013-05-24 17:58:18 <barthalion> i don't think so
2013-05-24 17:58:29 <barthalion> you can probably find similiar reports in other applications
2013-05-24 17:59:07 <alpinenoob1> yes, seen similar error in other apps, but never explained what it is
2013-05-24 17:59:47 <barthalion> badly written code based upon glibc-only feature
2013-05-24 18:01:25 <alpinenoob1> unappreciated feature strikes again.
2013-05-24 19:10:51 <kaniini> alpinenoob1: it is due to underlinking
2013-05-24 19:11:09 <kaniini> alpinenoob1: export LIBS="-lintl" ./configure ... should get you on track
2013-05-24 19:12:21 <alpinenoob1> cool, will try that in a bit
2013-05-24 19:18:22 <alpinenoob1> kaniini, what is this ./configure file you speak off? is this something that should be in the project i am trying to compile, or … ?
2013-05-24 19:18:52 <kaniini> yes.
2013-05-24 19:19:01 <kaniini> if it doesn't have configure
2013-05-24 19:19:06 <kaniini> then you have to hack the build system
2013-05-24 19:19:15 <alpinenoob1> it just has Makefiles
2013-05-24 19:19:41 <kaniini> yeah...
2013-05-24 19:19:44 <kaniini> try uhm
2013-05-24 19:19:48 <kaniini> make LIBS="-lintl"
2013-05-24 19:19:54 <kaniini> if it works, cool
2013-05-24 19:20:03 <kaniini> if it explodes violently, you have to edit some makefiles
2013-05-24 19:20:05 <kaniini> :/
2013-05-24 19:21:29 <alpinenoob1> got a new error
2013-05-24 19:21:52 <alpinenoob1> undefined reference to `getpackageversion'
2013-05-24 19:21:59 <kaniini> try...
2013-05-24 19:22:04 <kaniini> make LIBS+="-lintl"
2013-05-24 19:22:43 <alpinenoob1> same error again
2013-05-24 19:22:47 <kaniini> hmm
2013-05-24 19:23:16 <kaniini> seems to be a symbol in lshw
2013-05-24 19:23:19 <kaniini> try grepping for it
2013-05-24 19:23:31 <kaniini> you'll want to install real grep
2013-05-24 19:23:36 <kaniini> busybox grep is kinda... dumb...
2013-05-24 19:24:04 <alpinenoob1> ok, will do
2013-05-24 19:25:26 <alpinenoob1> yeah, it is used. here are included in one of the sources:
2013-05-24 19:25:36 <alpinenoob1> #include "hw.h"
2013-05-24 19:25:36 <alpinenoob1> #include "print.h"
2013-05-24 19:25:36 <alpinenoob1> #include "main.h"
2013-05-24 19:25:36 <alpinenoob1> #include "version.h"
2013-05-24 19:25:36 <alpinenoob1> #include "options.h"
2013-05-24 19:25:36 <alpinenoob1> #include "osutils.h"
2013-05-24 19:25:37 <alpinenoob1> #include "config.h"
2013-05-24 19:25:37 <alpinenoob1> #include <unistd.h>
2013-05-24 19:25:39 <alpinenoob1> #include <stdio.h>
2013-05-24 19:25:39 <alpinenoob1> #include <string.h>
2013-05-24 19:25:40 <alpinenoob1> #include <stdlib.h>
2013-05-24 19:25:40 <alpinenoob1> #include <iostream>
2013-05-24 19:25:40 <alpinenoob1> #ifndef NONLS
2013-05-24 19:25:42 <alpinenoob1> #include <locale.h>
2013-05-24 19:25:51 <barthalion> stahp
2013-05-24 19:26:01 <alpinenoob1> ?
2013-05-24 19:26:13 <barthalion> use pastebin for longer pastes
2013-05-24 19:26:17 <alpinenoob1> sorry!
2013-05-24 19:29:27 <alpinenoob1> there is a const char *getpackageversion() function with implementation in the project.
2013-05-24 19:29:55 <kaniini> Odd.
2013-05-24 19:30:04 <kaniini> oh.
2013-05-24 19:30:05 <kaniini> well.
2013-05-24 19:30:07 <kaniini> hmm
2013-05-24 19:30:16 <kaniini> you might need to mess with it a bit
2013-05-24 19:30:30 <kaniini> point is, you need to link against libintl
2013-05-24 19:30:42 <kaniini> so, might have to hack build system after all
2013-05-24 19:30:51 <alpinenoob1> hm, or not. i have multiple versions of the same project
2013-05-24 19:30:51 <alpinenoob1> checking again!
2013-05-24 19:44:45 <alpinenoob1> so i am not good with c++ .h files. the missing bits are declared here, http://pastebin.com/QM41D1Z5 can i just delete some portions to make it visible to all or something like that?
2013-05-24 20:02:26 <alpinenoob1> i am not good with make files either.
2013-05-25 07:10:11 <barthalion> ncopa: Is it possible to skip one patch while building vim?
2013-05-25 08:21:04 <barthalion> \o/
2013-05-25 08:45:05 <clandmeter> :)
2013-05-25 20:03:40 <build64-2-5> files from v2.5.4-121-gdccf8cc uploaded
2013-05-25 21:37:49 <build64-2-4> files from v2.4.11-20-gba11fef uploaded
2013-05-25 21:38:00 <build-2-4> files from v2.4.11-20-gba11fef uploaded
2013-05-25 22:30:13 <build64-2-5> files from v2.5.4-122-g8bab026 uploaded
2013-05-25 22:30:23 <build-2-5> files from v2.5.4-122-g8bab026 uploaded
2013-05-25 22:34:09 <build-2-4> files from v2.4.11-21-g43ac323 uploaded
2013-05-25 22:34:44 <build64-2-3> files from v2.3.6-214-g2857721 uploaded
2013-05-25 22:36:03 <tdtrask> ncopa: I'm having problems installing postgresql on alpine 2.2
2013-05-25 22:36:05 <tdtrask> /usr/bin/psql: symbol 'CRYPTO_memcmp': can't resolve symbol in lib '/usr/lib/libssl.so.1.0.0'.
2013-05-25 22:36:20 <tdtrask> seems like the recent security fixes caused some problems
2013-05-25 22:48:32 <jbilyk__> tdtrask: can you try bumping pkgrel on postgresql package in alp 2.2?
