2018-03-16 16:44:52 ACTION  2018-03-16 20:18:25 hey guys, has anyone had any luck installing alpine using grub as a bootloader? for some reason when rebooting after running setup-alpine the efi is saying an os couldn’t be found 2018-03-16 20:19:09 if i mount sda1 it looks like grub is there, not a clue what to look for next - not had an installer fail to install a bootloader for me before lol 2018-03-16 20:20:09 <_ikke_> JordanDoyle: is the partition bootable? (or do you use gpt)? 2018-03-16 20:21:11 it should be, i just passed ‘sys’ to setup-disk and let it do its thing 2018-03-16 20:21:32 i assume that formats the partition to fat32 for me? 2018-03-17 02:09:04 dalias: i will make a pull request for the perf package soon that you requested a month ago. there are some things omited for the minimum requirements but you should ask for them if you want them enable like libunwind. 2018-03-17 02:09:37 hold up i will pull the full list of dependencies 2018-03-17 02:21:45 :))) 2018-03-17 02:27:02 https://pastebin.com/jH4BSpmU are the possible optional features enabled and disabled 2018-03-17 02:27:40 some might be inadvertantly be on because I am running with abuild -rK 2018-03-17 18:46:51 Hello, anyone active? 2018-03-17 18:47:09 no 2018-03-17 18:47:10 Hi, yes. 2018-03-17 18:47:12 <_ikke_> no 2018-03-17 18:47:14 <_ikke_> :P 2018-03-17 18:47:19 I've got alpine install in VMware and all the text is unusably tiny 2018-03-17 18:47:36 <_ikke_> in the console? 2018-03-17 18:47:38 3000x2000 on a 13.5 inch screen :( 2018-03-17 18:47:47 yes 2018-03-17 18:47:58 I installed a very small i think 10 MB iso 2018-03-17 18:48:03 I dont need a desktop env 2018-03-17 18:48:48 is setfont something I can use? 2018-03-17 18:48:57 i see it for ubuntu on SO 2018-03-17 18:49:40 You could try to connect to the installed system using ssh. Other than that there are boot parameters (vga=...) you could set. 2018-03-17 18:50:50 does ssh work over local network? I dont have a lot of experience with it 2018-03-17 18:50:55 yes 2018-03-17 18:51:15 you need to start sshd first 2018-03-17 18:51:36 I don't know VMware (I assume you mean the workstation and not ESXI) but in VirtualBox you also have a slider to adjust the scaling. 2018-03-17 18:52:06 VMware prompts you to install its vmware tools but i dont wanna do that 2018-03-17 18:52:38 I have tried it with debian in the pat. All is did was alow dynamic resizing, but kept scaling the same 2018-03-17 18:52:56 So i just end up with a lot of space for everything which is still tiny 2018-03-17 18:54:27 It says sshd requires re-exec with an absolute path? 2018-03-17 18:54:38 Does it need someplace to set up a unix socket or something? 2018-03-17 18:58:00 I never saw that message before to be true. 2018-03-17 19:16:04 question... I'm trying to use mkimage.sh. I got it to include custom packages. but what about installing them into the iso rather then including the packages on the iso? 2018-03-17 19:16:59 kfox1111: thats the same on alpine 2018-03-17 19:17:32 the installer system is built in ram from the packages and apkovl on the iso 2018-03-17 19:17:36 I don't understand. 2018-03-17 19:17:42 I'm following: https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage 2018-03-17 19:18:11 I can add apk's to the list but it doesn't install them into the ISO, just make them available for installing. 2018-03-17 19:18:35 trying to customize the iso to have a few things already done when the iso boots. 2018-03-17 19:19:10 ah, it should get added to /etc/apk/world in the apkovl then 2018-03-17 19:19:20 but im not sure how its done with the mkimage.sh script 2018-03-17 19:19:40 <_ikke_> I was checking the source if I could find out 2018-03-17 19:22:21 looks like that may come from mkiso.base.sh? 2018-03-17 19:23:38 <_ikke_> you mean mkimg.base.sh? 2018-03-17 19:23:38 is there any hooks that I could use to run commands before the iso's built to do further customizations? 2018-03-17 19:23:42 yeah. 2018-03-17 19:23:51 sorry. typo. 2018-03-17 19:24:04 <_ikke_> kfox1111: There seems to be some kind of way yes 2018-03-17 19:24:23 <_ikke_> ~/.mkimage can contain plugins 2018-03-17 19:25:13 <_ikke_> it looks for files called mkimg.*.sh in there 2018-03-17 19:26:29 <_ikke_> It's expected to have functions in there called 'profile_* {' or 'section_* {' 2018-03-17 19:27:11 <_ikke_> https://git.alpinelinux.org/cgit/aports/tree/scripts/mkimg.base.sh#n274 2018-03-17 19:27:23 the example overrides some env variables. but is it ok to tweak the image as its getting layed down that way? 2018-03-17 19:28:12 <_ikke_> I don't know 2018-03-17 19:28:54 hmm.. apkovl seems to be an image. 2018-03-17 19:31:21 for _script in "$PWD"/"$apkovl" $HOME/.mkimage/$apkovl \ 2018-03-17 19:31:31 that looks promising. 2018-03-17 19:38:07 so, it looks like I may need to generate a custom apkovl image? 2018-03-17 19:39:57 <_ikke_> I think so 2018-03-17 19:40:08 <_ikke_> You can use lbu to do that 2018-03-17 19:44:03 ok. thanks. 2018-03-17 20:05:45 Hey, I've been trying to get the armhf image to run (on a BB-xM); is there any documentation on it? 2018-03-17 20:19:13 I couldn't get the MLO/u-boot that comes with it to work at all, but judging from the folder name it's probably for the bb black, so I instead used the ones that come with the official image. Only I can't seem to get it to boot the Alpine kernel 2018-03-17 20:19:56 Has anyone tried running on the xM before? 2018-03-17 20:23:54 Orochimarufan: whar is xM 2018-03-17 20:24:08 s/whar/what/ 2018-03-17 20:24:17 https://beagleboard.org/beagleboard-xm 2018-03-17 20:24:37 the followup design of the original BeagleBoard 2018-03-17 20:26:03 is it arm32 or arm64 2018-03-17 20:26:17 32 2018-03-17 20:26:28 which SOC 2018-03-17 20:26:35 OMAP3 2018-03-17 20:26:48 Alpine should work 2018-03-17 20:27:12 I didn't tried on OMAP3 though 2018-03-17 20:27:26 Yes, it's even shipping the dtb, but I'm not sure how to get it to boot 2018-03-17 20:27:58 you should find proper u-boot for the board 2018-03-17 20:28:12 Where do I look? 2018-03-17 20:28:15 install it and see if the u-boot works 2018-03-17 20:28:44 Well, I have the stock u-boot from the bb site and it works 2018-03-17 20:28:44 I don't know for OMAP boards because I don't have any of them 2018-03-17 20:28:58 ah, so far so goog 2018-03-17 20:29:01 but it won't boot the kernel 2018-03-17 20:29:14 s/goog/good/ 2018-03-17 20:29:22 (Though I'm not even entirely sure how to configure it correctly) 2018-03-17 20:29:47 bah, crapy new keyboard on my side 2018-03-17 20:31:44 to repeat myself, I don't have experience with OMAP boards, but on some other which I tried it works by making boot partition and install kernel image and initram on that partition 2018-03-17 20:32:49 I think there is guide on Alpine wiki which describes proccess for the A20 boards 2018-03-17 20:33:57 https://wiki.alpinelinux.org/wiki/DIY_Fully_working_Alpine_Linux_for_Allwinner_and_Other_ARM_SOCs 2018-03-17 20:34:14 It *should* work, according to https://wiki.alpinelinux.org/wiki/Alpine_on_ARM 2018-03-17 20:34:22 hmm 2018-03-17 20:34:31 Are you booting with the right DTB from Das U-boot? 2018-03-17 20:35:00 It might be called something different in ,mainline from the beaglebone kernel 2018-03-17 20:51:16 My U-Boot doesn't seem to want to read extlinux.conf 2018-03-17 20:51:38 Can u-boot even read extlinux config files? 2018-03-17 20:51:48 it should 2018-03-17 20:52:31 https://wiki.alpinelinux.org/wiki/DIY_Fully_working_Alpine_Linux_for_Allwinner_and_Other_ARM_SOCs#Boot-loaders_aka_U-Boot 2018-03-17 20:54:29 But I can't find any documentation that has specifics 2018-03-17 20:55:15 make /extlinux/extlinux.conf on the boot partition 2018-03-17 20:55:32 and mark boot partition bootable 2018-03-17 20:55:48 If you didn't already 2018-03-17 20:56:35 is musl supposed to put the symbol "puts" in the .rela.dyn section of a dso as if the symbol exists in the dso itself 2018-03-17 20:56:38 000000200fc8 000100000006 R_X86_64_GLOB_DAT 0000000000000000 puts + 0 2018-03-17 20:56:51 that way it works on few my boards 2018-03-17 20:59:39 as if the symbol "puts" is not in the PLT/got of hello.so then how does it know if the symbol is meant to be externally retrieved from libc.so itself instead of hello.so 2018-03-17 21:00:43 void* handle = dlopen("./hello.so", RTLD_LAZY); 2018-03-17 21:00:44 dlerror(); 2018-03-17 21:00:46 hello_t hello = (hello_t) dlsym(handle, "hello"); 2018-03-17 21:01:29 The trick was marking both partitions bootable 2018-03-17 21:01:46 But now it's stuck on "Starting kernel..." 2018-03-17 21:03:09 does it use a symbol value of 0000000000000000 in replacement of the PLT/GOT table? (.plt) 2018-03-17 21:04:27 like Sym. Value 0000000000000000 == *_JUMP_SLOT 2018-03-17 21:04:29 also, the dtb seems to be the correct one (omap3-beagle-xm.dtb) 2018-03-17 21:05:13 == R_$ARCH_JUMP_SLOT * 2018-03-17 21:07:45 Any more ideas? 2018-03-17 21:11:46 i have alpine installed and working on my raspberry pi 3b. the SD card does not boot on the new raspberry pi 3b+. does anyone know what i need to do to get it to boot on the new hardware? 2018-03-17 21:21:25 Orochimarufan: is your console video card or serial interface 2018-03-17 21:21:30 serial 2018-03-17 21:22:17 did you put console=ttyS0 (or is it ttyACM0 for that board) in extlinux.conf 2018-03-17 21:22:42 it should be ttyO2 2018-03-17 21:23:10 but you're right, SOMETHING is happening on the display 2018-03-17 21:23:31 are you sure? to me it looks strange name for tty 2018-03-17 21:23:46 some omap special case apparently 2018-03-17 21:23:53 ah, ok then 2018-03-17 21:24:11 but you mean 02 and not O2, right 2018-03-17 21:24:25 nah, Oh Two 2018-03-17 21:24:52 heh, really strange, but ok if it work 2018-03-17 21:27:31 Fun! it depends on the kernel driver which one to use :( 2018-03-17 21:27:32 https://github.com/torvalds/linux/blob/master/drivers/tty/serial/8250/Kconfig#L384 2018-03-17 21:30:21 soo... maybe ttyS2 then? 2018-03-17 21:31:09 really, really strange. I still think it is 0 (zero) and not O (ou) but till I see by my eyes I can't be sure 2018-03-17 21:31:32 yes, try ttyS2 2018-03-17 21:32:20 doesn't work either 2018-03-17 21:32:25 I think it is ttyS2 on my arm64, but not sure and cannot look right now 2018-03-17 21:33:31 wait a moment, I will power up my arm board which is near here to see parameters and show to you 2018-03-17 21:35:33 Also, Alpine has CONFIG_SERIAL_8250_OMAP set to m 2018-03-17 21:36:02 can you even output to a module in early boot? 2018-03-17 21:37:13 ah, well, if the module is in initram it could after module is loaded 2018-03-17 21:39:16 and, in my experience, it is better to build your own kernel image and modules than use stock from distribution 2018-03-17 21:40:44 Sure doesn't look like it ships 2018-03-17 21:41:15 I kind of wanted to avoid going down the rabbit hole of compiling a custom kernel 2018-03-17 21:44:40 then you can extract kernel, initram and modules from some other distro which supports your board 2018-03-17 21:45:58 I did it for lamobo R1 to test it's ethernet switch downloading Armbian and put it's kernel and modules to the Alpine installed on the disk 2018-03-17 21:46:30 I'll just take the kernel from the official image 2018-03-17 21:46:50 that sounds good idea 2018-03-17 21:47:37 la_putin_: This is a bad question. You're providing absolutely no context whatsoever. Go and ask on #musl, but reformulate your question so that people can answer it. 2018-03-17 21:48:46 I am having a hard time figuring out how to customize this iso within the iso image. 2018-03-17 21:49:15 the image seems to think it is a shell script. 2018-03-17 21:49:22 but its really unclear what should go into the script. 2018-03-17 22:03:02 duncan^: https://paste.pound-python.org/raw/pYCXCQ3mCQKJndQXIPi8/ 2018-03-17 22:03:51 la_putin_: You were already told why it wouldn't work, it was to do with the way you're specifying fpic and stuff 2018-03-17 22:03:58 la_putin_: You're aksing this question over and over 2018-03-17 22:04:54 ik but does musl use sym value 0 in placement of R_x86_64_JUMP_SLOT 2018-03-17 22:05:26 ask on #musl 2018-03-17 22:05:55 ...which you were banned from for asking awful questions 2018-03-17 22:06:51 as compiling exactly the same on gnu produces https://paste.pound-python.org/show/b7UVSWo0LDC6zSyVciej/ 2018-03-17 22:11:21 is there a difference between DT_JMPREL and SHT_DYNSYM ? 2018-03-17 22:12:19 as both appear to hold the same entries but SHT_DYNSYM also appears to hold a few more then DT_JMPREL does 2018-03-17 22:20:32 for example, aswell as these symbols: puts, __cxa_finalize, __deregister_frame_inf, _ITM_registerTMCloneTa, _ITM_deregisterTMClone, _Jv_RegisterClasses, and __register_frame_info, it also has _edata, _end, _init, __bss_start, hello, and _fini, in which those have values that are non 0 2018-03-17 22:26:38 All right, I got output from kernel now 2018-03-17 22:27:29 duncan^, la_putin_ is banned from #musl for this 2018-03-17 22:31:32 mps: Now it's complaining that it doesn't have a rootfs. Any ideas? 2018-03-17 22:32:54 Orochimarufan: what filesystem you create on root partition 2018-03-17 22:33:14 I mean, fs type 2018-03-17 22:33:20 ext4 2018-03-17 22:34:11 but doesn't alpine do something fancy with ramdisks or something? 2018-03-17 22:34:53 no, if you put kernel parameter to boot from partition 2018-03-17 22:35:08 That doesn't seem to be the default tho 2018-03-17 22:35:51 I don't have initrd in extlinux.conf 2018-03-17 22:37:19 APPEND root=/dev/mmcblk0p2 rw rootwait console=ttyS0,115200 2018-03-17 22:37:34 I see 2018-03-17 22:37:53 and no initram, just bare Alpine on that partition 2018-03-17 22:38:00 there's nothing on that partition though 2018-03-17 22:38:32 you didn't put Alpine on the root partition? 2018-03-17 22:38:57 I only got the arm image from the downloads section 2018-03-17 22:39:11 tar.gz ? 2018-03-17 22:39:22 yeah 2018-03-17 22:39:24 Orochimarufan: For BeagleBone xM, it will looj either for fat32 (or ext4?) partition marked as bootable or for u-boot spl at the start of the partition table (IIRC) 2018-03-17 22:39:58 try getting uart to see what u-boot is doing 2018-03-17 22:40:20 duncan^: Already got into the kernel. Now just trying to figure out Alpine stuff 2018-03-17 22:41:00 mps: there's only boot and apks on my "root" partition 2018-03-17 22:41:00 there are two armhf files to download, 'Mini root filesystem' and 'Generic ARM' 2018-03-17 22:41:10 I see 2018-03-17 22:41:46 I can't remember which one I used to create root fs 2018-03-17 22:42:04 So I take it you don't know how the whole diskless thing works? 2018-03-17 22:42:08 but, one of them for sure :) 2018-03-17 22:42:47 well, few months ago I did it with qemu VM 2018-03-17 22:43:04 but forgot, and didn't wrote docs about it 2018-03-17 22:43:37 but, it could be done for sure 2018-03-17 22:44:13 you just need to unpack downloaded tar.gz on the root filesystem 2018-03-17 22:44:40 i.e. on your main machine do: 2018-03-17 22:44:57 Orochimarufan: its not a "root" partition then, but "boot" 2018-03-17 22:45:07 and essentially carries the semantics of /boot 2018-03-17 22:45:27 isn't orochimaru the snake guy from Naruto? 2018-03-17 22:46:24 linxon: True enough I guess. Though there's a second "boot" partition with just the board MLO and u-boot. And yes. 2018-03-17 22:46:33 i guess you meant me 2018-03-17 22:46:40 oh yeah sorry 2018-03-17 22:46:46 you should unpack alpine-minirootfs on your root partition 2018-03-17 22:48:44 got it. copied the dhcp example and tweaked it. 2018-03-17 23:01:16 the minimal rootfs isn't very complete it seems 2018-03-17 23:01:42 complete for what? 2018-03-17 23:01:47 it is called minimal for reason 2018-03-17 23:02:02 enough to run apk update; apk add alpine-base 2018-03-17 23:02:40 hmm 2018-03-17 23:02:52 I suspect I know what I'm missing 2018-03-17 23:03:16 i guess its openrc 2018-03-17 23:04:18 it is 2018-03-17 23:04:40 which makes the kernel complain and not do anything useful 2018-03-17 23:05:46 Orochimarufan: which device do you have? 2018-03-17 23:05:59 usually one can do the rootfs setup with setup-alpine 2018-03-17 23:06:06 BeagleBoard-xM 2018-03-17 23:10:40 Guess I could have just chrooted in... oh well 2018-03-17 23:13:18 scratch that, forgot it's a different arch 2018-03-17 23:34:33 do most of you use alpine for just docker? 