2013-05-25 22:48:42 <jbilyk__> that will probably fix the issue
2013-05-25 22:58:28 <build64-2-5> files from v2.5.4-123-g09b34fd uploaded
2013-05-25 22:58:37 <build-2-5> files from v2.5.4-123-g09b34fd uploaded
2013-05-26 02:00:47 <Mp5shooter> eek
2013-05-26 02:01:10 <Mp5shooter> whoever upgraded dovecot - there's a README in /etc/dovecot :O
2013-05-26 06:09:43 <barthalion> uf, it wasn't me this time
2013-05-26 06:11:50 <barthalion> tdtrask: ^
2013-05-26 06:16:22 <build64-2-2> files from v2.2.3-232-gad39fe8 uploaded
2013-05-26 06:16:29 <build-2-2> files from v2.2.3-232-gad39fe8 uploaded
2013-05-26 12:51:37 <tdtrask> barthalion / jbilyk: thanks
2013-05-26 12:52:00 <tdtrask> how do we know that's the only package needing rebuild?
2013-05-26 13:32:35 <tdtrask> btw, the rebuild of postgresql fixed the problem
2013-05-26 19:15:06 <jbilyk__> tdtrask: i think a grep of anything with openssl in makedepends ought to do it
2013-05-26 19:35:38 <barthalion> tdtrask: actually it would be handy to have a method to check abi/api change other than upstream-tracker.org
2013-05-26 21:03:54 <barthalion> 
ACTION almost added apk support to Ansible

2013-05-26 21:04:20 <barthalion> I need to add upgrading only specified packages and forcing
2013-05-27 07:17:25 <kaniini>  v/buffer 74
2013-05-27 07:46:57 <clandmeter> morning
2013-05-27 07:48:44 <clandmeter> ncopa: do you know how to disable logging in awall, or move it to some seperate logfile?  my dmesg is completely filled up.
2013-05-27 07:50:15 <clandmeter> i checked the wiki, but awall's documentation is a little hard to read from enduser point of view.
2013-05-27 07:50:27 <kaniini> unfortunatly iptables does not log to file
2013-05-27 07:50:32 <kaniini> only to dmesg
2013-05-27 07:50:33 <kaniini> :/
2013-05-27 07:51:35 <clandmeter> i would prefer to disable it if logfile is not supported
2013-05-27 08:48:41 <kaniini> interesting
2013-05-27 08:48:45 <kaniini> openbsd was forked
2013-05-27 08:48:50 <kaniini> how did i miss that
2013-05-27 08:49:44 <barthalion> when?
2013-05-27 08:49:46 <barthalion> to what?
2013-05-27 08:49:49 <barthalion> :o
2013-05-27 08:50:27 <barthalion> ah, you mean bitrig probably
2013-05-27 08:50:30 <kaniini> bitrig
2013-05-27 08:50:32 <kaniini> yes
2013-05-27 08:57:28 <vkrishn> some interesting way of using git, https://github.com/joeyh
2013-05-27 08:58:15 <barthalion> lol
2013-05-27 08:58:28 <barthalion> 79.207.25.10 authored 2 hours ago
2013-05-27 09:01:31 <ncopa> clandmeter: i dont know
2013-05-27 09:25:37 <barthalion> libpng 1.5 seems to be discontinued
2013-05-27 09:25:48 <barthalion> or just website is misleading
2013-05-27 11:16:18 <vkrishn> some links for al: http://dpaste.de/f4qdT/
2013-05-27 11:16:29 <vkrishn> maybe good for wikipedia
2013-05-27 12:01:22 <tdtrask> anyone care to dive into the /. comments on virtualization and recommend alpine? http://ask.slashdot.org/story/13/05/26/1813216/ask-slashdot-which-100-user-virtualization-solution-should-i-use
2013-05-27 12:01:59 <tdtrask> 
ACTION isn't too familiar, but knows others here care

2013-05-27 13:49:36 <vkrishn> https://vimeo.com/21570609 - 2yrs old but goood for wikipedia
2013-05-27 14:01:09 <build64-2-4> files from v2.4.11-22-gaac3ab6 uploaded
2013-05-27 14:22:28 <build64-2-3> files from v2.3.6-230-gdc3a97f uploaded
2013-05-27 14:26:33 <ncopa> build-2-3: retry
2013-05-27 14:30:26 <rnalrd> build64-edge: retry
2013-05-27 14:32:39 <rnalrd> ncopa: what's the new syntax for edge bots?
2013-05-27 14:32:56 <ncopa> algitbot: build master
2013-05-27 14:33:07 <rnalrd> tnx
2013-05-27 14:33:50 <ncopa> rnalrd: thanks for updating linux-virt-grsec
2013-05-27 14:33:59 <rnalrd> yw
2013-05-27 14:34:05 <ncopa> it is some work with it...
2013-05-27 14:34:10 <ncopa> as you notice :)
2013-05-27 14:34:19 <rnalrd> it builds just fine on brand new env
2013-05-27 14:34:21 <rnalrd> mm
2013-05-27 14:35:21 <ncopa> algitbot: build master
2013-05-27 14:41:23 <clandmeter> ncopa: regarding webapps, if we have a default httpd group. isnt that easier to setup?
2013-05-27 14:41:47 <ncopa> i suppose it would
2013-05-27 14:41:58 <ncopa> i think debian has a group www
2013-05-27 14:42:10 <clandmeter> makes upgrade webapps easier
2013-05-27 14:42:10 <hl> debian uses www-data
2013-05-27 14:42:24 <build64-2-5> files from v2.5.4-124-g9375938 uploaded
2013-05-27 14:43:05 <ncopa> and the webservers are in this groups i suppose
2013-05-27 14:43:30 <clandmeter> that would be the right way i suppose
2013-05-27 14:44:31 <ncopa> i would think so yes
2013-05-27 14:45:46 <build64-2-3> files from v2.3.6-231-g1a986b6 uploaded
2013-05-27 14:51:44 <build64-2-5> files from v2.5.4-125-gc42ea19 uploaded
2013-05-27 14:51:49 <build-2-5> files from v2.5.4-125-gc42ea19 uploaded
2013-05-27 14:53:42 <build-2-4> files from v2.4.11-24-g2f6ffeb uploaded
2013-05-27 14:54:22 <build64-2-3> files from v2.3.6-232-g337ebba uploaded
2013-05-27 15:02:14 <rnalrd> hope that fixes it
2013-05-27 15:02:16 <rnalrd> gotta go
2013-05-27 15:10:48 <clandmeter> there will be infra maintainance on NL location. this could create downtime for bugs/www/mirror/rsync between 18:00 and 20:00 hours CET.
2013-05-27 15:11:11 <ncopa> thanks for letting us know
2013-05-27 15:11:36 <jbilyk> clandmeter: that's today?
2013-05-27 15:12:23 <clandmeter> yes
2013-05-27 15:12:27 <jbilyk> ok
2013-05-27 15:12:45 <jbilyk> clandmeter: want me to send a notice to alpine-devel ML too?
2013-05-27 15:17:59 <clandmeter> im sending now
2013-05-27 15:18:19 <jbilyk> thx
2013-05-27 16:09:00 <build-2-3> files from v2.3.6-232-g337ebba uploaded
2013-05-27 16:33:19 <barthalion> rnalrd: 3.9.4 is out
2013-05-27 17:00:03 <clandmeter> seems nl kvm has been updated recently, it fails to restart
2013-05-27 17:00:22 <clandmeter> im not sure what ncopa did, but i get a segfault at booting :(
2013-05-27 17:00:52 <clandmeter> i have another vm which has boot issues, so this is my prio atm.
2013-05-27 17:58:17 <clandmeter> fabled, here?
2013-05-27 17:58:54 <clandmeter> anybody has ever had this error?
2013-05-27 17:59:06 <clandmeter> ERROR: Assigning locked name
2013-05-27 18:16:22 <jbilyk> clandmeter: i haven't ever seen that error
2013-05-27 18:16:40 <clandmeter> im upgrading to 2.6
2013-05-27 18:20:22 <clandmeter> reboot it now
2013-05-27 18:20:27 <clandmeter> should come back online i hope
2013-05-27 18:22:21 <clandmeter> we have lift off
2013-05-27 18:23:19 <clandmeter> hail running from RAM!
2013-05-27 18:23:21 <jbilyk> clandmeter: i can load www and nl mirror :)
2013-05-27 18:23:58 <clandmeter> now i need to fix a broken esxi server :|
2013-05-28 00:13:13 <kaniini> oh man
2013-05-28 00:13:31 <kaniini> i am watching that talk vkrishn linked on vimeo
2013-05-28 00:13:53 <kaniini> apparently alpine is the tool of choice for building infrastructure amongst activists these days
2013-05-28 00:15:27 <kaniini> apparently the big wins are
2013-05-28 00:15:31 <kaniini> 1. it's tamperproof (duh)
2013-05-28 00:15:47 <kaniini> 2. it can be deployed on cheap outdated hardware that it doesn't matter if the hardware is looted/confiscated/etc
2013-05-28 00:18:38 <Mp5shooter> ;p;
2013-05-28 07:10:10 <barthalion> to be honest, I stumbled upon Alpine because I needed distro that could be run from usb key on thinclient without big performance loss
2013-05-28 08:02:48 <clandmeter> ncopa: around?
2013-05-28 08:03:46 <ncopa> morning clandmeter
2013-05-28 08:03:54 <ncopa> i see you had issues yesterday
2013-05-28 08:03:57 <ncopa> how did it go?
2013-05-28 08:04:11 <ncopa> looks like www and bugs are up
2013-05-28 08:04:30 <clandmeter> i dont know what happend
2013-05-28 08:04:33 <clandmeter> you tell me :p
2013-05-28 08:04:42 <clandmeter> did you do anything recently to kvm?
2013-05-28 08:05:02 <clandmeter> i know you upgraded some parts of the guests
2013-05-28 08:07:03 <clandmeter> the time is off for the kvm guests
2013-05-28 08:08:00 <clandmeter> fixed
2013-05-28 08:08:14 <clandmeter> forgot i was inside vserver and kvm :)
2013-05-28 08:14:07 <ncopa> :)
2013-05-28 08:14:14 <ncopa> i think i upgradeed alpine-vserver
2013-05-28 08:14:28 <ncopa> atleast i upgraded some of the vserver guests there
2013-05-28 08:14:34 <ncopa> i dont remember if i upgraded alpine-vserver
2013-05-28 08:14:36 <ncopa> i wanted to
2013-05-28 08:14:58 <ncopa> no i didnt upgrade it
2013-05-28 08:15:14 <ncopa> maybe should do so now
2013-05-28 08:15:19 <ncopa> is it ok if i do?
2013-05-28 08:15:34 <ncopa> we should reboot it afterwards
2013-05-28 08:16:03 <ncopa> downtime for bugs and www should be tops 5min
2013-05-28 08:27:47 <ncopa> lo
2013-05-28 08:27:48 <ncopa> lol
2013-05-28 08:27:59 <ncopa> http://www.phoronix.com/scan.php?page=news_item&px=MTM3OTY
2013-05-28 08:28:21 <ncopa> "dropping a README.md or README.rst or README.txt into your git repository will cause the /repo/about page to have markdown/restructuredtext/plain-converted output, in much the same way that Github does it."
2013-05-28 08:31:17 <barthalion> new maintainer does really good job
2013-05-28 08:31:24 <barthalion> do we have cgit in repos?