2018-03-17 23:36:24 ACTION mostly doesn't use docker 2018-03-17 23:36:55 ACTION only uses it for desktop 2018-03-17 23:37:35 phpdave11 when possible 2018-03-17 23:37:43 i use it on servers and embedded devices, i have a mips fork i'm working a little on, but haven't had much time 2018-03-17 23:40:03 phpdave11: i run it as desktop and server 2018-03-17 23:41:17 phpdave11: like liwakura I run it on desktop/notebooks and server 2018-03-17 23:44:55 i'm going to use it as a source-based distro on powerpc, but i'm not there yet 2018-03-17 23:54:51 there is something odd happening 2018-03-17 23:55:12 id4rk: what is it? 2018-03-17 23:58:58 installed alpine-linux on this laptop (i686_standard) on the apu (x86_64_extended) and on the laptop of my gf (x86_64_standard),installed tmux on all 3, and installed tmux-config (custom script) and right out of the box only on my gfs laptop the status bar is on the top 2018-03-18 00:00:10 also, when i put the set-option statusbar top it stays on the bottom 2018-03-18 00:00:59 are there some kind of version differences in the repositories? 2018-03-18 02:27:18 So.. I finally got it working :) 2018-03-18 02:45:34 you are looking for trouble, young man 2018-03-18 02:45:44 installing Alpine on your girlfriend's computer 2018-03-18 02:46:08 unless she's a Linux hacker, that move spells disaster! 2018-03-18 02:47:25 (let's see just how many details I got wrong...) 2018-03-18 02:52:16 :) 2018-03-18 02:52:37 the only winning move is not to install 2018-03-18 02:52:56 yhankfully alpine makes it easy to use that way :) 2018-03-18 03:51:49 lel assuming females no nothing about technology is never a good idea 2018-03-18 03:51:55 know* 2018-03-18 04:13:30 I cannot get Oracle java to run, I installed it following the Installing_Oracle_Java wiki page, and when I go to /opt/java/jre8.0_161/bin, and type in "./java" ash tells me "./java: not found" What am I doing wrong? 2018-03-18 04:18:59 ldd ./java 2018-03-18 04:19:29 then "file ./java | grep interp" 2018-03-18 04:20:09 also does musl support runtime linker auditing? 2018-03-18 04:20:42 if so is runtime linker auditing considered portable/standard? 2018-03-18 04:36:08 @la_putin_ ldd: /java: Not a valid dynamic program 2018-03-18 04:38:47 petek, you need the glibc compat package (gcompat iirc) to have any chance of it working 2018-03-18 04:39:03 dalias: oracle releases native musl builds now 2018-03-18 04:39:09 oh? 2018-03-18 04:39:12 yep 2018-03-18 04:39:16 wow 2018-03-18 04:39:21 for alpine explicitly 2018-03-18 04:39:26 then i dont know 2018-03-18 04:39:40 its possibly not that version that PeteK has 2018-03-18 04:39:48 ah 2018-03-18 04:40:04 also might be 32/64 bit mismatch 2018-03-18 04:40:37 @la_putin_ ./java: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.9, BuildID[sha1]=c90f19ee0af98c47ccaa7181853cfd14867bc931, not stripped 2018-03-18 04:41:58 not a 32/63 mismatch, running 64 bit alpine, downloaded 64 bit java. Did not know about Oracle releasing a musl version of Java. 2018-03-18 04:42:22 http://jdk.java.net/10/ 2018-03-18 04:42:28 it's available here 2018-03-18 04:42:32 early access only still 2018-03-18 04:47:38 does /lib64/ld-linux-x86-64.so.2 itself exist 2018-03-18 04:48:04 ( ls -L /lib64/ld-linux-x86-64.so.2 ) 2018-03-18 04:48:42 no, thats glibc 2018-03-18 04:48:52 or the gcompat stub 2018-03-18 04:49:42 ik but still, it may not exist and may atually be in /lib/ld-linux-x86-64.so.2 instead (rare but possible) 2018-03-18 04:49:55 actually* 2018-03-18 04:56:59 tho u could always try to strace ./java to see if a file is failing to be opened or found 2018-03-18 05:08:14 la_putin_ execve("./java", ["./java"], 0x7e0fb8744e50 /* 13 vars */) = -1 ENOENT (No such file or directory) 2018-03-18 05:08:14 writev(2, [{iov_base="strace: exec: No such file or di"..., iov_len=39}, {iov_base="\n", iov_len=1}], 2strace: exec: No such file or directory 2018-03-18 05:08:14 ) = 40 2018-03-18 05:08:14 writev(2, [{iov_base="", iov_len=0}, {iov_base=NULL, iov_len=0}], 2) = 0 2018-03-18 05:08:14 getpid() = 3719 2018-03-18 05:08:14 exit_group(1) = ? 2018-03-18 05:08:14 +++ exited with 1 +++ 2018-03-18 05:10:41 @la_putin /lib64 does not exist 2018-03-18 05:10:43 wtf 2018-03-18 05:11:17 ln -s /lib /lib64 2018-03-18 05:11:51 (varify that /lib/ld-linux-x86-64.so.2 exists and is 64 bit first tho) 2018-03-18 05:12:25 otherwise check if /usr/lib/ld-linux-x86-64.so.2 exists and is 64 bit 2018-03-18 05:13:27 if /lib doesnt exist but /usr/lib exists AND it is 64 bit the "ln -s /usr/lib /lib ; ln -s /usr/lib /lib64 2018-03-18 05:13:36 then* 2018-03-18 05:13:42 no ld-linux-x86-64.so.2 does not exist either in /lib or /usr/lib 2018-03-18 05:14:02 lib/ld-musl-x86_64.so.1 * 2018-03-18 05:14:21 did you install gcompat? :p 2018-03-18 05:14:42 yes, ld-msl exists in /lib 2018-03-18 05:14:51 is it 64 bit 2018-03-18 05:14:53 @scv no I did not 2018-03-18 05:15:10 well, that's necessary if you're trying to use the non-musl java build 2018-03-18 05:15:16 21:38 < dalias> petek, you need the glibc compat package (gcompat iirc) to have any chance of it working 2018-03-18 05:15:18 @la_putin_ ld-musl-x86_64.so.1 2018-03-18 05:15:51 "non-musl java build" why would a non musl package/build even exist for a distro that is musl based? 2018-03-18 05:16:28 it's not a repo package 2018-03-18 05:16:37 o.o 2018-03-18 05:16:45 have you installed java before la_putin_ ? 2018-03-18 05:16:57 oracle/sun java, not openjdk 2018-03-18 05:17:03 once, never installed it again 2018-03-18 05:17:08 ah 2018-03-18 05:17:39 anyway if it is being compiled then gcc SHOULD link it with musl instead of gnu 2018-03-18 05:18:22 @scv It's not in the Installing_Oracle_Java wiki page, so I did not add it 2018-03-18 05:18:50 unless it specifically overridess the default linker with -Wl,-dynamic-linker,LINKER_PATH 2018-03-18 05:19:11 but even so it would still link with musl's libc itself 2018-03-18 05:19:59 its a binary la_putin_ 2018-03-18 05:20:53 so just mkdir /lib64 ; /usr/lib/ld-musl-x86_64.so.1 /lib64/ld-linux-x86-64.so.2 2018-03-18 05:21:03 scv: so it is pre compiled? 2018-03-18 05:21:06 yes 2018-03-18 05:21:14 then it would not work 2018-03-18 05:21:22 it will with gcompat lol 2018-03-18 05:21:27 how o.o 2018-03-18 05:21:37 I downloaded a tarball from Oracle 2018-03-18 05:21:46 https://code.foxkit.us/adelie/gcompat 2018-03-18 05:21:52 even IF u symlink the linker it STILL needs to use gnu libc instead of musl libc 2018-03-18 05:22:03 read the lin 2018-03-18 05:22:05 link* 2018-03-18 05:22:48 otherwise all applications compiled using glibc can just be used on musl with absolutely no recompilation at all 2018-03-18 05:23:21 unless there statically compiled in which case they will always work regardless 2018-03-18 05:23:32 right 2018-03-18 05:24:45 gcompat (missing): 2018-03-18 05:24:47 required by: world[gcompat] 2018-03-18 05:25:28 PeteK: you should either try the native musl version (http://jdk.java.net/10/) or if you need the specific version you're trying then gcompat (in community repos currently? iirc) is your best bet 2018-03-18 05:26:49 la_putin_ it is in alpine edge community 2018-03-18 05:26:51 0.2 can run... probably 90% of applications 2018-03-18 05:26:59 we are working on closing the gap 2018-03-18 05:27:44 According to the author, gcj will not work. 2018-03-18 05:28:22 how do I pull gcompat from edge without pulling everything else from edge? 2018-03-18 05:29:31 thanks kaniini 2018-03-18 05:30:26 unfortunately you can't really do that safely right now 2018-03-18 05:31:38 shouldnt this just be a single printf? 2018-03-18 05:31:40 printf("This is the gcompat ELF interpreter stub.\n"); 2018-03-18 05:31:41 printf("You are not meant to run this directly.\n"); 2018-03-18 05:32:01 lines 39 and 40 of gcompat/loader/loader.c 2018-03-18 05:32:23 thanks for the unsolicited code review, but it is separate for style reasons 2018-03-18 05:34:25 why 2018-03-18 05:34:48 because I like the way it looks 2018-03-18 05:34:53 lol 2018-03-18 05:35:10 hince "style reasons" 2018-03-18 05:35:35 u could just use \ to move it to a new line without needing a new printf 2018-03-18 05:35:48 you are right. i could. 2018-03-18 05:35:49 but I don't want to. 2018-03-18 05:35:56 ._. rip 2018-03-18 05:35:56 so, I won't. 2018-03-18 05:37:21 that is the beauty of open source software: if it bothers you that much, you can patch it yourself, and release your modified version 2018-03-18 05:37:31 wonder what would happen if u decided to printf every single letter and backslash escape seperately like printf("a"); printf("b"); ... , wonder how long it would take XD 2018-03-18 05:38:13 the time difference, under normal circumstances, would not perceivable by human beings 2018-03-18 05:38:49 unless you are running it on a computer that could not run gcompat anyway 2018-03-18 05:39:47 lel 2018-03-18 05:40:32 like call printf 1 milllion times XD 2018-03-18 05:41:21 better watch out kaniini, gonna have a patch from la_putin_ coming soon replacing all the printfs with individual putcs 2018-03-18 05:41:45 LMFAO 2018-03-18 05:41:50 which will go directly to the trashcan, your point ? 2018-03-18 05:41:55 LD 2018-03-18 05:41:58 :D 2018-03-18 05:41:59 rather 2018-03-18 05:42:28 oh well, at least it's not like the time i had the inspircd source code printed out on toilet paper and used it to wipe my ass for months 2018-03-18 05:42:38 bahaha 2018-03-18 06:17:49 funny how musl dynlink.c contains "do_relocs(p, laddr(p, dyn[DT_JMPREL]), dyn[DT_PLTRELSZ], 2+(dyn[DT_PLTREL]==DT_RELA));" yet there are no DT_JMPREL's in a EXEC/DYN ELF file compiled using musl 2018-03-18 06:28:29 or at least none that i have descovered 2018-03-18 06:28:35 ... 2018-03-18 06:29:24 that would be -z bindnow 2018-03-18 06:33:08 in gcc or in ld 2018-03-18 06:34:24 in alpine's hardened gcc defaults 2018-03-18 06:34:44 nothing to do with musl 2018-03-18 06:35:07 so i would do "gcc -Wl,-z\ bindnow file.c" ? 2018-03-18 06:35:32 or would i do "gcc -z bindnow file.c" ? 2018-03-18 06:37:09 bash-4.4# cd examples/example1/ ; g++ -z bindnow hello.cpp -fPIC -shared -o hello.so 2018-03-18 06:37:11 /usr/lib/gcc/x86_64-alpine-linux-musl/6.4.0/../../../../x86_64-alpine-linux-musl/bin/ld: warning: -z bindnow ignored. 2018-03-18 07:55:45 danieli: i am interested in your mips fork of abuild/alpine. would you mind to share a link? 2018-03-18 08:33:52 danieli: i just saw that the mips triplets are there in upstream abuild, but not yet in a released version 2018-03-18 08:48:14 morning 2018-03-18 08:48:32 i fixed the tmux issue, i dont know what it was, but its working now 2018-03-18 08:48:58 is there gpm for alpine linux? cannot find it in the repositories 2018-03-18 08:49:12 (cli mouse support) 2018-03-18 10:56:36 Hello, when trying to install shadow-doc, I got an error: 2018-03-18 10:56:46 ERROR: shadow-doc-4.5-r0: trying to overwrite usr/share/man/man1/groups.1.gz owned by coreutils-doc-8.29-r2. 2018-03-18 10:56:49 ERROR: shadow-doc-4.5-r0: trying to overwrite usr/share/man/man8/nologin.8.gz owned by util-linux-doc-2.31-r0. 2018-03-18 10:56:55 Should i report it as a bug or not? 2018-03-18 11:12:55 mh, conflict between packages 2018-03-18 11:14:36 Youbi: i think its fine to do 2018-03-18 11:15:02 probably not noticed because installing coreutils and util-linux is less common for alpine linux 2018-03-18 11:15:26 *alpine linux users 2018-03-18 11:58:49 hi 2018-03-18 11:58:59 fabled: do you have some news about this pull request: https://github.com/alpinelinux/aports/pull/2505 ? 2018-03-18 12:00:18 I'm using it on my routers for 6 months now, no issue about it, but it could be cool if it was mainstreamed 2018-03-18 13:42:39 liwakura: ok :) Yeah, coreutils and util-linux installed here because I have it on my desktop machine ^^" 2018-03-18 13:42:55 IMHO not required 2018-03-18 13:43:09 never needed any of those two on my alpine laptops 2018-03-18 13:43:29 :) 2018-03-18 14:19:37 Hi all, I have alpine linux 3.6.2 installed on aws through docker cloudformation template. From what I understand it uses "data" installation type. The problem is - I want to autostart cron on boot and whenever I use rc-service, I'm getting errors like "read-only file system" (https://gist.github.com/maokomioko/64623778b9d7db9e46cf8ee78e861b57). 2018-03-18 14:20:20 <_ikke_> \ 2018-03-18 14:21:11 maokomioko: is sysfs mounted? because it looks like it doesn't 2018-03-18 14:21:29 let me check 2018-03-18 14:21:30 i guess there are many startup jobs missing 2018-03-18 14:21:37 likely not added via rc-update 2018-03-18 14:23:08 There is only /sys/fs/cgroup mounted to tmpfs 2018-03-18 14:23:30 And yes, only docker processes are running on boot, everything else is disabled 2018-03-18 14:24:43 oh, docker 2018-03-18 14:25:36 wait, is it in Docker or a docker host? 2018-03-18 14:26:48 in a host 2018-03-18 14:26:58 >< maokomioko> And yes, only docker processes are running on boot, everything else is disabled 2018-03-18 14:27:12 did you disable the other stuff? 2018-03-18 14:27:44 Nope, that was already there, as a part of amazon image 2018-03-18 14:28:01 can i access that image somehow? 2018-03-18 14:28:13 Mmm. Let me check 2018-03-18 14:28:26 or wait, can you post the output of `ls /etc/runlevels/*` ? 2018-03-18 14:29:15 https://gist.github.com/maokomioko/a7ae6e4c65f5e06e1f4b0da2d183d847 2018-03-18 14:29:48 not sure how sysfs is supposed to get mounted from that situation.. 2018-03-18 14:30:17 no sysfs, no cgroups, and docker wont like not having cgroups 2018-03-18 14:30:55 Added df -h output to that link 2018-03-18 14:32:02 this looks like the inside of an container 2018-03-18 14:34:58 albeit it's a host os. Is it possible to mount sysfs and then run rc service? 2018-03-18 14:35:12 I've been playing around with ACF; is there a way to manage the system (i.e. not ACF) users? 2018-03-18 14:35:43 maokomioko: its not a host, / is an overlayfs, with external fragments of an actual hard-disk mounted in 2018-03-18 14:36:49 especially the docker.sock mounted in from above gives it away.. 2018-03-18 14:37:47 Hmm. 2018-03-18 14:38:04 the error you do get, "non-writable file system" comes when trying to interact with the file system 2018-03-18 14:38:29 Because it's mounted into memory, right? 2018-03-18 14:38:32 so you should probably look why its not writable when it should 2018-03-18 14:38:35 no idea 2018-03-18 14:38:46 df -h doesn't give the mountflags 2018-03-18 14:38:50 maybe try findmnt ? 2018-03-18 14:39:20 alternatively `cat /proc/self/mountinfo` because findmnt wont be installed per default 2018-03-18 14:40:12 Added findmnt, https://gist.github.com/maokomioko/50a0041fca3295c8c1738a9991fa98eb 2018-03-18 14:41:15 i see, indeed, quite alot of stuff is mounted read-only 2018-03-18 14:42:36 i dont know why, its not alpine who does this 2018-03-18 14:44:48 Where could the mount config be? I was thinking of changing mount options and checking it they can persist after reboot 2018-03-18 14:46:07 probably outside of the container 2018-03-18 14:46:22 is that kubernetes? or some coreOS stuff? 2018-03-18 14:47:08 moby (customized alpine from docker). This particular node acts as docker swarm manager 2018-03-18 14:47:24 you could try to use alpine with runit 2018-03-18 14:47:37 or maybe disable cgroup functionality on openrc 2018-03-18 14:47:43 but i dont know how to do the latter 2018-03-18 14:48:49 I'll try both. Hopefully something will work 2018-03-18 15:17:45 2 2018-03-18 15:27:03 jn__: it's not finalized nor even close to ready for production 2018-03-18 15:27:09 i've gotten it to boot, with some minor issues 2018-03-18 15:50:43 hello 2018-03-18 15:53:13 is anyone using alpline linux on pine64+? 2018-03-18 15:55:30 tmcm: just ask right away 2018-03-18 15:55:42 there are chances your problem is not specific to pine64 anyways 2018-03-18 15:56:27 My home server keeps halting unexpectedly, cannot find any pointers in /var/log/messages. Can someone tell me how to increase the kernel loglevel and change the location of the kernel logfile? 2018-03-18 15:56:42 well primarily, i'm looking for the most appropriate installation document 2018-03-18 15:56:57 i admit that i'm not clear how to use the aarch64 tarball to install 2018-03-18 15:57:18 e.g., what kinda fs is it supposed to be extracted to on the sd, etc. 2018-03-18 15:57:40 jay-v: if the disk sync fails, the log will likely not contain the relevant stuff.. 2018-03-18 15:58:00 adding "debug" to the kernel command line could do it, tho 2018-03-18 15:58:01 tmcm: The aarch64 tarball is a root filesystem. Does Pine64 have support in Linux kernel mainline? 