2013-05-28 08:31:30 <ncopa> yes
2013-05-28 08:31:35 <barthalion> oops
2013-05-28 08:31:43 <ncopa> the funny thing is
2013-05-28 08:31:47 <barthalion> let me update is
2013-05-28 08:31:49 <barthalion> it*
2013-05-28 08:31:54 <barthalion> not funny is security bug
2013-05-28 08:32:03 <ncopa> i implemented README.md in cgit a few weeks ago
2013-05-28 08:32:08 <barthalion> :D
2013-05-28 08:32:21 <ncopa> before you commit it.. i'm gonna assign a issue for it
2013-05-28 08:32:32 <barthalion> sure
2013-05-28 08:33:21 <ncopa> #2017
2013-05-28 08:34:08 <barthalion> I'm not authorized to access this
2013-05-28 08:34:13 <ncopa> i'm on it
2013-05-28 08:34:16 <ncopa> you need access there
2013-05-28 08:34:30 <ncopa> i think all package maintainers and devs should have access
2013-05-28 08:36:17 <ncopa> i htink you have now
2013-05-28 08:36:36 <ncopa> you can just add: ref #2017 in commit message
2013-05-28 08:36:49 <ncopa> I'll create subtasks for the stable branches
2013-05-28 08:37:24 <barthalion> hm, zx2c4.com is down here
2013-05-28 08:37:46 <ncopa> i got a screenshot
2013-05-28 08:39:03 <barthalion> yup, for some reason I can't fetch cgit-0.9.2.tar.xz
2013-05-28 08:40:29 <ncopa> heh
2013-05-28 08:40:34 <ncopa> i think i its "slashdotted"
2013-05-28 08:41:02 <ncopa> its down yes
2013-05-28 08:46:11 <barthalion> I'll leave this git release for now
2013-05-28 08:46:20 <barthalion> I mean, 1.7.4
2013-05-28 08:46:30 <barthalion> I'll upgrade it later for edge
2013-05-28 08:47:14 <barthalion> or not
2013-05-28 08:50:45 <barthalion> make[1]: ../cgit.mk: No such file or directory hrm
2013-05-28 09:05:25 <barthalion> ln -s was bad idea
2013-05-28 09:05:32 <barthalion> cp -r worked
2013-05-28 09:32:47 <clandmeter> ncopa: kvm is already updated to 2.6 now.
2013-05-28 09:33:10 <clandmeter> since yesterday. i had to do something to make it boot agian.
2013-05-28 09:33:50 <clandmeter> thats why i asked you changed something to kvm host. it failed booting when i rebooted it.
2013-05-28 09:34:18 <ncopa> hm
2013-05-28 09:34:49 <clandmeter> the error i msg yesterday in the chan
2013-05-28 09:34:59 <clandmeter> after that error it segfaulted
2013-05-28 09:35:04 <clandmeter> looked like some apk msg
2013-05-28 09:35:24 <ncopa> ERROR: Assigning locked name
2013-05-28 09:35:27 <ncopa> that one?
2013-05-28 09:36:51 <clandmeter> yes
2013-05-28 11:57:44 <ncopa> kaniini: i think it would be nice with native upstream support for alpine linux in unetbootin
2013-05-28 11:58:04 <kaniini> ncopa: i dunno, i only use unetbootin to build windows install usb media
2013-05-28 11:59:04 <kaniini> ncopa: but yes, i will talk to them about it.
2013-05-28 11:59:08 <kaniini> ncopa: it shouldn't be hard.
2013-05-28 11:59:20 <kaniini> and, i had the same thought.
2013-05-28 11:59:32 <ncopa> i looked at it some year ago or so
2013-05-28 11:59:43 <ncopa> the list of distros etc was coded into the c++ code
2013-05-28 11:59:46 <ncopa> a bit ugly...
2013-05-28 11:59:52 <kaniini> yeah it's still that way
2013-05-28 12:00:20 <ncopa> but it would be nice if windows users could dl unetbootin.exe and be able to build alpine boot usbs
2013-05-28 12:31:14 <barthalion> jbilyk: I moved my mailbox to different server, does alias need to be refreshed?
2013-05-28 12:31:26 <jbilyk> nope, as long as address is the same
2013-05-28 12:31:30 <jbilyk> just wait for ttl to expire
2013-05-28 12:31:42 <jbilyk> email address the same that is
2013-05-28 12:32:12 <barthalion> just wondering, because everything spotted the change, but alias not yet
2013-05-28 13:07:01 <jbilyk> barthalion: so relay=mail.bpiotrowski.pl[178.33.81.96] is incorrect?
2013-05-28 13:08:02 <jbilyk> seems like dns on mail.a.o is still picking up that address
2013-05-28 13:08:16 <barthalion> jbilyk: it should be 188.40.51.140 now
2013-05-28 13:08:47 <barthalion> or mail.server-speed.net
2013-05-28 13:09:59 <jbilyk> barthalion: the ttl on the record is still good for 2h on mail.a.o
2013-05-28 13:10:15 <jbilyk> ;; ANSWER SECTION:
2013-05-28 13:10:15 <jbilyk> bpiotrowski.pl.		7918	IN	MX	1 mail.bpiotrowski.pl.
2013-05-28 13:10:16 <jbilyk> ;; ADDITIONAL SECTION:
2013-05-28 13:10:16 <jbilyk> mail.bpiotrowski.pl.	7918	IN	A	178.33.81.96
2013-05-28 13:10:31 <barthalion> thanks
2013-05-28 13:12:00 <jbilyk> barthalion: if it's still not working then, pls feel free to ping me
2013-05-28 14:55:12 <build64-2-3> files from v2.3.6-237-g2a73f39 uploaded
2013-05-28 14:55:24 <build-2-4> files from v2.4.11-28-g8fa342f uploaded
2013-05-28 14:55:26 <build64-2-5> files from v2.5.4-127-g90230ac uploaded
2013-05-28 15:00:25 <ncopa> algitbot: build master
2013-05-28 23:04:51 <shafire> ah
2013-05-28 23:04:52 <shafire> hi
2013-05-28 23:34:45 <Mp5shooter> [    1.121317] /home/buildozer/aports/main/linux-grsec/src/linux-3.6/drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
2013-05-28 23:34:47 <Mp5shooter> this shows in dmesg
2013-05-28 23:34:51 <Mp5shooter> just noting
2013-05-29 04:45:03 <kaniini> clandmeter: when do you plan to get the arm builder/porting machine
2013-05-29 06:42:03 <clandmeter> kaniini: only if somebody would like to port it.
2013-05-29 06:44:55 <clandmeter> kaniini: the ODROID-U2 is not that interesting other than nice cpu and memory (which should be powerfull enough for our infra)
2013-05-29 06:51:29 <clandmeter> would be nice if we could get a ARM Cortex-A57
2013-05-29 07:23:30 <kaniini> clandmeter: i am interested if we have a plan to get acceptable equipment to facilitate the port
2013-05-29 07:23:45 <kaniini> clandmeter: ODROID-U2 is okay if you can find a supplier
2013-05-29 07:29:36 <kaniini> clandmeter: the problem with ODROID-U2 is local storage
2013-05-29 07:30:42 <kaniini> cubieboard has SATA
2013-05-29 07:31:10 <kaniini> i mean, do we want to deal with external USB HDDs and such in the server room?