2018-03-18 15:58:26 checking 2018-03-18 15:58:29 ha 2018-03-18 15:58:32 it's an allwinner chip 2018-03-18 15:58:32 so 2018-03-18 15:58:39 lucky guess wins a cookie 2018-03-18 15:58:39 soc is the allwinner a64 2018-03-18 15:58:39 ;) 2018-03-18 15:58:43 If not, it probably won't boot the tarball as it is. You'll need to add the appropriate kernel and DTB. 2018-03-18 15:58:59 scv: does it need DTB ? 2018-03-18 15:59:08 dunno, i actually just got mine yesterday lol 2018-03-18 15:59:13 All ARM boards need a DTB pretty much 2018-03-18 15:59:14 waiting on a new sd card so i can play with it 2018-03-18 15:59:15 it is standard 2018-03-18 15:59:16 ok, i've gone through something similar with openbsd on this board 2018-03-18 15:59:35 maybe i can plod through this 2018-03-18 15:59:51 https://linux-sunxi.org/Pine64 2018-03-18 16:00:01 This page will tell you what to do. 2018-03-18 16:00:19 fwiw, i hate these stupid boards 2018-03-18 16:00:33 but i feel obligated to make some use of them 2018-03-18 16:00:45 i like the idea of the sopine but i have yet to actually do anything with one of em 2018-03-18 16:00:51 Why do you feel obligated to make use of these single board computers? 2018-03-18 16:01:03 form factor is super neat though 2018-03-18 16:01:09 they were given to me. there's a reason they only cost $15 or $25 or whatever 2018-03-18 16:01:17 i have 3 raspberry pi's here 2018-03-18 16:01:22 all dusty 2018-03-18 16:01:28 cool to play with, but never found any use 2018-03-18 16:01:31 ACTION uses the macchiatobin with Linux kernel mainline 2018-03-18 16:01:36 at least they are alpine-ized 2018-03-18 16:02:13 for me it's the ease of running a bunch of them if i wanted to play with distributed services or clustering or w/e 2018-03-18 16:02:15 https://www.cnx-software.com/wp-content/uploads/2017/08/PINE64-SoPine-Cluster-Board-Large.jpg 2018-03-18 16:02:48 liwakura: thanks any other pointers in finding out why the system halts? 2018-03-18 16:03:02 jay-v: do you have access to the kprint logs at runtime? 2018-03-18 16:03:08 external monitor or serial console? 2018-03-18 16:03:12 setup a netconsole 2018-03-18 16:04:38 I hooked up an external monitor 2018-03-18 16:05:03 jay-v: i think thats the way to go 2018-03-18 16:05:15 scv: noticed netconsole docs are removed from ArchWiki 2018-03-18 16:05:27 i guess because poettering didn't put support for it into systemd lol 2018-03-18 16:05:36 wut 2018-03-18 16:05:56 its a kernel module, it shouldn't have anything to do with systemd ?! 2018-03-18 16:06:08 >it shouldn't have anything to do with systemd ?! 2018-03-18 16:06:11 oh boy 2018-03-18 16:06:19 i cry every time 2018-03-18 16:06:24 https://wiki.archlinux.org/index.php/General_troubleshooting#netconsole 2018-03-18 16:06:59 https://wiki.archlinux.org/index.php/Netconsole 2018-03-18 16:07:01 <_< 2018-03-18 16:08:01 its marked for redirection to a site which does not contain netconsole at all 2018-03-18 16:08:23 premium archwiki 2018-03-18 16:08:25 https://www.kernel.org/doc/Documentation/networking/netconsole.txt 2018-03-18 16:09:14 not sure if i trust the display or network subsystems more when in a crash 2018-03-18 16:09:27 i've used it to great success in debugging bizarre panics 2018-03-18 16:11:43 oh man 2018-03-18 16:11:45 https://github.com/apritzel/pine64 2018-03-18 16:11:50 how did i not find that before 2018-03-18 16:11:53 that's a goldmine 2018-03-18 16:12:29 Thanks, I don't have any other system running at home 2018-03-18 16:12:46 But can fetch a rpi to set it up as receiver 2018-03-18 16:12:55 Thanks, netconsole should help :) 2018-03-18 16:13:20 good luck :D 2018-03-18 16:27:48 Hallo, has anyone experience to boot alpine linux arm with uboot? 2018-03-18 16:29:09 gerald: just ask right away 2018-03-18 16:29:36 i installed alpine linux aarch64 on a SD 2018-03-18 16:29:39 i never booted alpine linux with arm & uboot, but your question is likely not specific to alpine linux 2018-03-18 16:29:57 yes, it is not specific 2018-03-18 16:30:09 "installed" as in, extracted a tarball to an partition? 2018-03-18 16:30:41 yes i have an odroid c2 2018-03-18 16:30:53 for that i compiled uboot 2018-03-18 16:31:10 and flashed it to the SD 2018-03-18 16:31:28 before i created a fat partition at block2048 2018-03-18 16:31:49 at extracted the tar there 2018-03-18 16:32:08 uboot starts but don't load the kernel 2018-03-18 16:32:53 odroid c2 has a good mainline kernel support 2018-03-18 16:33:16 so i think the problem is uboot 2018-03-18 16:33:22 "flashed to SD", can you elaborate that? 2018-03-18 16:33:39 on my raspberry pi 2 alpine linux worked great 2018-03-18 16:33:47 i thought uboot was already included in the generic ARM tarball, so im not sure about the self-compilation thing 2018-03-18 16:34:02 with dd 2018-03-18 16:34:20 the primary bootloader to block 0 to 96 2018-03-18 16:34:21 ah, to the beginning, i guess, and thats why the fat at the later block 2018-03-18 16:34:38 and uboot binary to block 97 to 1431 2018-03-18 16:35:00 is there an partition table? 2018-03-18 16:35:36 yes the fat partition at block 2048 2018-03-18 16:36:00 and before the uboot binary 2018-03-18 16:36:26 yes i saw a uboot binary in the tarbal 2018-03-18 16:36:43 but i don't know how to use it 2018-03-18 16:36:49 did you follow some guide or something? 2018-03-18 16:37:03 i followed the wiki from alpine linux for arm 2018-03-18 16:37:30 and there it is explained to use the uboot binary from the board 2018-03-18 16:38:56 https://wiki.alpinelinux.org/wiki/Alpine_on_ARM 2018-03-18 16:39:13 currently reading it 2018-03-18 16:39:14 n 2018-03-18 16:39:42 does uboot give some kind of verbose output? 2018-03-18 16:39:47 at partitioning it is explained to make the first partition at block 2048 2018-03-18 16:40:02 no nothing only black screen 2018-03-18 16:40:16 but a blue led is on 2018-03-18 16:40:29 that indicates that uboot is running 2018-03-18 16:40:45 if the blue led is blinking then the kernel is loading 2018-03-18 16:40:52 kahiru_: did you get alpine running on your odroid? 2018-03-18 16:40:53 but that don't happens 2018-03-18 16:41:35 uboot from odroid needs also a boot.ini 2018-03-18 16:41:41 in the boot directory 2018-03-18 16:42:12 but i don't know what the uboot environment partition is 2018-03-18 16:42:52 https://wiki.odroid.com/odroid-c2/software/partition_table 2018-03-18 16:43:29 there is the partition table of the odroid 2018-03-18 16:43:38 nah, you made your own, you said 2018-03-18 16:43:57 yes like it is explained here: 2018-03-18 16:44:18 https://wiki.odroid.com/odroid-c2/software/building_u-boot 2018-03-18 16:45:04 can you get a serial console going? 2018-03-18 16:45:37 its my first time with uboot 2018-03-18 16:45:49 i know grub from ubuntu 2018-03-18 16:46:01 i don't know how to get in the serial console 2018-03-18 16:46:17 external adapter 2018-03-18 16:46:18 i is not good documented 2018-03-18 16:46:39 i don't have this 2018-03-18 16:46:55 can i enter to uboot also with usb otg? 2018-03-18 16:47:21 i connected the odroid c2 with hdmi 2018-03-18 16:47:45 i dont think u-boot is able to drive hdmi 2018-03-18 16:48:00 ARM doesn't have BIOS-like firmware like x86 does 2018-03-18 16:48:28 but what is the uboot directory in the tarbal of alpine linux? 2018-03-18 16:48:45 i think you should get an USB to serial adapter, hook it to the odroid so you can at least talk to uboot and read the boot log 2018-03-18 16:49:02 ok 2018-03-18 16:49:35 has kahiru installed alpine linux on the odroid board? 2018-03-18 16:50:04 gerald: he asked once about it 2018-03-18 16:50:18 so i pinged them, maybe they got it working 2018-03-18 16:50:36 ok 2018-03-18 16:51:18 i tried alpine linux on my raspberry pi 2 2018-03-18 16:51:25 it performs really fast 2018-03-18 16:51:29 RPi 2 is also well tested 2018-03-18 16:51:44 and from the tarball i would assume there is good support for it 2018-03-18 16:51:54 yes ... it would be great to see alpine linux on more boards 2018-03-18 16:52:17 if you get it running, you can document it 2018-03-18 16:52:23 alpine linux would be ideal for that low powered boards 2018-03-18 16:52:41 yes ok 2018-03-18 16:52:59 thank you for your help 2018-03-18 16:53:02 i recommend you to get an USB-to-Serial adapter 2018-03-18 16:53:15 i will buy a serial adapter 2018-03-18 16:53:22 debugging embedded issues without serial or JTAG is some hopeless adventure 2018-03-18 16:53:31 yes i see 2018-03-18 16:53:36 maybe a local hackspace has some laying around 2018-03-18 16:53:56 yes i will see 2018-03-18 16:54:31 ok, thankyou bue 2018-03-18 16:54:54 sorry my english 2018-03-18 16:54:58 bye 2018-03-18 16:55:01 bye 2018-03-18 16:55:09 gerald: you could also wait here 2018-03-18 16:55:24 maybe someone else comes online who already got it working 2018-03-18 16:55:35 ok 2018-03-18 16:56:07 i have an Pandora here, which is also uboot and armhf, but i haven't tried much getting it to work... 2018-03-18 16:59:30 ok 2018-03-18 16:59:46 i have also a rock64 2018-03-18 16:59:58 but there it is more complicate 2018-03-18 17:00:17 because the mainline kernel support is still in early stage 2018-03-18 17:00:29 so i must also compile a kernel 2018-03-18 17:00:34 for that 2018-03-18 17:01:04 and uboot is not good documented there 2018-03-18 17:01:48 do you know a good documentation of uboot? 2018-03-18 17:03:03 no 2018-03-18 17:16:09 looking through channel history I see question about Alpine on aarc64 machines, yes it could be installed and works pretty wel for months on my chromebook 2018-03-18 17:16:57 but with custom built kernel 2018-03-18 17:24:04 hallo 2018-03-18 17:24:27 i saw that you installed alpine on aarch64 2018-03-18 17:24:39 on a chromebook 2018-03-18 17:24:54 how you managed to boot? 2018-03-18 17:26:27 gerald: chromebook is specific, I do not use u-boot but put kernel directly to boot partition 2018-03-18 17:27:14 with chromebook boot binary must be signed with chromebook tools 2018-03-18 17:27:15 ok ... odroid c2 has a good mainline kernel support 2018-03-18 17:27:34 then it should be easy 2018-03-18 17:27:34 so the kernel should work 2018-03-18 17:27:43 of course 2018-03-18 17:27:49 the problem for me is to get uboot work 2018-03-18 17:28:26 I see, but could you get prompt from the u-boot on that board 2018-03-18 17:29:33 someone told me already to buy an adapter for monitoring the output 2018-03-18 17:29:49 good advice, IMHO :) 2018-03-18 17:29:59 i have only connected a monitor with hdmi 2018-03-18 17:30:08 but there is no output for uboot 2018-03-18 17:31:30 then, that is a problem with this u-boot. It should show propmt on the hdmi 2018-03-18 17:32:11 yes, i have only experience with grub and ubuntu 2018-03-18 17:32:23 there it is easier 2018-03-18 17:32:27 you should check docs for that board 2018-03-18 17:32:58 yes i will, i also wait for answers in the forum 2018-03-18 17:33:06 I have few arm boards and on all of them u-boot works with hdmi 2018-03-18 17:33:29 i only looked here if someone already installed alpine linux on odroid c2 2018-03-18 17:34:06 the problem is uboot from odroid 2018-03-18 17:34:24 they turned off the interactive mode 2018-03-18 17:34:34 for uboot 2018-03-18 17:34:47 I didn't because I don't have odroid, but I took a lot of docs about arm from odroid site and all were very usefull 2018-03-18 17:35:20 what boards do you have? 2018-03-18 17:35:27 but, I didn't worked with their boards so I cannot tell anything for sure 2018-03-18 17:35:43 yes the wiki is good 2018-03-18 17:36:00 better as for pine64 2018-03-18 17:36:01 I have exynos, bananapi and two arm64 chromebooks 2018-03-18 17:36:12 i have also a rock64 2018-03-18 17:36:28 and a raspberry pi 2 2018-03-18 17:36:37 there alpine linux works great 2018-03-18 17:36:46 my samsung chromebook plus one is rockchip 3399 2018-03-18 17:36:55 but odroid c2 would have more power and ram 2018-03-18 17:37:23 and you managed to boot alpine linux on the boards? 2018-03-18 17:37:45 yes, I looked at rock64, adn I would by it but it is not available where I live 2018-03-18 17:38:56 and also looked at pine64, but I think they stoped to make it anymore, but I'm not sure is that true 2018-03-18 17:39:20 for now i have android on the rock64 board 2018-03-18 17:39:29 it performs great 2018-03-18 17:39:45 but there are no security patches 2018-03-18 17:39:58 so security is not good in place 2018-03-18 17:40:05 hey guys how do i add something from apk with an extra ./configure flag? 2018-03-18 17:40:08 any easy way 2018-03-18 17:40:18 You mean, rebuild the package. 2018-03-18 17:40:52 This can easily be done: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2018-03-18 17:41:12 i guess, does apk just pull down compiled artifacts? 2018-03-18 17:41:22 a realy great board comes out soon 2018-03-18 17:41:46 https://www.wandboard.org/ 2018-03-18 17:42:06 the wand-pi-8m 2018-03-18 17:42:53 with a nxp i.mx8m quad core processor 2018-03-18 17:43:13 gerald: I don't know, I'm litle tired with arm politics about their GPUs 2018-03-18 17:43:17 What is significant about that? 2018-03-18 17:43:36 yes me too 2018-03-18 17:43:52 the future will be risk-v processors 2018-03-18 17:44:53 I hope so, but till we have to 'play' with arm non open GPU drivers 2018-03-18 17:45:02 yes 2018-03-18 17:45:27 there is already a risc-v board 2018-03-18 17:45:42 but it is about 1000 Euro 2018-03-18 17:45:49 very expensive 2018-03-18 17:46:09 so we have to wait 2018-03-18 17:46:19 yes, but that is the case with every new technology 2018-03-18 17:46:27 1000eur? 2018-03-18 17:46:52 the 60usd one ships next month :D 2018-03-18 17:47:13 this is prereleases mostly for system integratorts to make products around these new CPU 2018-03-18 17:48:08 scv: The 60USD board in question is more like the Arduino. 2018-03-18 17:48:32 $999 2018-03-18 17:48:58 i didn't know they made quad core arduinos 2018-03-18 17:49:08 or am i mixing the two boards up 2018-03-18 17:49:08 :/ 2018-03-18 17:49:11 its from Hi-Five 2018-03-18 17:49:13 It's still a microcontroller-like board 2018-03-18 17:49:19 whatever the core count 2018-03-18 17:49:54 10:38 < mps> and also looked at pine64, but I think they stoped to make it anymore, but I'm not sure is that true 2018-03-18 17:50:06 they have a "lifetime guarantee" policy 2018-03-18 17:50:13 5 years minimum or something 2018-03-18 17:50:31 scv: of course 2018-03-18 17:50:33 https://www.crowdsupply.com/sifive/hifive-unleashed 2018-03-18 17:51:42 it can run linux 2018-03-18 17:51:51 scv: but I found chromebook with less than 1 Kg weight for about 400 euros, with 2400x1600 display 2018-03-18 17:52:13 https://www.techrepublic.com/article/hi-five-unleashed-the-first-linux-capable-risc-v-single-board-computer-is-here/ 2018-03-18 17:53:33 gerald: yes, I read their anounces earlier, that really promising but industry is not driven by best technology but by something else called politics and money 2018-03-18 17:53:35 mps: that seems like a bit of a strange comparison 2018-03-18 17:53:52 does the chromebook have some convenient form factor for its board ? 2018-03-18 17:54:15 yes i know but i have hope 2018-03-18 17:54:23 there was only windows 2018-03-18 17:54:34 now you have open alternatives 2018-03-18 17:54:42 in software 2018-03-18 17:55:09 now it is time that also hardware will get open alternatives 2018-03-18 17:55:22 gerald: yes, I have seen a loot of good technology during decades long work in IT, which failed and mediocre technology succed 2018-03-18 17:55:58 *cough*js*cough* 2018-03-18 17:56:03 scv: 'convenient form factor', what you mean to say 2018-03-18 17:56:25 no, nevermind, i see what you mean now 2018-03-18 17:56:30 i didn't realise there was a pinebook :) 2018-03-18 17:58:19 gerald: when I started to work (about 35 years ago) I thought that the RISC pervail for sure, but - oh how I was wrong 2018-03-18 17:58:39 but, who knows, maybe in 'next' future :) 2018-03-18 17:59:00 we will see 2018-03-18 17:59:19 chromebooks are already a good alternative 2018-03-18 17:59:29 you have to trust google 2018-03-18 18:00:02 well, if you are brave and have some tools you can reflash everything 2018-03-18 18:00:19 even u-boot on flash chip 2018-03-18 18:00:28 but with android apps and soon linux applications running on it chromeos will be a platfrom for all applications 2018-03-18 18:01:07 I run clean linux on three chromebooks 2018-03-18 18:01:37 on one is Debian, one is dual boot - Alpine and Debian, and third is only Alpine 2018-03-18 18:01:48 i run ubuntu 2018-03-18 18:02:20 but with the meltdown and spectre i looked to arm boards 2018-03-18 18:02:39 which are not affected 2018-03-18 18:02:52 I mean, I run three ARM chromebooks, not x86 2018-03-18 18:03:01 that is not always true gerald 2018-03-18 18:03:12 many new ARM chips are affected 2018-03-18 18:03:18 yes 2018-03-18 18:03:29 anything with simpler in-order exec is ok though 2018-03-18 18:03:30 so i bought the rock64 2018-03-18 18:03:42 arm A53 is not affected 2018-03-18 18:03:50 thankfully :) 2018-03-18 18:03:58 so odroid c2 and rock64 have them 2018-03-18 18:04:44 but there android has no security patches 2018-03-18 18:04:58 and debian/ubuntu are slow 2018-03-18 18:05:06 so i want to try alpine linux 2018-03-18 18:05:17 are android devices affected by spectre 2018-03-18 18:05:27 yes 2018-03-18 18:05:45 maybe some high-end but most are not, AFAIK 2018-03-18 18:05:45 only arm a53 and arm a7 are not 2018-03-18 18:06:25 So, my server keeps crashing. Does this look like a broken SSD to you? https://pastebin.com/raw/QZtQGXUN 2018-03-18 18:06:42 a8, a9, a15, a17, a57, a72, a73 and a75 are affected 2018-03-18 18:07:37 jay-v: could be :( 2018-03-18 18:07:55 oof 2018-03-18 18:08:04 if possible check SMART counters on the disk? 2018-03-18 18:08:17 should see stats for errors, reallocated blocks etc 2018-03-18 18:08:29 looks pretty hosed though :( 2018-03-18 18:10:05 gerald: I have to go, best regards :) 2018-03-18 18:10:54 thank you 2018-03-18 18:10:56 i too 2018-03-18 18:11:03 best regards too 2018-03-18 18:13:49 No errors to be seen in smart stats unfortunatelu 2018-03-18 18:33:05 Could be the controller or cable too, but yes 2018-03-18 18:33:29 That's definitely a hardware issue with the disk subsystem 2018-03-18 18:57:59 hello. 2018-03-18 18:58:38 /proc/interrupts shows interrupt counters for various devices. does anyone know, where i can see how many system calls occured (inside /proc/?)? 2018-03-19 00:17:02 welp i think i finally got the PLT relocation correct for musl compiled applications but now i get into a lookup loop ._. puts>__cxa_finalize>__deregister_frame_info>puts>... https://paste.pound-python.org/raw/09J1mBpmLC0iMiL1ttTz/ 2018-03-19 01:49:57 trying to create a custom iso... there are no kernel modules in /lib/modules 2018-03-19 01:50:03 is there something I need to enable? 2018-03-19 03:23:52 where can i aquire execinfo.h in alpine 2018-03-19 03:24:14 so i can use backtrace() and backtrace_symbols() 2018-03-19 03:32:49 backtrace.c:18:22: fatal error: execinfo.h: No such file or directory 2018-03-19 03:32:51 #include 2018-03-19 03:52:35 la_putin: https://pkgs.alpinelinux.org/contents?file=execinfo.h&path=&name=&branch=edge 2018-03-19 03:52:38 use the package search 2018-03-19 03:59:31 even if i manually include the definitions from the glibc execinfo.h musl complains that backtrace() and backtrace_symbols() are undefined ;-; 2018-03-19 04:00:07 ./files/backtrace.so: In function `bt': 2018-03-19 04:00:08 /loader/backtrace.c:53: undefined reference to `backtrace' 2018-03-19 04:01:36 https://paste.pound-python.org/raw/xzSINJZEaT5kAeEewggK/ 2018-03-19 05:07:24 that's not what I told you 2018-03-19 05:07:26 and of course that wouldn't work 2018-03-19 05:07:45 you want libexecinfo 2018-03-19 06:31:28 ok 2018-03-19 06:32:37 /usr/lib/gcc/x86_64-alpine-linux-musl/6.4.0/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find -llibexeinfo 2018-03-19 06:32:44 /usr/lib/gcc/x86_64-alpine-linux-musl/6.4.0/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find -lexeinfo 2018-03-19 06:33:06 /usr/lib/gcc/x86_64-alpine-linux-musl/6.4.0/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find -lexecinfo 2018-03-19 06:33:18 and so on 2018-03-19 06:55:34 did you install libexecinfo-dev 2018-03-19 09:02:04 no o.o 2018-03-19 12:06:48 Guys, which package is needed for launch VLC with gui? 2018-03-19 12:07:42 I installed vlc-qt but woes about: "no suitable interface module" 2018-03-19 12:20:39 Hello! 2018-03-19 12:32:11 Hello 2018-03-19 13:35:16 What is the default framebuffer driver alpine (vm flavor) is using? 2018-03-19 13:36:22 Can it be configured via some boot parameter to start with a higher resolution? It defaults to a rather small one. I know SSH is the way, but for quick fiddling it would be nice to have a little more space. :P 2018-03-19 13:36:44 Anyone else getting 503's for http://dl-cdn.alpinelinux.org/alpine/v3.6/main/x86_64/APKINDEX.tar.gz ? 2018-03-19 13:41:44 hello, Are you aware of the varnish 503 on the pkgs cdn? 2018-03-19 13:42:06

Details: cache-ams4145-AMS 1521466915 165672672

2018-03-19 13:45:17 morning 2018-03-19 13:46:55 Trunet2, I was wondering if I were the only one getting those - thanks for clarifying. 2018-03-19 13:47:48 Hi all, I am currently unable to install any package in Alpine 3.7 2018-03-19 13:48:09 Not sure if it was discussed earlier (I joined 5 minutes before you did... and IRC logs -_-) ... not sure who to contact or if anyone is on it. 2018-03-19 13:48:27 bobotux, the alpine cdn is down right now 2018-03-19 13:48:47 this is SOOO bad... all CI/CD failing here. 2018-03-19 13:48:59 ok thanks awippler 2018-03-19 13:49:29 I was hoping this was only a problem on my side 2018-03-19 13:50:28 is anybody getting error 503 Backend unhealthy from the cdn? 2018-03-19 13:50:36 yes, we are 2018-03-19 13:50:38 Does anybody now when the cdn will be up again? 2018-03-19 13:50:41 this looks like an Europe problem 2018-03-19 13:50:50 I tried some US/Canada mirrors and they are working 2018-03-19 13:51:05 i'm actually in Canada getting this error 2018-03-19 13:51:18 curl http://mirror.csclub.uwaterloo.ca/alpine/v3.6/main/x86_64/APKINDEX.tar.gz 2018-03-19 13:51:21 this mirror is working 2018-03-19 13:52:02 thanks 2018-03-19 13:52:23 http://mirror.leaseweb.com/alpine/ 2018-03-19 13:52:26 this is also working 2018-03-19 13:53:03 it is ok now 2018-03-19 13:53:06 had the same problem for 15 minutes, it's working again (EU) 2018-03-19 13:53:11 curious - would it be `apk --repo http://mirror.leaseweb.com/` ? 2018-03-19 13:53:39 My build process just worked 2018-03-19 13:53:49 Varnish probably fixed itself 2018-03-19 13:54:21 well, looks like varnish was working just fine, problem was the backend unreacheable 2018-03-19 13:54:24 Yep 2018-03-19 13:54:30 how long was it down approximately? 2018-03-19 13:54:36 10-15 minutes 2018-03-19 13:55:02 It worked at :20 2018-03-19 13:55:34 then did not work at :30 2018-03-19 13:55:39 gitlab-ci builds started failing 35minutes ago 2018-03-19 13:55:49 So I am thinking 30-40 minutes 2018-03-19 13:57:00 Well, for those affected, consider telling your company to do: https://wiki.alpinelinux.org/wiki/How_to_setup_a_Alpine_Linux_mirror 2018-03-19 13:57:22 or donate to the project 2018-03-19 13:57:23 thanks 2018-03-19 14:03:19 @awippler, how to donate? 2018-03-19 14:04:27 There are many ways: https://wiki.alpinelinux.org/wiki/Alpine_Linux:Contribute 2018-03-19 14:05:05 will ask our company to do it 2018-03-19 14:05:08 For money, I think they want funds to go to https://www.patreon.com/musl 2018-03-19 15:25:36 i have no clue where money donated to alpine goes, if any 2018-03-19 15:27:46 https://wiki.archlinux.org/index.php/Netconsole ? 2018-03-19 15:28:10 whoops 2018-03-19 15:28:17 awippler, funds going to musl patreon go to me, helping keep it practical for me to spend maintenance and development effort on project direction independent of what consulting clients might want 2018-03-19 15:28:49 dalias: thanks for autentication 2018-03-19 15:28:50 if alpine is recommending ppl make pledges there, that's very nice :) 2018-03-19 15:29:57 if the amount increases significantly i might also make some one-time or even ongoing donations to major musl contributors in thanks 2018-03-19 17:57:03 Is there some way to depend on a device in openrc? 2018-03-19 18:21:49 Orochimarufan: yes 2018-03-19 18:38:36 Hello together, I search for a way to distribute the team members ssh keys to our different alpine hosts (now more than 50). How do you do this? 2018-03-19 18:42:16 bernhardgruen: i have an ssh CA 2018-03-19 18:42:29 one key to bind them, one key to rule them all 2018-03-19 18:42:37 the server all trust the CA 2018-03-19 18:42:51 and adding a user is just a matter of creating a new certificate 2018-03-19 18:43:35 sounds nice. Never did that before but read about it. What if we have different teams that shouldn't have access to all systems but only their systems? 2018-03-19 18:43:56 multiple CA's, i guess 2018-03-19 18:44:12 liwakura, do you know if a root ca and several intermediate CA are possible? 2018-03-19 18:44:26 afaik its not 2018-03-19 18:44:51 that would solve my problem. 2018-03-19 18:45:20 how so? 2018-03-19 18:45:40 you can create an CA for each team, and if a team should be allowed to access an machine, you add the CA to the sshd config 2018-03-19 18:46:00 so you at least save editing the authorized_keys file for every user 2018-03-19 18:46:36 The root CA is only used by me (as it is my company I most likely don't quit) and the intermediate certificates are the team members. 2018-03-19 18:47:09 the SSH certificates are not transitive, you cant make chains 2018-03-19 18:47:38 with multiple CA i mean, multiple keys 2018-03-19 18:48:14 alternatively, you could use AuthorizedKeysCommand to fetch the list of allowed public keys from a central location 2018-03-19 18:48:37 Anyway I will look at it that could a nice addition for our SFTP systems. I assume this CA thing can be used additionally to normal keys. 2018-03-19 18:48:45 yes. 2018-03-19 18:49:18 i use it because i rotate my laptop's keys often, and the CA saves me from going to all servers and edit some files 2018-03-19 18:49:19 The AuthorizedKeysCommand is not safe as it is prone to DNS attacks. 2018-03-19 18:49:32 you could use TLS key pinning there 2018-03-19 18:49:55 Oh - nice idea. wow 2018-03-19 18:50:12 whole new possibilities. 2018-03-19 18:59:59 liwakura, one more question: how to revoke a certificate in case an employee leaves the company? 2018-03-19 19:01:45 update the revocation list, but then you need to distribute the updated revocation list 2018-03-19 19:02:54 if you time-limit the certificate validity for like an month, you might be able just by not creating a new certificate 2018-03-19 19:03:31 Yes - that would be even nicer. 2018-03-19 19:03:41 or restricting the certificate until exactly the last day of work of the employee 2018-03-19 19:04:00 but that depends on how well your users will like the certificate renewal 2018-03-19 19:05:56 I assume they will not like it that much. And they could also add their key directly to the systems. 2018-03-19 19:07:10 do they have root? 2018-03-19 19:08:16 yes - or at least they can use sudo. 2018-03-19 19:09:10 so they could install a large variety of backdoors anyways or vandalize the systems 2018-03-19 19:09:32 but thats some implicit risk of having employees 2018-03-19 19:10:12 this is right. But they need to have some root privileges as they are DevOps and I don't want to update each and every system by myself. 2018-03-19 19:22:13 bernhardgruen: saw this the other day, looks neat https://github.com/ierror/ssh-permit-a38 2018-03-19 19:24:26 danieli, thank you. I will look at it too (a38 looks like the one from asterix) 2018-03-19 19:24:35 bernhardgruen: that's where it's from! 2018-03-19 19:24:46 permit A 38 2018-03-19 19:24:55 i thought it was some strange spec until i recalled that 2018-03-19 19:25:07 (it's in the readme) 2018-03-19 19:28:29 danieli, the demo video already looks quite nice to me 2018-03-19 19:29:07 bernhardgruen: you could lab it and give it a spin 2018-03-19 19:32:34 yes - coincidentally the author has the same first name. 2018-03-19 19:51:28 Anyone knows which packages are needed to launch VLC with gui? 2018-03-19 19:52:08 I installed vlc-qt and vlc-plugins-gui but I got: "no suitable interface module" 2018-03-19 19:57:19 Yay, finished turning my beagleboard into a simple wifi AP :D 2018-03-19 20:07:59 Hello 2018-03-19 20:08:06 anyone active please? 2018-03-19 20:08:25 <_ikke_> birdbolt1: Usually helps to ask a question 2018-03-19 20:09:58 lol ok 2018-03-19 20:10:03 just, this...https://imgur.com/B0idtNy 2018-03-19 20:10:17 3000x2000 screen on a 13.5 inch device 2018-03-19 20:10:26 The text is unusably small 2018-03-19 20:10:36 How can I increase the size? 2018-03-19 20:11:18 If I install vmware-tools, it simply makes dynamic resizing possible. That in turn will just give a large expanse of teeny tiny text on my screen 2018-03-19 20:15:08 And how do i logout? 2018-03-19 20:15:21 I've just been restarting and logging in as root 2018-03-19 20:15:32 I dont wanna install sudo 2018-03-19 20:26:22 birdbolt1: Take it up wiht vmware, Alpin e takes the size that it thinks the screen is 2018-03-19 20:26:30 this is controlled by your vmware installation 2018-03-19 20:27:05 birdbolt1: To log in/out as someone other than root, of course you have to create a new user, who is not root. 2018-03-19 20:27:19 the 'adduser' command is what you would use, in that case. 2018-03-19 20:28:08 sudo is a useful program, in that it allows you to set permissions and you'll see a list of commands run by a user as root in auth.log. Similarly you'll have a record of the sudo commands run by your individual user. 2018-03-19 20:28:15 This can be useful in certain circumstances. 2018-03-19 20:31:42 well my aim was having a low privilege user, and root, thats it. I am trying out a few commands for a Dockerfile I'm working on. 2018-03-19 20:32:34 Then use adduser as recommended. But don't necessarily dismiss sudo. It is useful. 2018-03-19 20:32:50 duncan^, I understand that, and it's the reason why i wont bother with installing vmwaretools, because that still provides the same pixel density, and doesnt fix my problem 2018-03-19 20:33:16 Take it up with vmware support; I mean, vmware costs enough 2018-03-19 20:33:26 sigh, alright :( 2018-03-20 03:12:58 Hello! 2018-03-20 03:12:58 Would anyone be able to help me run Alpine on a Beaglebone black? 2018-03-20 03:13:32 The documentation on ARM is a little cryptic 2018-03-20 03:19:08 i have alpine running on a bbb 2018-03-20 03:19:22 it's probably mostly a kernel thing 2018-03-20 03:19:31 i'm sure it'll work with armhf 2018-03-20 03:19:38 i built my own armv7 image 2018-03-20 03:19:44 mepholic: Oh good! 2018-03-20 03:21:22 Yeah I have RobertCNelson's bb-kernel built and u-boot built, but not quite sure how to proceed. 2018-03-20 03:21:23 I think I need to init the rootfs somehow 2018-03-20 03:21:52 ethanrjones97[m]: It runs out of the box 2018-03-20 03:21:56 use the standard alpine initrd 2018-03-20 03:22:10 https://wiki.alpinelinux.org/wiki/Alpine_on_ARM 2018-03-20 03:22:17 well 2018-03-20 03:22:28 you may need to build your own with the bb-kernel mods 2018-03-20 03:22:30 idk 2018-03-20 03:22:36 It's not necessary 2018-03-20 03:22:47 Unless you want *blegh* powervr 2018-03-20 03:23:13 mepholic: That's in the tarball? 2018-03-20 03:23:24 the standard initrd? 2018-03-20 03:23:29 idk man I built my own kernel 2018-03-20 03:23:32 and initrd 2018-03-20 03:24:48 duncan^: Yeah I've read over the ARM wiki page but it doesn't tell me much on how to install 2018-03-20 03:52:55 the kernel, initrd, and requisite DTB for your board are in the tarball 2018-03-20 03:53:05 and for all the other ones listed on that wiki page 2018-03-20 03:55:58 duncan^: and the rootfs I just initialize with apk? 2018-03-20 05:00:26 if a symbol cannot be found it is normally appropriate to abort execution right? 2018-03-20 05:01:12 if so would the same be true but if the symbol CAN be found but it is of value 0 and cannot be found in the files dependancies 2018-03-20 05:51:38 la_putin, you cannot run musl programs with a nonmusl dynamic linker. you have been told this countless times. it will appear to work with toys but blow up horribly 2018-03-20 05:52:21 now stop bringing your context-omitted nonsense questions here or i'll ask ops to ban you here too 2018-03-20 05:52:36 and stop doing it on #gdb too 2018-03-20 05:54:58 whats the difference? is it libc specific? is it locked to a specific nlibc and thats it? 2018-03-20 05:55:03 libc* 2018-03-20 05:56:32 if i created my own libc does that mean i cannot use musl's dynamic linker with it? 2018-03-20 05:56:54 or glibc's dynamic linker? 2018-03-20 05:57:15 like it must be my own linker or it will not work no matter what 2018-03-20 06:29:09 yes yes and yes 2018-03-20 06:42:52 ._. 2018-03-20 10:05:00 ok now this is just wtf 2018-03-20 10:31:08 hi 2018-03-20 10:31:25 i get this with edge: ERROR: rsyslog-8.31.0-r0: BAD signature -- something broken there? 2018-03-20 10:31:41 <_ikke_> micw: what mirror? 