2013-05-29 07:49:03 <shafire> hi
2013-05-29 08:09:57 <build64-2-5> files from v2.5.4-128-g4646c3e uploaded
2013-05-29 08:09:58 <royger> kaniini: if you get an A7/A15 it will also have the virt extensions
2013-05-29 08:11:44 <royger> there's the arndale and the Allwiner A20, but I haven't been able to find any A20 that has a serial output
2013-05-29 08:41:15 <shafire> how can I use Makefile.dkms right on alpine?
2013-05-29 08:41:20 <shafire> I want to install flashcache
2013-05-29 08:50:24 <build64-2-4> files from v2.4.11-29-ga2ad5e0 uploaded
2013-05-29 08:50:32 <build-2-4> files from v2.4.11-29-ga2ad5e0 uploaded
2013-05-29 09:10:52 <vkrishn> rnalrd: does inotify-tools require anything enabled in kernel?
2013-05-29 09:19:15 <ncopa> aww... we have problems
2013-05-29 09:19:29 <ncopa> abuild-2.14.2.tar.bz2: FAILED
2013-05-29 09:19:30 <ncopa> md5sum: WARNING: 1 of 1 computed checksums did NOT match
2013-05-29 09:19:33 <ncopa> i have updated cgit
2013-05-29 09:19:40 <ncopa> and not checksums does not match
2013-05-29 09:19:44 <ncopa> now*
2013-05-29 09:21:23 <ncopa> i suppose what we need to do is have a git hook that does git-archive when there are tags
2013-05-29 09:34:49 <ncopa> whatta
2013-05-29 09:35:08 <ncopa> seems like /bin/addgroup moved to /usr/sbin/addgroup
2013-05-29 09:59:41 <vkrishn> I thought /dev/inotify was needed by inotify-tools, but it works, thanks
2013-05-29 12:49:45 <kaniini> royger: does arndale have sata
2013-05-29 12:51:49 <kaniini> arndale does not have sata
2013-05-29 12:52:01 <kaniini> why do they make all these boards and don't put sata on any of them
2013-05-29 13:08:17 <StarWarsFan> how long is the usual time until a patch via mailing list will be applied?
2013-05-29 13:09:23 <StarWarsFan> i've sent a simple patch (my first one here... ;-) on 22th but nothing happend until now
2013-05-29 13:10:08 <StarWarsFan> so i'm wondering if i did something wrong, used wrong format or just have to wait any longer...
2013-05-29 13:15:57 <jbilyk> StarWarsFan: ncopa might've just missed it
2013-05-29 13:16:40 <ncopa> StarWarsFan: what can i search for?
2013-05-29 13:16:49 <ncopa> it probably just fell between the cracks
2013-05-29 13:17:26 <StarWarsFan> [PATCH] Typo on setup-timezone
2013-05-29 13:17:38 <StarWarsFan> 2013-05-22
2013-05-29 13:17:51 <ncopa> ah
2013-05-29 13:17:56 <ncopa> i saw it
2013-05-29 13:18:05 <ncopa> but i just forgot
2013-05-29 13:18:06 <ncopa> sorry
2013-05-29 13:18:35 <ncopa> that would probably been nice to have in v2.6.1
2013-05-29 13:18:47 <ncopa> a tip
2013-05-29 13:19:00 <ncopa> create an issu on bugs.a.o
2013-05-29 13:19:08 <ncopa> then a patch that yo usend to alpine-devel
2013-05-29 13:19:18 <ncopa> in commit message you include: ref #<issue>
2013-05-29 13:19:37 <ncopa> issue tracker helps to not forget
2013-05-29 13:19:50 <StarWarsFan> ah ok
2013-05-29 13:20:02 <ncopa> but it shoudl be enough what you did
2013-05-29 13:20:10 <ncopa> i'm just flooded..
2013-05-29 13:20:20 <ncopa> tsunami of emails
2013-05-29 13:22:11 <ncopa> other thing
2013-05-29 13:22:16 <ncopa> User-Agent: Roundcube Webmail/0.9-beta
2013-05-29 13:22:31 <ncopa> it helps if i can: |git am
2013-05-29 13:22:34 <ncopa> it from mutt
2013-05-29 13:22:47 <ncopa> so using git send-mail helps too
2013-05-29 13:23:29 <StarWarsFan> yes i know
2013-05-29 13:23:39 <ncopa> if you cannot use git send-mail for some reason, then its probably better to attach it to an issue on bugs.a.o
2013-05-29 13:23:49 <StarWarsFan> but for this i have to setup some more details on my mail system...
2013-05-29 13:24:18 <StarWarsFan> but actually i don't want to do this because it's more or less wasted time
2013-05-29 13:24:34 <StarWarsFan> because i think to switch my mail server to alpine ;-)
2013-05-29 13:24:35 <ncopa> you can add smtp server in your .gitconfig
2013-05-29 13:24:50 <ncopa> hehe :)
2013-05-29 13:24:55 <ncopa> np
2013-05-29 13:25:07 <StarWarsFan> but thx for the input!
2013-05-29 13:26:04 <ncopa> thats for the patch
2013-05-29 13:26:09 <ncopa> and sorry for beeing slow
2013-05-29 13:35:00 <ncopa> StarWarsFan: http://git.alpinelinux.org/cgit/alpine-conf/commit/?id=f65ddba2cf01de088877fa635610c3c94aa6c9c9
2013-05-29 13:35:03 <ncopa> looks ok?
2013-05-29 13:35:14 <ncopa> do you think you could doublecheck it?
2013-05-29 13:35:26 <StarWarsFan> just checking...
2013-05-29 13:35:29 <ncopa> if ok i'll tag 2.12.1 release of alpine-conf
2013-05-29 13:35:56 <StarWarsFan> looks ok to me <thumbs-up/>
2013-05-29 13:36:00 <StarWarsFan> :-)
2013-05-29 14:34:25 <kaniini> clandmeter: this looks like a better fit, it has SATA and can boot off of it with u-boot: http://boundarydevices.com/products/nitrogen6x-board-imx6-arm-cortex-a9-sbc/
2013-05-29 14:36:03 <kaniini> what we want is an arm board with sata and 2+gb ram and >1ghz multi-core cpu
2013-05-29 14:36:18 <kaniini> that way we can be certain the builder will keep up
2013-05-29 15:00:38 <kaniini> clandmeter: it seems arndale has SATA
2013-05-29 15:02:31 <kaniini> http://www.howchip.com/shop/item.php?it_id=AND5250B
2013-05-29 15:03:10 <kaniini> clandmeter: ^
2013-05-29 15:03:21 <kaniini> clandmeter: faster chip
2013-05-29 15:03:25 <kaniini> clandmeter: SATA
2013-05-29 15:12:21 <royger> kaniini: arndale has SATA3
2013-05-29 15:13:04 <royger> kaniini: http://www.arndaleboard.org/wiki/index.php/Main_Page
2013-05-29 15:13:12 <royger> but it's quite expensive...
2013-05-29 15:13:14 <kaniini> yap
2013-05-29 15:13:27 <kaniini> $250 isn't too bad
2013-05-29 15:13:34 <kaniini> considering serial and sata
2013-05-29 15:13:40 <kaniini> we need sata regardless
2013-05-29 15:13:51 <royger> why is sata so important?
2013-05-29 15:13:53 <kaniini> builders require many hundreds of gb of space
2013-05-29 15:14:05 <royger> external HDD using USB2/3?
2013-05-29 15:14:16 <kaniini> could work
2013-05-29 15:14:30 <kaniini> main thing is, we need hdd somehow
2013-05-29 15:14:35 <kaniini> ideally, more than one hdd
2013-05-29 15:14:37 <kaniini> for raid
2013-05-29 15:14:49 <kaniini> could boot off cf card
2013-05-29 15:15:05 <kaniini> or SD or whatever the rage is
2013-05-29 15:15:13 <royger> yes, you could probably also setup iscsi, but using iscsi without a dedicated ethernet I'm not sure if it's a good idea for speed
2013-05-29 15:15:56 <kaniini> i'd rather... not do that
2013-05-29 15:16:00 <kaniini> again
2013-05-29 15:16:06 <kaniini> we want the builder to keep up
2013-05-29 15:16:06 <royger> I guess you could also boot off PXE
2013-05-29 15:16:07 <kaniini> lol
2013-05-29 15:16:18 <kaniini> booting isn't the problem
2013-05-29 15:16:25 <royger> and then use USB/SATA or whatever
2013-05-29 15:16:27 <kaniini> sdcard, emmc, etc has that covered
2013-05-29 17:59:57 <barthalion> Is it expected? ERROR: linux-virt-grsec: Trying to overwrite usr/share/kernel/grsec/kernel.release owned by linux-grsec
2013-05-29 18:10:58 <barthalion> Anyway, I successfully installed Alpine on KVM machine from Front Range Hosting
2013-05-29 18:11:29 <barthalion> They're cheap and mostly useless, but still
2013-05-30 00:58:21 <Mp5shooter> huh
2013-05-30 00:58:29 <Mp5shooter> my terminal colors disappeared
2013-05-30 00:58:32 <Mp5shooter> i think after i installed coreutils
2013-05-30 01:03:13 <Mp5shooter> new dir appeared
2013-05-30 01:03:14 <Mp5shooter> /etc/terminfo
2013-05-30 01:03:32 <Mp5shooter> and why is coreutils installed with rutorrent o.O
2013-05-30 01:04:09 <Mp5shooter> yeah
2013-05-30 01:04:10 <Mp5shooter> so weird
2013-05-30 01:04:18 <Mp5shooter> purged rutorrent and terminal colors are back
2013-05-30 01:04:25 <Mp5shooter> must be coreutils messing with stuff
2013-05-30 07:03:39 <clandmeter> kaniini: i looked at SATA solutions. there are (what i can find) 2 solutions. one is 250 dollar (3x more expensive) one is A10 which is much lower in performance.
2013-05-30 07:08:08 <kaniini> clandmeter: hdd over usb should work
2013-05-30 07:10:15 <royger> just my two cents, at citrix they are using the arndale board to test xen on arm, and it seems to be a good platform
2013-05-30 07:15:15 <kaniini> citrix has a larger budget than we do
2013-05-30 07:15:22 <kaniini> we have lunch money compared to them ;p
2013-05-30 07:16:31 <clandmeter> :)
2013-05-30 07:16:44 <clandmeter> usb3 would also be nice
2013-05-30 07:17:12 <clandmeter> i think the new exnos 5 has it. but probably too expensive ones releaed.
2013-05-30 07:17:35 <kaniini> yes, usb3 is preferable, and we want at least two usb3 ports
2013-05-30 07:17:40 <kaniini> so we can do swraid
2013-05-30 07:26:40 <StarWarsFan> Hi
2013-05-30 07:27:06 <StarWarsFan> How is the hardware etc paid?
2013-05-30 07:27:28 <StarWarsFan> Is there some sort of a sponsor?
2013-05-30 07:28:42 <StarWarsFan> And where is it located?
2013-05-30 07:31:24 <kaniini> the hardware is paid for whoever buys it and sets it up ;p
2013-05-30 07:31:38 <kaniini> hince why i said our budget is 'lunch money'
2013-05-30 07:32:29 <StarWarsFan> i c
2013-05-30 07:41:43 <clandmeter> kaniini: are you serious about porting?
2013-05-30 07:42:50 <kaniini> clandmeter: enough where i consider it to be a hardblock for 2.7
2013-05-30 07:44:46 <clandmeter> do you need it to do the actual porting/toolchain?
2013-05-30 10:51:14 <kaniini> i can do versatile-express, but we need to do porting work to the target board as well
2013-05-30 10:54:31 <kaniini> with arm you have to ship kernel images for each individual device you want to support
2013-05-30 10:58:10 <clandmeter> kaniini: I want to keep my budget within 100$. I see http://cubieboard.org with SATA but the SoC is not as nice as the ODROID-U2. Which do you prefer?
2013-05-30 10:59:54 <kaniini> ODROID-U2 is fine
2013-05-30 11:00:24 <clandmeter> ok ill order it.
2013-05-30 11:00:29 <kaniini> we have a chicken and egg problem
2013-05-30 11:00:38 <kaniini> we need an arm port to get people interested in the arm port
2013-05-30 11:00:46 <kaniini> so, if we can ship in 2.7 a port to the ODROID-U2
2013-05-30 11:00:48 <StarWarsFan> *g*
2013-05-30 11:00:52 <kaniini> other stakeholders might show up
2013-05-30 11:00:56 <kaniini> is what i am hoping
2013-05-30 11:00:58 <kaniini> :)
2013-05-30 11:01:28 <kaniini> but, we had this same problem with x86-64 until i ported it (and fixed uclibc itself in the process)
2013-05-30 11:01:50 <kaniini> uclibc/amd64 was in a really bad state when i started that port.