2018-03-20 10:32:32 _ikke_, default from docker image (dl-cdn.alpinelinux.org) 2018-03-20 10:33:06 that already happened last week but i did not find time to investigate yet 2018-03-20 10:33:13 issue is still there 2018-03-20 10:33:26 i dont even know what to do anymore 2018-03-20 10:35:25 <_ikke_> micw: not sure if that's the problem, but someone updated the package without bumping pkgrel, this ussually gives issues with dl-cdn due to it's caching features 2018-03-20 10:36:34 <_ikke_> micw: Might be wrong though, might have been updated in a single merge 2018-03-20 10:37:27 <_ikke_> Last update was in december, this update was in februari, so that 2018-03-20 10:37:35 <_ikke_> that's most likely the case 2018-03-20 12:46:39 <_ikke_> micw: They pushed a new version, should be fixed soon 2018-03-20 12:47:39 _ikke_, sorry, was off for lunch. I'll check... 2018-03-20 12:48:02 <_ikke_> micw: no worries 2018-03-20 12:48:20 <_ikke_> you should see an -r1 now 2018-03-20 12:48:52 <_ikke_> But it can take a bit of time before it's updated everywhere 2018-03-20 12:48:54 could take 30 min or so 2018-03-20 12:49:05 <_ikke_> right 2018-03-20 13:01:57 hi. where can i read about the history of alpine and where can i find some operational info about the distribution? i'd like to know who works on it, when it was conceived/first released, if its development is formally funded in any capacity, etc 2018-03-20 13:08:52 lunaphyte, alpine was born around 2005-08 2018-03-20 13:12:31 lunaphyte, a list of more recent active developers can be found here: https://git.alpinelinux.org/cgit/aports/stats/?period=y&ofs=100 2018-03-20 13:14:19 oh, thanks, that's great 2018-03-20 13:14:33 we dont have any public info about who is working on what. there are probably some ppl missing from the above list because they focus on infra or other alpine related things. 2018-03-20 13:18:44 oh, yeah, that's ok - i wasn't really after specific individual's names so much, more looking for some insight along the lines of does the project have a particular organizational structure to it [e.g. a formal team with defined members/roles/etc], or is it generally more an ad-hoc community effort, where an arbitrary given individual focuses on whatever they might focus on in some given moment? 2018-03-20 13:19:02 i did just find this, btw: https://wiki.alpinelinux.org/wiki/Alpine_Linux:Developers 2018-03-20 13:24:24 lunaphyte, seems that's not really updated regularly. 2018-03-20 13:24:40 i kind of got that sense, yeah 2018-03-20 13:24:56 _ikke_, my docker bontainer built with no error now 2018-03-20 13:25:23 <_ikke_> micw: Good to hear 2018-03-20 13:29:42 is there a governing body of any sort that oversees development of alpine? 2018-03-20 13:31:44 lunaphyte, we have recently been putting effort in making some sort of formal team, but its still a work in progress. 2018-03-20 13:32:14 we noticed the project has grown where its kind of impossible to have a single person to deal with everyhing. 2018-03-20 13:34:02 oh, ok 2018-03-20 13:34:08 well that's not a bad problem to have :) 2018-03-20 13:34:41 maybe, maybe not. it depends on your project goals. :) 2018-03-20 13:34:53 yeah, agreed 2018-03-20 13:35:14 i'm hoping that alpine's goals stay predominately technical 2018-03-20 13:36:17 imho, that has been the single biggest cause of the ultimately unfortunate direction other good distros have gone in, and ultimately beget unfortunate concessions/compromise 2018-03-20 13:43:01 generally speaking, could i expect that it's unlikely alpine linux would ever adopt systemd? 2018-03-20 13:48:28 lunaphyte: I hope AL will not adopt systemd 2018-03-20 14:01:23 <_ikke_> Chances are very slim it will happen 2018-03-20 14:03:27 that's encouraging 2018-03-20 14:04:31 another question - hopefully i can ask it without being too polarizing. is there a general philosophy as far as dependencies among software packages in alpine? 2018-03-20 14:05:12 one thing that's becoming more and more frustrating to me is the overconservative approach to package dependencies in other distros 2018-03-20 14:06:19 can you provide an example? 2018-03-20 14:06:56 sure, let me find done 2018-03-20 14:07:17 oh, i just happened across one the other day 2018-03-20 14:08:39 <_ikke_> lunaphyte: alpine tries to keep packages as small as possible, but it's always a trade-off between size and functionality 2018-03-20 14:11:15 kahiru: here's an example that, while "trivial", i consider to illustrate the frustration really well: http://dpaste.com/16FX5R0.txt 2018-03-20 14:11:36 oh, this 2018-03-20 14:11:37 that's from an oracle linux install [e.g. redhat, in essence] 2018-03-20 14:13:43 this was driving me mad on debian 2018-03-20 14:14:19 yeah, that's exactly the boat i'm in. my passtime has now become filing bugs for ridiculous dependencies 2018-03-20 14:14:23 <_ikke_> lunaphyte: alpine tries t split packages up as much as possible 2018-03-20 14:14:35 _ikke_: that's good to hear 2018-03-20 14:14:37 <_ikke_> lunaphyte: so that extra dependencies are as optional as possible 2018-03-20 14:15:42 ubuntu really does a pretty good job at splitting things up [in fact, most distros do, tbh], but the trouble surfaces when they don't allow the admin to make the decision is some supporting package is truly necessary 2018-03-20 14:23:59 is alpine formally/officially funded in any capacity? 2018-03-20 17:43:59 what's the opinion on embedding things like monit.d and logrotate.d files in packages. I see some ship with logrotate files 2018-03-20 18:37:51 or would a secondary package like `chrony-monit` make sense 2018-03-21 07:50:06 how to enable snd-pcsp? i tried commenting-out "blacklist snd-pcsp" but it didn't work. modprobe -vf snd-pcsp fails with "Exec format error" 2018-03-21 08:47:41 hello everyone 2018-03-21 08:47:59 I am trying to set up a Dockerfile, and I'm totally new to alpine 2018-03-21 08:48:20 i know in debian i can set a -y flag for non interactive install 2018-03-21 08:48:42 in the alpine apk --help, the only relevant flag i found was -i 2018-03-21 08:48:55 that this mean it is non interactive by default? 2018-03-21 08:49:10 <_ikke_> birdbolt1: yeah, I'm not aware of any interactive part in APK 2018-03-21 08:49:20 awesome thanks 2018-03-21 08:50:16 Non to be off topic, but are you the only guy who ever talks in here lol 2018-03-21 08:50:27 Every time I've ocme in, you're the only one who's responded 2018-03-21 08:51:37 <_ikke_> It varies 2018-03-21 08:58:29 lel 2018-03-21 09:28:39 Hey! I'm having issues with the latest "curl" version installed through apk (https://pkgs.alpinelinux.org/packages?name=curl&branch=v3.6&arch=x86_64). Seems like a mismatch in curl & libcurl version in the build: curl=7.59.0,libcurl=7.57.0. What would be the appropriate channel to get some answers on this? 2018-03-21 09:29:50 either here or if you're convinced that's the reason why curl is misbehaving you could probably just file an issue at the bug tracker 2018-03-21 09:31:23 Thanks kahiru, found the bug tracker link just as you wrote. Will head over there. 2018-03-21 09:33:20 also I'm just looking and it seems both curl and libcurl should be 7.59.0 in the repositories 2018-03-21 09:37:03 Sorry, I posted you the wrong link. I'm on alpine 3.4.6 2018-03-21 09:37:12 oh 2018-03-21 09:37:29 kahiru, created bug: https://bugs.alpinelinux.org/issues/8703 2018-03-21 09:39:13 could you post content of /etc/apk/repositories ? 2018-03-21 09:39:59 Bazze: because it seems to work for me https://paste.pound-python.org/show/tDHfBD7X997ctGyuZ18i/ 2018-03-21 09:40:20 so my next guess would be the built packages did not completely propagated to all the mirrors, but I might be wrong 2018-03-21 09:42:11 Interesting. I updated the bug with the repositories info. 2018-03-21 09:42:31 interesting 2018-03-21 09:45:47 Hmm I'm more and more convinced this is an issue in the image we are building our image on and not Alpine itself. Cache issue probably. 2018-03-21 09:46:20 might be 2018-03-21 09:50:09 No, I'm sure now. I'll try to close the bug, it's some layer in our docker image that's been messed up. Running clean from alpine:3.4 or mhart/alpine-node it works fine. Sorry for the noise. 2018-03-21 09:50:48 I'm sure nobody minds all that much as long as it gets solved in the end 2018-03-21 09:50:51 Hello, I tried to build package using arm64v8/alpine Docker image and qemu-aarch64-static with binfmt configured. Everything was fine except from created apk file for main package having 6 KiB size and only one broken symlink inside (*-dev, *-libs are OK). Investigated it a bit and apparently `abuild-tar` is the problem here, tarring main package yields 27.5 MiB text file and passing it through `abuild-tar --hash` reduces it 2018-03-21 09:50:51 to 6 KiB (again it works fine for *-dev and *-libs). Any idea what to do about it? 2018-03-21 09:57:37 w3m is available here: https://pkgs.alpinelinux.org/packages?name=w3m&branch=edge 2018-03-21 09:58:05 but for some reason `apk add w3m` results in "w3m (missing)" 2018-03-21 09:58:26 birdbolt1: do you have edge enabled in /etc/apk/repositories ? 2018-03-21 09:58:57 no, but I just specified some more: https://pkgs.alpinelinux.org/packages?name=w3m&branch=v3.7&arch=x86_64 2018-03-21 09:59:03 and it was still available 2018-03-21 09:59:21 do you have the community repository enabled? 2018-03-21 09:59:38 I have not enabled any such thing. is it disabled by default? 2018-03-21 09:59:44 I think it is 2018-03-21 09:59:51 only main is enabled by default iirc 2018-03-21 10:00:30 where can I enable that 2018-03-21 10:01:22 nvm found instructions 2018-03-21 10:01:35 beat me to it before I could write it out :) 2018-03-21 10:07:32 Hi, i try to install Alpine Linux on a 1GB HDD in sys mode and i need help or a advice to install the boot loader, i created one partion (ext4), but not sure how to get these bootable, should i use syslinux with EFI or just use grub-install? 2018-03-21 10:32:53 How can I see all cd drives in console 2018-03-21 10:35:07 the cd drives are usually /dev/sr* iirc 2018-03-21 10:35:21 so just ls /dev/sr* 2018-03-21 10:52:01 kahiru, I tries `mount \dev\sr0`and it returned: can't find /dev/sr0 in /etc/fstab 2018-03-21 10:57:46 birdbolt1: What is "it" 2018-03-21 10:57:52 Oh, never mind. 2018-03-21 10:58:17 "mount " won't work, you need a second argument, unless you've specified the drive in fstab. 2018-03-21 10:58:42 see mount --help for the correct syntax. 2018-03-21 10:59:25 Also, curious about using backslashes there? That probably won't work. 2018-03-21 10:59:35 backslash is a control character, after all. 2018-03-21 10:59:57 birdbolt1: "mount /dev/sd0 " 2018-03-21 11:00:48 and mountpoint is any reasonable path of my choosing? 2018-03-21 11:01:06 Yes. 2018-03-21 11:01:10 and sorry i typed those backslashes by msitake 2018-03-21 11:01:36 Yeah, you wouldn't have seen that message otherwise. 2018-03-21 11:04:05 now it says it failed 2018-03-21 11:04:10 Invalid argument 2018-03-21 11:04:35 `mount /dev/sr0 /home/vmwaretools` is what i tried 2018-03-21 11:05:51 duncan^, 2018-03-21 11:05:57 root 2018-03-21 11:06:09 Does that directory even exist? 2018-03-21 11:06:27 i tried it the first time, thinking it would be created auto 2018-03-21 11:06:38 then i went ahead and made it, but still got that message 2018-03-21 11:06:55 One of the arguments is incorrect. 2018-03-21 11:07:12 If the directory exists, does this device "/dev/sr0" even exist? 2018-03-21 11:08:24 just checked both directories exist 2018-03-21 11:08:51 Then you probably have invalid permissions, etc 2018-03-21 11:09:09 running as root currently 2018-03-21 11:09:21 or, there's no filesystem on /dev/sr0 2018-03-21 11:09:44 Its a VMware tools setup cd 2018-03-21 11:09:56 in debian, when I have a gui i can usually just extract it 2018-03-21 11:10:09 I have no idea what that is. 2018-03-21 11:10:16 You may have to specify a filesystem 2018-03-21 11:10:19 etc 2018-03-21 11:10:29 or the filesystem may be on a different device from /dev/sr0 etc 2018-03-21 11:10:56 util-linux's "lsblk" may be able to show you 2018-03-21 11:11:00 or use busybox blkid 2018-03-21 11:11:55 mount -t iso9660 for iso 2018-03-21 11:12:09 blkid reveals that its the setup iso i hooked up to vmware 2018-03-21 11:12:14 odd 2018-03-21 11:12:36 i have 3 sda. These are only for partitions right? 2018-03-21 11:12:52 No idea. They could be anything. 2018-03-21 11:13:04 Depends on your setup. 2018-03-21 11:15:09 alright i finally gave up and just installed a desktop environment lol 2018-03-21 11:15:30 every. single. time. :( I fear I'll never break my dependence on gui's 2018-03-21 11:16:24 I don't understand what the issue with a GUI is 2018-03-21 11:16:32 it's a tool, what's wrong with using it 2018-03-21 11:17:30 guys, I just wanna say: great distro!! keep up the good work :) very happy with it 2018-03-21 11:19:23 ^^ 2018-03-21 11:20:03 I'd just like to NOT depend on a gui so if I'm ever faced with a machine which for whatever reason I can't get a GUI on, I can actually use it 2018-03-21 11:20:13 well then do it 2018-03-21 11:20:22 At this point, I'm basically limited to cd'ing and ls'ing around lol 2018-03-21 11:20:23 maybe just use a window manager 2018-03-21 11:20:27 its a tendious process of several months of learning 2018-03-21 11:20:31 ^^ 2018-03-21 11:20:47 yea, I finally got xfce here, although I would prefer kde 2018-03-21 11:20:54 I may switch later 2018-03-21 11:21:23 kde isn't supported on alpine I believe 2018-03-21 11:21:36 but i wouldn't use it, uses soo many resources 2018-03-21 11:22:02 really? xfce it is then 2018-03-21 11:22:20 Might as well get used to an extra environment 2018-03-21 11:22:24 dwm by suckless is what I'm swearing by for almost a year now 2018-03-21 11:22:28 before that, i3 2018-03-21 11:22:37 does xfce use qt nowadays? 2018-03-21 11:22:46 probably not 2018-03-21 11:25:26 to run something in alpine its `./executable` right? 2018-03-21 11:25:44 birdbolt1: it's that way on pretty much every *nix like system 2018-03-21 11:25:51 that doesn't include '.' in $PATH 2018-03-21 11:25:53 im in a directory and trying that on a file it says its not found 2018-03-21 11:25:57 i ls and its right there 2018-03-21 11:26:06 ldd $file 2018-03-21 11:26:12 what's the ouput of that? 2018-03-21 11:26:30 Look, if you are comfortable with a GUI, use a GUI 2018-03-21 11:26:51 lmao I'll never learn if I keep shying away from cli tools lol 2018-03-21 11:26:53 get actual work done instead of fooling about with using tools that you won't read the documenation for 2018-03-21 11:27:09 there's nothing magical about "CLI tools" 2018-03-21 11:27:23 there's no documentation for vmwaretools i believe. 2018-03-21 11:27:31 GUI exists for a reason and that is to make the computer do the work for you 2018-03-21 11:27:35 use the GUI 2018-03-21 11:27:39 or not 2018-03-21 11:28:50 but then again -- alpine doesn't package *that* many GUI tools 2018-03-21 11:29:00 (which doesn't bother me personally) 2018-03-21 11:29:38 It packaged enough for me to use it on a laptop mostly comfortably 2018-03-21 11:29:45 but there are other OSs that one can install on laptops that do package more tools. 2018-03-21 11:29:51 yup 2018-03-21 11:29:54 or install from source, etc 2018-03-21 11:29:58 there is no crisis 2018-03-21 11:29:59 oh man. even gui doesnt work 2018-03-21 11:30:23 Failed to execute child process "xterm" (No such file or directory). 2018-03-21 11:30:27 well 2018-03-21 11:30:30 install xterm? 2018-03-21 11:30:45 I thought that was default terminal emulator for xfce? 2018-03-21 11:30:54 birdbolt1: nope, that'd be xfce-terminal 2018-03-21 11:30:58 Set another one 2018-03-21 11:31:10 thats xfce4-terminal 2018-03-21 11:31:21 there is an option in settings to change default programs 2018-03-21 11:31:37 busybox has so many good utilities, barely need much more for a comfy daily system 2018-03-21 11:31:57 They exist, not all of them are "good" 2018-03-21 11:32:16 imho busybox is good enough 2018-03-21 11:32:17 some of them are better than their full-fat originals 2018-03-21 11:32:35 vi(1) could use unicode support tough 2018-03-21 11:32:44 <_ikke_> I usually miss a lot of usefull options in the busybox tools 2018-03-21 11:32:46 some of them basically are the full-fat originals, albeit with options disabled 2018-03-21 11:32:47 (not that I'm experienced enough to write a patch) 2018-03-21 11:33:30 udhcpc, ntpd, nc, acpid etc. do everything I need them to do, for example 2018-03-21 11:35:13 and there is not an effort towards POSIX compliancy in Busybox 2018-03-21 11:36:55 12:32 dunca some of them basically are the full-fat originals, albeit with options disabled 2018-03-21 11:37:02 sometimes also with important security patches missing 2018-03-21 11:37:19 because busybox maintainers are even less on top of their shit than debian "developers" 2018-03-21 11:37:46 Not sure why you put that in inverted commas, given they do actually do development 2018-03-21 11:37:47 they just put busybox on their CV and then they move on. 2018-03-21 11:38:03 duncan^: they do? 2018-03-21 11:38:18 some of them do 2018-03-21 11:38:22 zzz 2018-03-21 11:38:28 hiro: well 2018-03-21 11:38:39 oh i'm not claiming there aren't debian "developers" that don't develop anything. 