2013-05-30 11:01:53 <clandmeter> 
ACTION wonders if he should buy the optionally eMMC

2013-05-30 11:01:56 <kaniini> i am hoping arm is in a better place.
2013-05-30 11:02:00 <barthalion> Well, I have beaglebone in local hackerspace and Efika MX close at hand
2013-05-30 11:02:23 <kaniini> clandmeter: yes, i am thinking we flash the actual alpine environment to eMMC
2013-05-30 11:02:39 <kaniini> clandmeter: and then /home/buildozer on hdd
2013-05-30 11:03:05 <clandmeter> it can also boot from uSD
2013-05-30 11:03:13 <kaniini> that works
2013-05-30 11:03:23 <clandmeter> and if we run from ram, i dont see the benifits of eMMC
2013-05-30 11:03:39 <clandmeter> 2G should be enough to run from ram i guess.
2013-05-30 11:03:58 <kaniini> i think we shouldn't run from ram for a builder
2013-05-30 11:04:08 <kaniini> it may make porting e.g. java harder
2013-05-30 11:04:13 <kaniini> otoh
2013-05-30 11:04:20 <kaniini> we can just put a 30gb swapfile on the hdd
2013-05-30 11:04:37 <kaniini> the hdd may eventually fail, but...
2013-05-30 11:04:39 <kaniini> lol
2013-05-30 11:04:47 <clandmeter> i have a spare ssd around
2013-05-30 11:05:03 <kaniini> the ssd will eventually fail :P
2013-05-30 11:05:11 <kaniini> i'd say about 2 years
2013-05-30 11:05:14 <kaniini> haha
2013-05-30 11:05:50 <kaniini> also, i don't think run-from-ram is supported for builders
2013-05-30 11:06:02 <kaniini> at least super_duck ran into trouble there
2013-05-30 11:07:02 <kaniini> then in 2.8 we can expand arm port
2013-05-30 11:07:04 <kaniini> more devices
2013-05-30 11:07:07 <kaniini> possibly xen
2013-05-30 11:11:14 <kaniini> i think for 2.7 our goals should be
2013-05-30 11:11:17 <kaniini> ODROID-U2
2013-05-30 11:11:23 <kaniini> and raspberry pi
2013-05-30 11:11:29 <kaniini> for fully supported devices
2013-05-30 11:13:07 <kaniini> raspberry pi is biggest possible win for developer outreach in my opinion
2013-05-30 11:15:06 <clandmeter> 30 dollars shipping... they come over and bring it themselves?
2013-05-30 11:18:38 <kaniini> most of these boards are dropshipped from korea
2013-05-30 12:49:01 <super_duck> kaniini, you can run from ram just fine, but you have to #SUDO_APK="abuild-apk --force" abuild -r
2013-05-30 12:50:24 <super_duck> there was also something about tmpfs with something under /tmp or /etc, but I forget what that was exactly
2013-05-30 13:48:00 <kaniini> super_duck: it still sounds like something we don't want to do for a production builder ;p
2013-05-30 13:54:29 <shafire> kaniini: are you a alpine developer, or?
2013-05-30 13:55:19 <shafire> I compiled flashcache, but the last step is to start flashcache on boot
2013-05-30 13:55:34 <shafire> and there is a Makefile.dkms, which I can not use with alpine, or?
2013-05-30 13:57:52 <kaniini> i am not familiar with flashcache, but you will need the -dev package for whatever installed kernel flavour you are using
2013-05-30 13:58:22 <shafire> I compiled it and everything works
2013-05-30 13:58:45 <kaniini> if your question is: does alpine presently support dkms -- answer is no
2013-05-30 13:58:49 <kaniini> (perhaps we should)
2013-05-30 13:59:18 <shafire> ok
2013-05-30 14:06:41 <kaniini> most of my work in alpine has to do with x86_64, uclibc, toolchain and mozilla.
2013-05-30 14:06:57 <kaniini> kernel is mostly an ncopa thing
2013-05-30 15:45:10 <shafire> kaniini: who is responsible for the website?
2013-05-30 15:46:21 <kaniini> clandmeter or ncopa or someone who is definitely not me ;)
2013-05-30 15:47:07 <shafire> they should really take a look at the download site
2013-05-30 15:47:13 <shafire> it's broken since days
2013-05-30 15:48:44 <jbilyk> shafire: what URL?
2013-05-30 15:49:21 <kaniini> yes the download page is broken
2013-05-30 15:49:24 <kaniini> visibly
2013-05-30 15:49:50 <jbilyk> http://alpinelinux.org/downloads ?
2013-05-30 15:55:52 <shafire> yes
2013-05-30 16:00:18 <jbilyk> just the xen iso? the rest seem fine
2013-05-30 16:00:27 <jbilyk> at least when viewed from here in chromium
2013-05-30 16:03:30 <shafire> Alpine Linux VServer: right box, wrong download link
2013-05-30 16:03:40 <shafire> Alpine Linux Xen: left box missing
2013-05-30 16:03:45 <shafire> Alpine Linux Xen: right box empty
2013-05-30 16:03:51 <kaniini> well
2013-05-30 16:03:54 <kaniini> left box should be missing
2013-05-30 16:03:58 <kaniini> we do not offer 32-bit xen
2013-05-30 16:04:08 <kaniini> i know this as i made the call on that
2013-05-30 16:04:12 <shafire> ah okay
2013-05-30 16:04:38 <shafire> ok, you are right, i am sorry
2013-05-30 16:04:49 <shafire> Alpine Linux VServer: right box, wrong download link
2013-05-30 16:04:54 <shafire> Alpine Linux Xen: right box empty
2013-05-30 16:33:14 <shafire> http://wiki.alpinelinux.org/wiki/Xen_Dom0 <-- could someone change "apk add xen" to "apk add xen xen-hypervisor"?
2013-05-30 17:07:05 <kaniini> you can
2013-05-30 17:07:09 <kaniini> just make an account.
2013-05-30 17:09:33 <shafire> Is xen-hypervisor bind to xen?
2013-05-30 17:09:36 <shafire> How can I check it?
2013-05-30 17:09:40 <shafire> Because yesterday, it was not
2013-05-30 17:27:19 <barthalion> It seems to be linked
2013-05-30 17:27:38 <barthalion> I mean, xen-hypervisor provides xen
2013-05-30 17:27:44 <barthalion> and apk add xen works fine here
2013-05-30 17:43:18 <shafire> barthalion: mh, so apk add xen-hypervisor instead apk add xen?
2013-05-30 17:44:46 <barthalion> http://git.alpinelinux.org/cgit/aports/tree/main/xen/APKBUILD hypervisor is subpackage of xen
2013-05-30 17:46:01 <shafire> mh, but "apk add xen" did not installed xen-hypervisor yesterday
2013-05-30 17:46:09 <shafire> maybe I did something wrong
2013-05-30 17:46:46 <shafire> there is also the /var/run/xen bug, the directory should be created on boot up
2013-05-30 17:46:59 <barthalion> report it or blame royger
2013-05-30 17:47:31 <shafire> ok
2013-05-30 17:47:50 <barthalion> I think it should look like apk add xen xen-hypervisor :p
2013-05-30 18:07:13 <royger> apk add xen used to work fine, and AFAIK the package has not changed
2013-05-30 18:10:54 <Mp5shooter> can someone remove gzip and coreutils from depends in APKBUILD for rutorrent please
2013-05-30 18:12:12 <barthalion> Mp5shooter: Are they unneeded?
2013-05-30 18:12:19 <Mp5shooter> yep
2013-05-30 18:12:32 <Mp5shooter> i didn't need them when i installed rutorrent by hand
2013-05-30 18:13:00 <barthalion> sure, I'll remove them in a second
2013-05-30 18:13:04 <Mp5shooter> ok ty
2013-05-30 18:26:20 <barthalion> Mp5shooter: ^
2013-05-30 18:26:39 <Mp5shooter> ty barthalion
2013-05-30 18:26:42 <Mp5shooter> :)
2013-05-30 19:01:04 <shafire> When will I get the verification email from bugs.alpinelinux.org?
2013-05-30 19:50:29 <Mp5shooter> hmm
2013-05-30 19:50:34 <Mp5shooter> i wonder if rutorrent plugins should be packaged
2013-05-30 19:50:35 <Mp5shooter> there's a ton of em
2013-05-30 19:53:15 <barthalion> Hm, any ideas why pv-grub doesn't see menu.lst of host? Path looks correct
2013-05-30 19:53:33 <barthalion> extra = '(hd0)/boot/grub/menu.lst'
2013-05-30 19:54:09 <barthalion> it drops me to grub shell, if I enter root, kernel, initrd and then boot, everything works
2013-05-30 19:55:27 <barthalion> and of course there is no errors before grub shell is shown
2013-05-30 20:06:48 <StarWarsFan> 
ACTION need some sleep

2013-05-30 20:06:51 <StarWarsFan> n8@all
2013-05-30 20:10:02 <barthalion> goodnight
2013-05-30 20:37:00 <barthalion> algitbot: build master
2013-05-30 21:42:22 <barthalion> broken ayway...
2013-05-30 21:42:34 <barthalion> I should hack mosh sources to ignore locales
2013-05-31 02:23:34 <jbilyk_> clandmeter: I didn't get a chance to look at downloads page on www.a.o today - if you get a chance tomorrow, can you check out the broken links reported this afternoon?
2013-05-31 06:45:18 <freedomrun> gm
2013-05-31 10:23:13 <barthalion> ncopa: I'm going to rewrite vim apkbuild
2013-05-31 10:23:14 <kaniini> royger: you about?
2013-05-31 10:23:53 <royger> kaniini: yes
2013-05-31 10:24:16 <kaniini> royger: i am seeing a strange bug with qemu-backed disk devices on xen 4.2.1-r11
2013-05-31 10:24:42 <kaniini> royger: specifically, booting the finnix kernel + iso fails
2013-05-31 10:25:07 <kaniini> royger: says timeout waiting for device
2013-05-31 10:25:15 <kaniini> remote state '1'
2013-05-31 10:25:27 <kaniini> but xenstore-ls shows remote state as 3 (connected)
2013-05-31 10:26:41 <royger> by remote do you mean the state of the backend?
2013-05-31 10:30:51 <kaniini> [    0.308595] blkfront: xvda1: flush diskcache: enabled
2013-05-31 10:30:52 <kaniini> [    0.311121] blkfront: xvda2: flush diskcache: enabled
2013-05-31 10:30:54 <kaniini> [    5.404069] XENBUS: Waiting for devices to initialise: 25s...20s...15s...10s...5s...0s...235s...230s...225s...220s...215s...210s...205s...200s...195s...190s...185s...180s...175s...170s...165s...160s...155s...150s...145s...140s...135s...130s...125s...120s...115s...110s...105s...100s...95s...90s...85s...80s...75s...70s...65s...60s...55s...50s...45s...40s...35s...30s...25s...20s...15s...10s...5s...0s...
2013-05-31 10:30:56 <kaniini> [  270.404216] XENBUS: Timeout connecting to device: device/vbd/51824 (local state 3, remote state 1)
2013-05-31 10:31:02 <kaniini> 51824 is the qdisk
2013-05-31 10:32:52 <kaniini> aha
2013-05-31 10:32:54 <kaniini> i found it
2013-05-31 10:33:00 <kaniini> we no longer create /var/run/xen
2013-05-31 10:33:14 <kaniini> i created it by hand and then it works
2013-05-31 10:35:56 <royger> I'm not sure why it is no longer created
2013-05-31 10:36:30 <royger> we could add a mkdir -p /var/run/xen to the xencommons init script
2013-05-31 10:37:12 <barthalion> doesn't openrc support tmpfiles.d?
2013-05-31 11:00:07 <kaniini> royger: xencommons is dead
2013-05-31 11:08:52 <kaniini> http://cgit.freedesktop.org/pkg-config/commit/?id=409ee76ce1ba3372f34ca4d14b629bd8ad8347e1
2013-05-31 11:09:03 <kaniini> yeah... i think we won't do that in pkgconf and say we didn't
2013-05-31 11:11:28 <kaniini> 
ACTION idly wonders what the new pkg-config maintainer smokes

2013-05-31 11:11:36 <kaniini> whatever it is, he hits it really hard...
2013-05-31 12:13:25 <super_duck> kaniini: you are right, for production purposes the build system should not care about the underlying file system except in the case that its out of space prior to build.