2018-03-21 11:38:52 i'm just saying there are debian "developers" who have never written a line of code. 2018-03-21 11:39:20 liwakura: hm? 2018-03-21 11:39:21 hiro: busybox seems better maintained than alpine stable 2018-03-21 11:39:32 really? :( 2018-03-21 11:39:45 liwakura: how did you measure that though? 2018-03-21 11:40:06 hiro: looking at hardened 2018-03-21 11:40:18 4.9.6*-something 2018-03-21 11:40:29 liwakura: i agree we should remove hardened. 2018-03-21 11:40:37 actually its at 73 now 2018-03-21 11:40:46 74 is last version 2018-03-21 11:40:52 upstream is at 88 2018-03-21 11:41:14 yes 2018-03-21 11:41:40 liwakura: so your definition of maintainership is how often they upgrade their linux kernels? 2018-03-21 11:42:07 hiro: v3.7 still hasn't the KPTI patches 2018-03-21 11:42:14 liwakura: so? 2018-03-21 11:42:25 >< hiro> sometimes also with important security patches missing 2018-03-21 11:42:46 the thing is: security is not so trivial 2018-03-21 11:43:00 indeed, KPTI patches cause a performance regression 2018-03-21 11:43:01 is spectre1 fixed ? 2018-03-21 11:43:15 but in this case, alpine stable is not better than busybox 2018-03-21 11:43:27 stay on point. 2018-03-21 11:43:47 until there are no sensible patches, its still intel who is at fault 2018-03-21 11:44:13 because its essentially an hardware failure 2018-03-21 11:45:19 liwakura: no that's not the point. i mean, there's one great thing about spectre/meltdown: it's a great wakeup call to all those morons that try to shove more complexity and useless non-generalizable heuristic methods into everything for security as the only goal. then spectre comes along, and it doesn't matter how detailed all your silly heuristics and syscall whitelists and blacklists would have been 2018-03-21 11:45:19 - everything crumbles down, and people have to realize that process separation was NEVER a hard security guarantee, and while people liked to pretend (for performance, price, comfort reasons), it's all too complex to be guaranteed well. 2018-03-21 11:45:42 hiro: i dont care about that 2018-03-21 11:45:56 12:43 liwak stay on point. 2018-03-21 11:45:57 i care about how you shit on busybox because you dont know crap about it 2018-03-21 11:46:00 you are not to judge what's on point 2018-03-21 11:46:28 liwakura: how do YOU know what i know about busybox? 2018-03-21 11:46:34 liwakura: aren't you a bit presumptious here? 2018-03-21 11:46:49 you just demonstrated it? 2018-03-21 11:47:13 i dont know it perfectly, either, but im aware of it enough not to spew up FUD about it 2018-03-21 11:47:34 12:43 liwak until there are no sensible patches, its still intel who is at fault 2018-03-21 11:47:35 wrong 2018-03-21 11:47:52 perhaps for meltdown you could claim intel was just stupid 2018-03-21 11:48:17 but to turbooptimize all the out of order stuff was an industry decision. 2018-03-21 11:48:31 everybody started putting those predictors 2018-03-21 11:48:35 Intel broke virtual memory isolation, a thing they assert in the datasheet. They broke that assertion. Thats a bug. 2018-03-21 11:48:42 amd even prided themselves they use some machine learning crap for it!! 2018-03-21 11:49:17 liwakura: it's more than just intel 2018-03-21 11:49:22 liwakura: the whole industry is affected 2018-03-21 11:49:33 it's a more fundamental problem. 2018-03-21 11:49:44 cut that crap, most architectures aren't even remotely affected 2018-03-21 11:49:58 if you really care you shouldn't rely on process isolation in the first place and just put more of your shit on physically isolated machines. 2018-03-21 11:50:07 i dont disagree with that. 2018-03-21 11:50:10 all that happened is that people got reminded about this fact. 2018-03-21 11:50:19 it was never very secure. 2018-03-21 11:50:27 nobody SHOULD have believed it was secure. 2018-03-21 11:50:53 the whole cloud bullshit is based on this insecure crap. and i wish more people would acknowledge it 2018-03-21 11:50:58 Yeah, people should have made informed decisions 2018-03-21 11:51:11 the openbsd guy kept on telling people that virtual machines don't create secure isolation either 2018-03-21 11:51:24 few people like him kept reminding, but nobody listened, ever. 2018-03-21 11:51:29 This is no news for me 2018-03-21 11:51:40 that's not intel's fault. people WANTED insecure shitty CPUs, they delivered. 2018-03-21 11:52:19 12:46 liwak you just demonstrated it? 2018-03-21 11:52:19 12:47 liwak i dont know it perfectly, either, but im aware of it enough not to spew up FUD about it 2018-03-21 11:52:24 i've asked the maintainer to fix some shit 2018-03-21 11:52:39 Were you the rude asshole on the ML? 2018-03-21 11:52:48 no, i only talked privately to him 2018-03-21 11:52:58 where are the init directories on alpine 2018-03-21 11:53:04 "talked" ? Not "submitted a patch" ? 2018-03-21 11:53:10 is it /etc/init.d/ ? 2018-03-21 11:53:39 liwakura: there was no patch necessarry. an old change they did wrongly reverted earlier changes that just needed to be reapplied 2018-03-21 11:53:57 liwakura: this is why i brought up debian "developers" in the first place. even THEY had their shit together better than busybox 2018-03-21 11:54:14 liwakura: they were the ones that sent them patches for security, because they knew busybox weren't tracking upstream. 2018-03-21 11:54:35 liwakura: then the busybox people wrongly fucked everything up when they applied that patch 2018-03-21 11:54:43 debian people try to wrap their software into pretty frameworks and abstractions and fuck up the distro for exotic use cases 2018-03-21 11:54:56 the installer and FAI are the best example of it 2018-03-21 11:55:17 liwakura: fast forward many years and busybox keeps 10 people busy all day trying to figure out why nothing works, everybody apart from me didn't even want to contact them cause they figured they will probably not answer anyway. 2018-03-21 11:55:43 hiro: that could as well be because you are the asshole, not them 2018-03-21 11:55:44 but then to the current maintainer's defense: he did read my mail carefully, fixed everything AND explained how this fuckup happened historically 2018-03-21 11:56:02 i think he wasn't the maintainer yet when it happened. 2018-03-21 11:56:15 wtf liwakura 2018-03-21 11:57:01 liwakura: first you open your mouth real big and claim i "demonstrated" how i know nothing about busybox, now after hearing my actual first-hand experience you resort to insult?! 2018-03-21 11:57:17 liwakura: who are you trying to convince, and of what?! 2018-03-21 11:57:48 Exhibit A: < hiro> because busybox maintainers are even less on top of their shit than debian "developers" 2018-03-21 11:58:04 Exhibit B: < hiro> because busybox maintainers are even less on top of their shit than debian "developers" 2018-03-21 11:58:17 Exhibit C: < hiro> they just put busybox on their CV and then they move on. 2018-03-21 11:58:28 oh, i duplicated B 2018-03-21 11:58:32 but the backlog is public 2018-03-21 11:58:38 ok, and the point? 2018-03-21 11:58:40 no, you were the one with the big mouth 2018-03-21 11:58:52 look, what you don't get is even who i blame 2018-03-21 11:58:56 and your opionated FUD is the only think why i felt the urge to argument against it 2018-03-21 11:59:04 do you even know about busybox maintainership history???? 2018-03-21 11:59:13 most people invovled already claimed there were problems 2018-03-21 11:59:23 I dont claim there are no problems 2018-03-21 11:59:27 it's not that they are all stupid incompetentet individuals. 2018-03-21 11:59:29 But you were just an asshole in the channel 2018-03-21 12:00:09 the way busybox is managed, and the number of developers involved it doesn't work. there aren't the necessarry ressources, the proper leadership, and the people who could have done better were offended and left, too. 2018-03-21 12:00:10 Also, you could join #musl and talk to them face-to-face 2018-03-21 12:00:34 instead of spreading your FUD here 2018-03-21 12:00:51 i'd prefer more insults rather than stupid distractions 2018-03-21 12:01:33 first you tell me to stay "on topic", then when you don't like the experiences that formed my PERSONAL OPINION, you try to threaten me with how publically visible my opinion is even to other people who haven't insulted me yet. 2018-03-21 12:02:16 you were an asshole, and when i argumented against it, you just talked about the spectre crap 2018-03-21 12:02:44 maybe i should have just said, the way you behave is shitty, instead of trying to argument it directly 2018-03-21 12:02:56 wtf, you brought up the kernel lagging behind and missing meltdown mitigation. 2018-03-21 12:03:22 to back up my point that alpine stable is not better maintained than busybox 2018-03-21 12:03:46 mostly due to a lack of non-lazy people 2018-03-21 12:03:46 yeah, a dumb point from the beginning. apples and oranges. 2018-03-21 12:03:54 i shouldn't have gotten into such stupid discussion 2018-03-21 12:04:06 haha, lack of non-lazy people 2018-03-21 12:04:27 how about you breed some and leave us alone instead? 2018-03-21 12:05:20 anyway. while i'm not putting my hope into busybox i hope toybox will become usable one day :( 2018-03-21 12:06:05 alternatively, you could stop talking crap about people behind their backs 2018-03-21 12:06:17 which is technically not true, because several of them are joined here 2018-03-21 12:06:46 stating ones opinion by insulting others and accusing them of incompetence is a bad pattern in general 2018-03-21 12:07:00 liwakura: behind their backs? 2018-03-21 12:07:00 wtf 2018-03-21 12:07:04 as you said it's public 2018-03-21 12:07:10 i'm not scared to share my opinion. 2018-03-21 12:07:16 doesn't mean the'll read it 2018-03-21 12:07:21 i'll tell people in the face if they want to hear, so don't misjudge me. 2018-03-21 12:07:56 anyway. enough personal advice from you. 2018-03-21 12:09:37 perhaps i should make clear that i didn't see incompetence of any individual involved in the busybox project (mainly cause i didn't care to check out the commiter's names too closely). but all in all i can tell that how the project is maintained, organized, how the code is copy-pasted but not tracking upstream properly, there are parts in busybox that are a net negative. while there are others, that 2018-03-21 12:09:37 are busybox specific that don't need to track any upstream, which i find more useful in general. i wish they would concentrate on those. 2018-03-21 12:09:49 make a fork 2018-03-21 12:09:59 maintainers rarely owe you anything 2018-03-21 12:10:23 wtf, where did i talk about somebody owing me? 2018-03-21 12:10:44 you are the only one who keeps on making unwarranted personal insults and dumb recommendations without knowing my background. 2018-03-21 12:10:53 you just voiced your expectations on the project 2018-03-21 12:11:06 what braindamage. 2018-03-21 12:12:02 everything i said was about shittily maintained software. i can produce the same, and i would have a reason it is shitty and all that. 2018-03-21 12:12:53 complaining doesn't do shit except harass the people who are already spending their free time on the project. 2018-03-21 12:12:59 can any of your recommend a good text editor on alpine? 2018-03-21 12:13:10 <_ikke_> vim :P 2018-03-21 12:13:10 I've been using sublime on other systems 2018-03-21 12:13:14 eww pls no 2018-03-21 12:13:16 birdbolt1: good for what? GUI? cli? 2018-03-21 12:13:19 GUI 2018-03-21 12:13:42 <_ikke_> I don't know any good GUI editors 2018-03-21 12:13:46 aw 2018-03-21 12:14:16 <_ikke_> I try them sometimes, but they keep anoying me in some way 2018-03-21 12:14:38 editors like vim and emacs have GUI versions 2018-03-21 12:14:41 (please don't stone me to death) emacs 2018-03-21 12:14:51 kahiru: dont be afraid, i use nano 2018-03-21 12:15:00 Use whatever you feel most comfortable with 2018-03-21 12:15:00 used mcedit previously 2018-03-21 12:15:37 Some people like joe, nvi, vim, emacs, mg, nano, ed, […] 2018-03-21 12:16:05 hiro: if you really dislike busybox and whatever, just dont use it, and the whole issue is solved. 2018-03-21 12:16:30 just not using a project is the biggest sanction you can give to project without being an total asshole 2018-03-21 12:17:23 liwakura: why do you keep on telling me what to do? what is your benefit? 2018-03-21 12:18:00 liwakura: software can have net negative impact. 2018-03-21 12:18:10 liwakura: that's why negative feedback to shitty developers would be useful. 2018-03-21 12:18:23 liwakura: give them a chance to stop what they're doing and find something they're better at. 2018-03-21 12:18:57 when i say busybox is shit it is bec. i already found all usable parts in busybox and usable alternatives for the rest. 2018-03-21 12:19:07 why do you assume i personally need busybox to be better for my own benefit? 2018-03-21 12:19:11 This is not the place to give negative feedback to busybox devs 2018-03-21 12:19:35 this is a place that - like me - uses busybox for many things 2018-03-21 12:19:57 and i'm just pointing out that while parts of it are useful it's not all high quality, for various reasons, many of which documented by ex-busybox maintainers 2018-03-21 12:20:11 You made your point some time ago 2018-03-21 12:20:12 other reasons are more obvious: for example time ressources of actual developers. 2018-03-21 12:20:31 a warning can go long ways. 2018-03-21 12:20:39 hiro: i dont understand your attitude 2018-03-21 12:20:47 if people read one sided praise of busybox it is my responsibility to show the other side 2018-03-21 12:20:47 people make busybox for free, and you expect more 2018-03-21 12:21:23 ofc software is usually shitty, but thats responsibility of the person using it 2018-03-21 12:21:24 liwakura: oh god, again. i don't want anything from busybox devs. the people i hope to read are the ones reading this channel and trying to decide which piece of softare to consider for their basic linux userland 2018-03-21 12:22:07 busybox is pretty nice 2018-03-21 12:22:17 far from shit 2018-03-21 12:23:26 hiro: educating them instead of opionating around would be better than 2018-03-21 12:23:42 liwakura: also the "for free" argument is real dangerous. advertisment is for free, too. 2018-03-21 12:23:56 people spend their free time on OSS 2018-03-21 12:24:07 liwakura: there's a bunch of things that are free and stil harmful. 2018-03-21 12:24:08 they dont owe anybody anything 2018-03-21 12:24:12 and? 2018-03-21 12:24:17 nobody forces you to use it 2018-03-21 12:24:31 "opionating around" 2018-03-21 12:24:37 obviously my "education' is based on my opinion 2018-03-21 12:24:51 you can state your's and i state mine, the reader will decide what to do with it. 2018-03-21 12:24:53 but thats like, just your opinion man 2018-03-21 12:25:06 teach people to read the code instead 2018-03-21 12:25:17 of course now i feel bad cause if you HAD a technical opinion it probably got lost in all the personal insult and shitty rhetorics. 2018-03-21 12:25:52 you wanna hear MY technical opinion on busybox? 2018-03-21 12:25:59 if you have one, i guess? 2018-03-21 12:26:15 it does the job fine for me, and the few bugs i encountered were like, nothing devastating 2018-03-21 12:26:21 im content with it 2018-03-21 12:26:35 but i dont understand why i should now walk around and convince people to use it 2018-03-21 12:26:41 or convince people not to use it 2018-03-21 12:26:55 "they dont owe anybody anything" -> no? how about just basic human decency? or not creating more unnecessarry damage on this world? 2018-03-21 12:27:30 so your *technical* opinion is: "works for me"? 2018-03-21 12:27:34 hiro: so, what kind of damage is it doing ? 2018-03-21 12:28:22 hiro: yeah. And the kbuild respects the flags i give in. I do actually like that. 2018-03-21 12:28:53 Because GNU/autotoools-based project sometimes to pretty dumb things with the inputs i give 2018-03-21 12:29:00 s/to/do/ 2018-03-21 12:29:22 liwakura: agreed. 2018-03-21 12:29:33 many times busybox is better than the gnu alternative. 2018-03-21 12:31:11 i remember that the uid namespaces in the unshare applet were just plain broken 2018-03-21 12:31:50 but i assumed vda had no previous experience with setting up user namespaces manually, so i intended a patch and saw it was already fixed upstream 2018-03-21 12:32:23 and that adduser segfaults under certain circumstances was also fixed before i needed to act upon it 2018-03-21 12:33:29 uhhh don't remember me of unshare haha 2018-03-21 12:33:48 right now i'm having so many problems with unshare(newnetns 2018-03-21 12:33:55 kernel is just fucked lately 2018-03-21 12:34:09 that syscall never finishes here. 2018-03-21 12:34:18 what timeframe is "lately" by your standards? 2018-03-21 12:34:21 not alpine though, but just a warning, in case anybody here uses containers :D 2018-03-21 12:34:41 since the meltdown/spectre stuff im trying to keep away from namespaces 2018-03-21 12:34:42 kahiru: kernel on ubuntu 16, upgraded like... monday? 2018-03-21 12:34:59 liwakura: yep, i've been seeing a LOAD of problem since meltdown/spectre stuff particularly 2018-03-21 12:35:04 this is just the latest and most insane :D 2018-03-21 12:35:06 ah, ubuntu/debian has custom LTS branch afaik 2018-03-21 12:35:19 they might have fucked up backporting some stuff? no clue 2018-03-21 12:35:25 happens *shrugs* 2018-03-21 12:35:27 but i also saw upstream discussions... 2018-03-21 12:35:30 wow i cant find one fucking gui editor for alpine :( 2018-03-21 12:35:33 just one... 2018-03-21 12:35:58 birdbolt1: im not too much surprised, but if you want to get rid of the GUI anyways, i'd recommend you nano or mcedit (part of mc) 2018-03-21 12:36:02 a lot of code that hasn't been touched in ages is now modified by people unrelated to the code - for security. 2018-03-21 12:36:04 <_ikke_> birdbolt1: geany :p 2018-03-21 12:36:21 I recall having geany back when i used linux mint 2018-03-21 12:36:39 is the one on alpine gui? 2018-03-21 12:36:45 <_ikke_> https://pkgs.alpinelinux.org/packages?name=geany&branch=edge 2018-03-21 12:36:59 How do i know from that lol ^ 2018-03-21 12:37:29 i like notepad.exe 2018-03-21 12:37:29 it depends on gtk+2.0 2018-03-21 12:37:40 works in wine i'd assume. 2018-03-21 12:38:15 well geany isnt quite the "text editor" i hoped for but thank you 2018-03-21 12:38:20 I'll settle for that 2018-03-21 12:38:40 i use vim, but i guess you could use gvim? 2018-03-21 12:38:43 birdbolt1: what did you hope for? 2018-03-21 12:38:54 isnt geany an IDE 2018-03-21 12:38:59 though i have no clue how many gigabytes of gtk3 it includes nowadays 2018-03-21 12:39:13 i was after just a plain text editor like notepad or sublime. and dint consider it 2018-03-21 12:39:15 my coworker is using some shit in a chrome tab now as his editor :D 2018-03-21 12:39:47 hiro, gvim (missing) 2018-03-21 12:39:52 LMAO 2018-03-21 12:39:55 there's leafpad 2018-03-21 12:40:04 I can feel his frustration 2018-03-21 12:40:08 birdbolt1: then i guess it had too many dependencies! :D 2018-03-21 12:40:09 man mousepad 2018-03-21 12:40:16 s/man/and/ 2018-03-21 12:40:39 beautiful 2018-03-21 12:40:39 birdbolt1: i'd never think of using it cause i rely too much on ssh, and all the modern gui features are very latency sensitive so forwarding x11 became useless. 2018-03-21 12:40:46 leafpad is just what i needed 2018-03-21 12:41:09 why didnt leafpad show up in any of my searches man 2018-03-21 12:41:18 leafpad is gtk+ 2018-03-21 12:41:20 about 1 hour wasted tring to figure this out smh 2018-03-21 12:41:27 is that a bad thing 2018-03-21 12:41:34 birdbolt1: you didn't search for *pad, i guess 2018-03-21 12:42:00 nah, i spent my time doing variations of "text editor alpine" 2018-03-21 12:42:03 birdbolt1: what is smh? 2018-03-21 12:42:08 shaking my head 2018-03-21 12:42:18 smdh == shaking my damn head 2018-03-21 12:42:25 haha 2018-03-21 12:43:57 is gtk bad? 2018-03-21 12:45:04 no, it is just fine 2018-03-21 13:44:00 Fixed 2018-03-21 13:44:07 EFI disabled. xD 2018-03-21 13:45:09 well, that's one way to go about it 2018-03-21 13:46:23 I always wondered what smh meant 2018-03-21 14:02:49 Hei, do you guys use rustup on alpine? 2018-03-21 14:04:51 Invader_Bork: I don't see rustup in packages though. 2018-03-21 14:05:45 I am really impressed: https://nopaste.xyz/?1b46100c60e1211a#8Nvgckj9nT1IJSpqQN/tUF0imJd/Bb1Hh0ysUDE7pNw= 2018-03-21 14:06:06 Last time I tried there wasn't rust for musl but later it came out but never tried. 2018-03-21 14:07:18 pickfire_, it's not. i normally get it with curl https://sh.rustup.rs -sSf | sh -s but it seems it still downloads the glibc version which does not run on alpine 2018-03-21 14:08:06 pickfire_, i wanted to manage my local rust installation with rustup because it updates frequently and I wanted to still be on alpine stable 2018-03-21 14:09:01 Invader_Bork: I am on stable but I don't see rust though. 2018-03-21 14:12:09 pickfire, maybe you need the community repo enabled 2018-03-21 14:12:43 pickfire, it has rust-1.22.1 2018-03-21 14:13:40 Invader_Bork: Oh wait, it's there. 2018-03-21 14:14:03 pickfire, but I am specifically asking about rustup which is a rust installation manager 2018-03-21 14:14:16 pickfire, my goal is to be on alpine stable and the latest rust stable as well 2018-03-21 14:14:42 Invader_Bork: I think if you install it with alpine stable, it's better to stick with distro stable. 2018-03-21 14:14:53 Unless it's unmaintained (like some of my packages) 2018-03-21 14:16:27 Invader_Bork: But I saw that jacob is maintaining it then probably it's good to go. 2018-03-21 14:16:34 Need to to move to rustup. 2018-03-21 14:16:38 not* 2018-03-21 14:30:55 pickfire, in conclusion because it turns out that they are still merging in patches upstream in order to properly support musl i will for the time being just switch to edge if I want to use 1.24.1 2018-03-21 14:37:30 Stupid question, first time on ash, is there a file like .ash.rc? :D For aliases e.g.? 2018-03-21 14:38:19 Bodenhaltung: its done via the POSIX-standardized mechanism 2018-03-21 14:38:30 in .profile, you define ENV=$HOME/.ashrc or something like that 2018-03-21 14:38:37 or a file of your choice 2018-03-21 14:38:51 needs re-login to take effect 2018-03-21 14:39:02 Ah, nice, thank you. 2018-03-21 14:40:19 its just zsh and bash that are non-standard 2018-03-21 14:40:29 bash even in a IMHO dumb way 2018-03-21 14:41:21 my .bashrc: 2018-03-21 14:41:21 export ENV="$HOME/.env" 2018-03-21 14:41:22 . "$ENV" 2018-03-21 14:41:54 dalias: i just have an single file, ~/.profile is an symlink to it, and it sets itself as ENV 2018-03-21 14:42:09 its reentrant to hack around bash weirds behavior 2018-03-21 14:42:22 .profile sets up $ENV too, but i have different behavior for login vs nonlogin shell 2018-03-21 14:42:44 s/reentrant/can be executed multiple times without duplicating stuff/ 2018-03-21 14:42:47 non-idempotent stuff like 2018-03-21 14:42:50 PATH="$HOME/bin:$PATH" 2018-03-21 14:43:01 idempotent is what im looking for! 2018-03-21 14:44:16 bash is annoying because it executes bashrc when interactive (except not login shell), and when invoked via ssh, despite not being interactive nor login 2018-03-21 14:44:50 and BASH_ENV has different semantics than ENV... 2018-03-21 14:46:14 Works fine, thanks again :) 2018-03-21 14:49:27 Neext, sorry, new user and doesnt find it on the wiki, how can i search a binary belongs to which paket? In gentoo speach "equery b name" (sorry, gentoo user mostly) 2018-03-21 14:50:46 Ah, like "info --who-owns" but on uninstalled packages 2018-03-21 14:51:18 you can do it vuia pkgs.alpinelinux.org 2018-03-21 14:53:40 Ah, ok, thanks 2018-03-21 15:01:22 Last question, how to remove a package with they dependencies? 2018-03-21 15:01:59 Bodenhaltung: Removing a package removes the dependencies. 2018-03-21 15:02:07 Unless they are shared by another package. 2018-03-21 15:02:20 apk del 2018-03-21 15:02:28 Hmm, ok, but the diskspace is still used? 2018-03-21 15:03:23 no 2018-03-21 15:03:51 Sure, and the package is still installed :( 2018-03-21 15:03:55 apk keeps all "top-level" dependencies in /etc/apk/world 2018-03-21 15:04:19 that is the list that apk will try to keep installed 2018-03-21 15:04:20 apk add will add to that list 2018-03-21 15:04:25 apk del will remove it 2018-03-21 15:04:44 apk will reoslve the entire dependency tree on both add and del 2018-03-21 15:04:52 so evertying not needed or explicitly added will be removed 2018-03-21 15:04:59 Ok, i get it, nice, a world list. :) 2018-03-21 15:05:16 I think my problem is 100% / 2018-03-21 15:05:34 so thats maybe why apk del foo does not work well 2018-03-21 15:05:58 you can manually remove a package from /etc/apk/world and on next apk add the package will be removed 2018-03-21 15:06:15 try purge logs first 2018-03-21 15:06:39 how? 2018-03-21 15:06:54 rm /var/log/ 2018-03-21 15:07:12 ore remove some big package 2018-03-21 15:07:17 ah, ok, sure, i that there is a fancy script or something 2018-03-21 15:07:20 rm /usr/bin/foo; apk del foo 2018-03-21 15:08:14 i normally try figure out what takes space: du -s /usr/* | sort -n 2018-03-21 15:08:19 or du -s /var/* | sort -n 2018-03-21 15:10:43 ah, strange, i have only 600MB on / so, this is all a little "fuzzy" 2018-03-21 15:11:05 275.3M /lib/modules, 211.8M /lib/firmware mäh 2018-03-21 15:12:05 Is there a smart and clean way to delete some stuff? 2018-03-21 15:12:50 linux-firmware has some sub-packages 2018-03-21 15:13:13 the modules directory can be reduced by removing unused kernel packages 2018-03-21 15:13:26 275MB is probably 2 or 3 kernels 2018-03-21 15:26:41 kernel is big yes 2018-03-21 15:31:59 Hmm, i think i have only one? 2018-03-21 15:32:44 Or is there a apk kernel command? I haven´t find anything about the kernel? 2018-03-21 15:36:31 Bodenhaltung: what does ls /lib/modules output? 2018-03-21 15:39:08 Only "4.9.65-1-hardened" 2018-03-21 15:40:41 i wonder why its so big then... 2018-03-21 15:44:47 Hmm 2018-03-21 15:56:51 I have no idea :( 2018-03-21 15:57:11 How can i add exfat? exfat-utils are only for armh? 2018-03-21 15:58:52 liwakura: Was my mistake, some really big dependency from ffmpeg 2018-03-21 15:59:29 huh 2018-03-21 16:00:02 hm, no, that was only ~80MB 2018-03-21 16:00:37 kernel/drivers is 192.2M 2018-03-21 16:00:54 looks normal 2018-03-21 16:05:21 Is there a way to use a smaller kernel? 2018-03-21 16:55:34 How can i add exfat? exfat-utils are only for armh? 2018-03-21 16:55:37 what kind of wget alpine linux has? im trying to mirror a website and the -m option is not available 2018-03-21 16:56:10 <_ikke_> id4rk: most likely busybox 2018-03-21 16:56:53 <_ikke_> You can separately install wget, which will get the original one 2018-03-21 16:59:11 true, thanks! will do 2018-03-21 17:01:45 worked =) 2018-03-21 17:34:38 Hey guys, just setup my first alpine machine (v3.7 VM image). Any reason why the repos default to http and not https? 2018-03-21 17:35:34 <_ikke_> Harm133: many mirros don't support https 2018-03-21 17:35:37 <_ikke_> mirrors 2018-03-21 17:36:34 _ikke_, Wouldnt it be nice to force them :)? 2018-03-21 17:37:09 Seeing as /secure/ is one of the 3 main points of this distro it seems 2018-03-21 17:37:14 Loving it so far, btw 2018-03-21 17:37:35 <_ikke_> Harm133: all packages and the index are signed, so that should cover most 2018-03-21 17:38:20 alright thx :) 2018-03-21 17:38:38 Harm133: maybe some mirrors support https 2018-03-21 17:38:57 yandex supports 2018-03-21 17:38:59 <_ikke_> dl-cdn.alpinelinux.org does not work, because it points to fastly 2018-03-21 17:39:00 mps, leaseweb does :) 2018-03-21 17:41:57 repository.fit.cvut.cz also support https 2018-03-21 17:43:29 nl.alpinelinux.org too 2018-03-21 17:43:35 Thats the thing :) using setup-apkrepos it defaults all mirrors to http, not https 2018-03-21 17:44:07 <_ikke_> I think most people don't care enough and trust the signatures 2018-03-21 17:44:09 Like I said, I just setup my first alpine linux machine and it stood out to me 2018-03-21 17:44:36 Harm133: alpine packages are signed anyways 2018-03-21 17:45:52 <_ikke_> Harm133: here is the list of mirrors: https://git.alpinelinux.org/cgit/aports/tree/main/alpine-mirrors/mirrors.yaml 2018-03-21 17:46:28 <_ikke_> Harm133: The build script only includes http:// mirrors though 2018-03-21 17:50:02 _ikke_, maybe https://mirrors.leaseweb.com should be added there :) 2018-03-21 17:50:17 <_ikke_> It's in there 2018-03-21 17:50:24 <_ikke_> line 41 2018-03-21 17:50:37 <_ikke_> ah, not https 2018-03-21 17:50:39 :) 2018-03-21 17:50:44 exeactly 2018-03-21 19:02:56 Harm133: as long as packages are signed and your Alpine machine uses Alpine dev's signing keys, I don't think https is that important. 2018-03-21 19:53:25 Orochimarufan: hi, still playing with ACF? just noticed your question from last week while catching up on backlog. 2018-03-21 19:54:04 I'm mostly done with what I wanted to do 2018-03-21 19:55:39 for system passwords, there is alpine-baselayout/password 2018-03-21 19:55:53 it's not included in the default menu, so it's kind of hidden 2018-03-21 19:57:15 ACTION is always happy to hear of more people using ACF 2018-03-21 19:59:02 Hmm I´m getting a bunch of ¨ttyS ttyS0: tty_port_close_start: tty->count = 1 port count = 2¨ on one of the two machines I installed. the one I´m getting this message on it 3.7 alpine-virt the other, without the message, is 3.7 alpine-extended. Any ideas? 2018-03-21 19:59:53 Both virtual machines, hosted on ESXi 6.5 2018-03-21 20:05:19 did you just use ümlauts instead of quotation marks? :') 2018-03-21 20:05:22 Hi, looking for some advice. Running alpine as an vm under proxmox. I have installed qemu-guest-agent but the process fails to start. "Error opening channel". Any suggestions? 2018-03-21 20:05:39 danieli, my irc program is bad :( 2018-03-21 20:05:43 danieli: what client / terminal emulator did you use? 2018-03-21 20:05:48 Harm133: no, i saw it correctly 2018-03-21 20:05:57 liwakura, on mine its also umlauts 2018-03-21 20:06:22 i just see two dots 2018-03-21 20:06:31 ¨ they're umlauts, " are quotation marks 2018-03-21 20:06:39 no 2018-03-21 20:06:48 thats the diacritic that is used in umlauts 2018-03-21 20:06:58 true true, close enough 2018-03-21 20:07:00 but it can also be used on other things that wont be umlauts then 2018-03-21 20:07:06 i didn't care to go into detail 2018-03-21 20:07:17 they're called diaeresis 2018-03-21 20:07:31 yeah! thats the accurate name for that specific one 2018-03-21 20:07:43 now im wondering how Harm133 even typed that thing 2018-03-21 20:07:51 >:) 2018-03-21 20:08:01 i'm norwegian, there's a button for it 2018-03-21 20:08:14 shift + the same button = ^, ctrl + the same button = ~ 2018-03-21 20:08:29 ¨/^/~ 2018-03-21 20:08:44 well, I use hexchat for windows and always had that issue. I want quotes but I get an ¨ 2018-03-21 20:08:51 Im gonna look into it :p 2018-03-21 20:09:43 <_ikke_> " 2018-03-21 20:10:03 shift+2 = " on scandinavian keyboards 2018-03-21 20:10:26 <_ikke_> On us keyboards it has a dedicated key 2018-03-21 20:10:41 <_ikke_> (still requires shift though) 2018-03-21 20:12:21 regarding my original question tho, any ideas? I´m getting those messages every 10 seconds 2018-03-21 20:17:06 altgr+shift+' twice on us(intl-altgr) kb 2018-03-21 20:18:15 more fun is altgr+shift+[/] 2018-03-21 20:18:17 “” 2018-03-21 20:18:24 anyway 2018-03-21 20:18:55 it sounds like the virtual serial port has some weirdness in how it's hooked up 2018-03-21 20:19:16 maybe there's a boot cmdline option that would disable it 2018-03-21 20:20:09 Hmm, its weird tho that one machine has it and the other doesnt 2018-03-21 20:20:22 lemme spin up another vm image and see if its the same way 2018-03-21 20:23:18 brb, updating, to get my quotes back 2018-03-21 20:23:38 :) 2018-03-21 20:31:24 Seems its a bug with the client. Fixed now " 2018-03-21 20:31:28 woo 2018-03-21 20:32:00 :) 2018-03-21 20:33:01 Spinning up another VM now to see if the error exists there as well 2018-03-21 20:40:55 same deal on the new machine 2018-03-21 20:41:18 "[ 2717.956047] ttyS ttyS0: tty_port_close_start: tty->count = 1 port count = 2 2018-03-21 20:41:18 " 2018-03-21 20:43:42 does your inittab have a serial port getty active? 2018-03-21 20:44:51 it sounds like something is opening/closing the tty device 2018-03-21 20:45:54 bingo 2018-03-21 20:47:34 https://pastebin.com/62N3qtv1 2018-03-21 20:48:25 didn't change anything in the configs, so maybe the image should be edited :o? 2018-03-21 20:55:21 very minor but opened a bug for it : https://bugs.alpinelinux.org/issues/8704 2018-03-21 23:23:21 I'm having some troubles upgrading from v2.7.9 to v3.x. The wiki lists a "few extra steps are necessary" but I haven't been able to find them so far. There was a link on the release notes for v3 but it 404s. 2018-03-21 23:23:50 Unfortunately I cannot reinstall this box either otherwise I would. 2018-03-22 01:15:39 Anyone else having trouble hitting the alpine package CDN? 2018-03-22 01:15:50 I'm getting "503 Backend is unhealthy" 2018-03-22 01:27:00 clandmeter: ^ 2018-03-22 01:27:04 idk what the backend is now 2018-03-22 01:40:54 has anybody succeeded in doing a sys install on a raspberry pi zero? I created an SD card as per the wiki instructions here (https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi), but the kernel doesnt seem to boot. It starts up and then powers itself off after about 2s. I was able to get raspbian working on it just fine, so I know the board works. 2018-03-22 08:22:40 How do I obtain a updated version for libtasn1 Alpine 3.7 x86_64 ? According to CVE-2018-6003 the problem is fixed in version 4.13. 2018-03-22 08:27:51 <_ikke_> hansthorsen: https://git.alpinelinux.org/cgit/aports/commit/?id=ebfdfada3106e7a72abccf4be1f6ea0defae3dcf 2018-03-22 08:27:59 <_ikke_> so the fix has been backported 2018-03-22 08:28:13 <_ikke_> 4.12-r3 2018-03-22 08:30:13 Thanx :-) 2018-03-22 11:03:31 Is there a cloudimage for alpine_ 2018-03-22 11:03:34 ? ) _ 2018-03-22 11:18:49 Bodenhaltung, what would that consist of :)? 2018-03-22 11:19:56 ssh, and dhcp? xD 2018-03-22 11:20:37 I think i have really a logical problem with the alpine iso vs. openstack volumes 2018-03-22 11:20:57 I need coffeine and/or alk....maybe fresh air...strange day. 2018-03-22 11:22:24 The VM iso is 37MB :) works pretty fine for me 2018-03-22 11:22:28 what do you mean logical problem 2018-03-22 11:38:20 If i boot the vm image...i have to "install" on a disk? Or 2018-03-22 11:38:43 ...or to do the setup scripts? 2018-03-22 11:40:14 Ah, wait..i get it now i think...i boot the image and attache the *new" volume to install on. haha. brb. 2018-03-22 14:03:29 Hello! Got a question regarding the package mongodb on edge, it appears that a dependency: `so:libyaml-cpp.so.0.5` got renamed (its now called `so:libyaml-cpp.so.06`) and so a rebuild of the package mongodb should probably fix that. How can I request that? 2018-03-22 14:06:20 hei, is anyone using rls-preview (rust component) on alpine? I see that there is no package for it, so i was about to build one myself and was wondering if anyone had any previous experience with it (rls-preview) on alpine. 2018-03-22 14:09:05 schnitzel, fastest way would be to directly contact the maintainer. Easiest way would be to file an issue at bugs.alpinelinux.org 2018-03-22 14:09:22 @bernhardgruen: allright! thank you 2018-03-23 02:57:19 Anybody using 9pfs with xen1:~# apk info xen-hypervisor xen-hypervisor-4.9.1-r4 description: Xen hypervisor 2018-03-23 02:58:23 Could you please share cfg file for the DomU? 2018-03-23 02:59:07 I am particularly interested in sharing common ZFS data pool among multiple DomUs 2018-03-23 02:59:40 Anybody running ZFS NAS on Alpine? 2018-03-23 04:51:46 How difficult is it to build a complete, official-style release tarball of Alpine? Namely, with custom -march 2018-03-23 04:57:06 tarball ? 2018-03-23 05:04:01 Actually it's an iso file, but really I'm interested in knowing if we can rebuild all packages in an automated way. 2018-03-23 05:13:59 yes, buildrepo will do it 2018-03-23 05:14:11 just set up abuild, clone aports.git and go to town 2018-03-23 05:15:52 kaniini, oh, that does sound pretty automated. debians sbuild is great for building a package or two but rebuilding the world is just a bear 2018-03-23 08:21:52 good morning folks 2018-03-23 08:21:56 a quick question 2018-03-23 08:22:13 the xen image is to run alpine as host or guest in Xen? 2018-03-23 08:28:47 nevermind found the solution 2018-03-23 08:29:52 btw if you want to build a aarch64 image for the rpi3, here I mage a simple guide https://gist.github.com/psammarco/1a444019869eb7acedb9b4bf495cc831 2018-03-23 10:02:38 I'm new to alpine. I want to install to disk, but i also want to use ZFS for the root partition. 2018-03-23 10:02:47 Does anyone know how to go about doing this? 2018-03-23 10:03:29 hi snappy, welcome 2018-03-23 10:03:35 use the extended iso 2018-03-23 10:04:05 i dont remember if we added zfs support to setup-alpine 2018-03-23 10:04:18 (i think not) 2018-03-23 10:04:41 what you do is run setup-alpine, select "none" on disk selection 2018-03-23 10:04:56 then you manually install zfs tools, create a zfs partition 2018-03-23 10:05:03 mount it under /mnt (forexample) 2018-03-23 10:05:17 and finally you run setup-disk /mnt 2018-03-23 10:05:18 i chose the standard iso, can i still proceed with that? (it's a hassle because i'm remotely setting this up) 2018-03-23 10:05:41 standard iso does not include zfs modules iirc 2018-03-23 10:05:53 i can just apk add them 2018-03-23 10:06:02 no you cant 2018-03-23 10:06:07 (i think) 2018-03-23 10:06:44 well maybe you can but you will have to mount overlayfs to work around read-only /lib/modules 2018-03-23 10:07:06 ok, i'll see about getting the extended iso 2018-03-23 10:11:25 apk says that there's an error, how do i check what that error actually is? 2018-03-23 10:12:52 i gues just apk fix and see what it does 2018-03-23 10:22:31 ncopa: when you do setup-alpine, you select "none" for disks, and it asks where do i want to store configs, do I also say none? 2018-03-23 11:24:14 ncopa: sorry to bother, but do you know how to layout the partition for a disk if doing zfs? I want to do ZFS raid-1, but i think i need separate /boot partition for the bootloader 2018-03-23 12:08:26 gpart show for the ZFS on the root (FreeBSD) http://collabedit.com/apgeh 2018-03-23 12:13:15 things are a bit more complicated 2018-03-23 12:13:40 with Linux we need the boot partitions as well 2018-03-23 12:35:22 snappy: I am very sure about zfs, but setup-disk should do the layout for you. 1st partition for /boot, 2nd for swap, 3rd for / 2018-03-23 12:35:27 *not very sure 2018-03-23 12:36:09 setup-disk also has option for you to do raid, although I am not well-versed with raid also 2018-03-23 12:36:15 give it a try with setup-disk 2018-03-23 13:40:31 hello folks, I've installed kubernetes v1.8.3 on alpine 3.7 and used flannel v0.9.1 as a network pod. after successfully installing the flannel with 'kubectl apply' command, the kubernenets pod k8s-dns-sidecar couldn't start. please has any one of you experienced this problem. PS I reproduced the setup on ubuntu and centos and it works like a charm. thank you 2018-03-23 13:41:01 for log files and pod logs: https://github.com/wajdihajji/kubernetes-alpine 2018-03-23 13:41:45 anyone please tested k8s and flannel on alpine? 2018-03-23 13:43:35 if i'm posting in the wrong channel, please let me know where can I ask for help on my inquiry 2018-03-23 14:08:27 cmpwhajj__: I assume its a missing bin 2018-03-23 14:08:30 is nsenter installed 2018-03-23 14:09:26 did you configure / install CNI? 2018-03-23 14:19:05 trying to boot alpine usb stick on my laptop as recovery and isolinux hangs after boot: prompt 2018-03-23 14:19:10 any ideas? 2018-03-23 14:19:52 could be missing framebuffer 2018-03-23 14:20:27 in which case it does boot, just does not show anything on the screen 2018-03-23 14:20:32 any cmdline opts to try? 2018-03-23 14:20:53 you can try type: rootmodprobe fbcon 2018-03-23 14:21:23 i wou'd think with no fb id still geg a few lines before swirching out of fb 2018-03-23 14:21:39 theres not eveb a loadibg vmlinux or anything 2018-03-23 14:25:40 hello 2018-03-23 14:26:20 I am trying to install alpinelinux on a apu2c4 via serial, though I can't seem to be able to do it 2018-03-23 14:26:40 I tried enabling serial as suggested here https://wiki.alpinelinux.org/wiki/Enable_Serial_Console_on_Boot 2018-03-23 14:27:16 though I cant modify the boot partition from the usb install since it has cdrom format 2018-03-23 14:27:46 I guess I will create the layout manually 2018-03-23 14:27:58 Latrina, 2018-03-23 14:28:06 hello 2018-03-23 14:28:06 what image are you using ? 2018-03-23 14:28:23 standard x86_x64 2018-03-23 14:28:50 You could try the VM image, it has serial enabled by default 2018-03-23 14:28:53 is there a specific image for serial boot ? 2018-03-23 14:29:14 but will that be okay kernel wise? 2018-03-23 14:29:25 Depends on what you want :) 2018-03-23 14:29:57 will the idea is to run xen on it 2018-03-23 14:30:12 so I will boot the xen kernel ultimately 2018-03-23 14:32:40 as far as I can see there is nog difference kernelwise between VM and XEN image 2018-03-23 14:32:44 not* 2018-03-23 14:32:48 hmm -t * 2018-03-23 14:32:55 uhg forgot i need edd=off 2018-03-23 14:35:20 Harm133: awesome 2018-03-23 14:35:56 well again, since I will be running Xen kernel (dom0) I should be okay I guess 2018-03-23 14:36:18 btw, does the VM image comes with some VM related tools or anything like that? 2018-03-23 14:37:30 Nope :) 2018-03-23 14:37:35 should be installable 2018-03-23 14:50:39 ncopa, edd=off should probably be in default boot args 2018-03-23 14:57:14 ncopa, the Kconfig for it explicitly says: 2018-03-23 14:57:19 This option is experimental and is known to fail to boot on some 2018-03-23 14:57:19 obscure configurations. Most disk controller BIOS vendors do 2018-03-23 14:57:19 not yet implement this feature. 2018-03-23 14:57:26 and it does not seem useful for anything 2018-03-23 14:57:40 (CONFIG_EDD) 2018-03-23 14:59:04 it makes at least my laptop (with no sata drives at all, just emmc, that's probably related) fail to boot at all with no error output indicating why. older kernels printed a message "Probing EDD (edd=off to disable)..." 2018-03-23 14:59:10 that's the only thing that reminded me 2018-03-23 15:00:43 ahhhh the "quiet" on the default alpine commandline is what suppresses the message 2018-03-23 15:00:46 *sigh* 2018-03-23 15:51:53 @deedubs the problem is fixed, it was because of nsenter 2018-03-23 15:52:00 thanks a lot !! 2018-03-23 16:41:40 dalias: there is an 'noquiet` option 2018-03-23 16:57:56 no problem cmpwhajj__, It's a problem space I'm working in as well 2018-03-23 17:10:17 ncopa, yep or you can just edit the commandline to remove quiet 2018-03-23 17:10:26 but i'm thinking especially about the isos shipped 2018-03-23 17:10:36 "fails to boot with no output at all" is a bad failure mode 2018-03-23 17:11:08 is there some useful functionality CONFIG_EDD is enabled for, or is it just bug surface with no benefit? 2018-03-23 20:17:49 i installed st but there is no terminfo entry; i think this is a bug 2018-03-23 20:19:34 markweston: install ncurses-terminfo 2018-03-23 20:21:51 it's already installed 2018-03-23 20:22:29 try it for yourself: apk add st; ls /etc/terminfo/s 2018-03-23 20:23:51 apk info -L ncurses-terminfo 2018-03-23 20:24:28 it is in /usr/share/terminfo not /etc/terminfo 2018-03-23 20:25:18 my bad 2018-03-23 20:25:25 then there is a bug in micro pachage 2018-03-23 20:25:34 it doesn't find st entry 2018-03-23 20:26:15 ncurses-terminfo-base contains small number of terminfo's 2018-03-23 20:26:44 and, ncurses-terminfo contains rest 2018-03-23 20:27:39 and micro package (a text editor) can't find terminal entry despite ncurses-terminfo being installed 2018-03-23 20:29:04 don't know why the text editor have to look after terminfo, that should be done by terminal emulator 2018-03-23 20:29:27 it's probably because the text-editor's fault 2018-03-23 20:30:07 I'm not expert in terminfo field so I cannot comment about that 2018-03-23 20:31:37 it doesn't use ncurses, i will submit a bug report on micro's github, see what they think 2018-03-23 21:43:09 mps: terminfo is used by programs interacting with the terminal 2018-03-23 21:43:48 so, the xterm terminfo describes how to interact with xterm 2018-03-23 21:44:34 naturally, i'd expect a text editor (CLI) to look after terminfo, while xterm doesn't need to 2018-03-23 21:45:59 for some reason, alpine doesn't have the terminfo for st packages 2018-03-23 21:46:52 so, its not a bug in terminfo, its more a bug in alpine for not shipping the necessary terminfo for st 2018-03-23 21:46:59 which is included in upstream 2018-03-23 21:47:08 *not a bug in micro editor 2018-03-23 21:49:33 ah, nevermind, i think i misunderstood the issue 2018-03-23 21:53:46 liwakura: I wrote that I'm not expert in terminfo field, but alpine has st (and variants) terminfo data in ncurses-terminfo 2018-03-23 21:53:55 indeed, i was wrong there 2018-03-23 21:54:14 ah, ok, nevermind ;-) 2018-03-23 21:54:16 https://github.com/zyedidia/micro/issues/20 2018-03-23 21:54:24 the bug seems to have been fixed 2018-03-23 21:55:11 but only included in 1.4.1, which is a pre-release 2018-03-23 21:55:25 alpine is still on 1.1.3 2018-03-24 00:27:54 Hey, is there anyone that could walk me through the intricacies of doing zfs on root for alpine. I'm not exactly sure how the partition layout should be, and how to then setup the disk using setup-disk during installer. 2018-03-24 00:52:56 snappy: I am well-versed in zfs on Alpine but pretty fluent with setup-disk. as ncopa said earlier, you should use extended image (or just standard? you should try yourself), run setup-alpine, and stop at installing on disk step (choose 'none' when asked for disk), install zfs tools, then run setup-disk /mnt 2018-03-24 00:53:09 *not well-versed... jesus ... 2018-03-24 00:53:37 you have any problem so far with setup-disk ? 2018-03-24 01:07:35 the problem i have is 2018-03-24 01:08:37 you can't boot directly to a ZFS disk, so there needs to be a partitioning layout, but i'm not sure exactly what needs to be in the partition map 2018-03-24 01:17:44 that was just discussed earlier today if I'm not mistaken 2018-03-24 01:18:20 sorry, i must have missed it, do you mind pasting to me in PM the relevant part 2018-03-24 01:18:30 I'll just quickly recap it 2018-03-24 01:19:06 I think the extended iso has what it takes to init your zfs root, then just mount it and run setup-disk 2018-03-24 01:19:53 the boot thing I think you can do in several ways depending on what you prefer to use to boot; most commonly you need a /boot outside of your zfs root 2018-03-24 01:20:51 but even that's not necessary if you just use your kernel as a boot loader, which you can do in an UEFI capable system 2018-03-24 01:23:07 The extended iso doesn't have a way to init zfs root, i have to do that manually 2018-03-24 01:53:41 oh wait, it was you who had a conversation with ncopa about this weren't you? 2018-03-24 02:03:33 correct 2018-03-24 02:15:02 okay, I'm pretty sure I can't match his knowledge on the topic :) 2018-03-24 02:16:13 snappy: the extended iso should have zfs kernel module + zfs userland application : https://git.alpinelinux.org/cgit/aports/tree/scripts/mkimg.standard.sh 2018-03-24 02:16:41 I thought clandmeter answered this a couple of days ago (to you? or somebody else with same question on zfs ) 2018-03-24 02:16:59 can you output specific log ? 2018-03-24 02:17:09 or anything specific ? 2018-03-24 02:17:09 ncopa did answer this, but my question was more to do on how to prepare the partitioning layout 2018-03-24 02:17:20 I just answered above :| 2018-03-24 03:35:34 hello. is llvm 6 coming to alpine edge soon? 2018-03-24 04:58:41 is there any armel (armv5) builds of Alpine? 2018-03-24 08:53:42 abuild -F fails with "do not run abuild as root" 2018-03-24 09:02:03 it does check if the flag -F is set, then sets forceroot="-F", then [ -z "$forceroot" ] && die "Do not run abuild as root" 2018-03-24 09:02:09 sounds strange 2018-03-24 09:03:24 i don't see why would you want to put arbitrary restriction on in the first place 2018-03-24 09:07:10 I once had written a Makefile which should have delete the build dir on clean. Unfortunately I wrote 'rm -rf "$build_dir" /'. This would have had a bad effect if ran as root. And I assume such things happen more often than we think. 2018-03-24 09:27:54 often 2018-03-24 09:28:34 well thank you for taking care of me and everyone else 2018-03-24 09:30:06 anyway, the quick fix is to remove the root check 2018-03-24 09:30:21 does fakeroot not work for whatever you're trying to build? 2018-03-24 09:30:48 i wont bother making a new user 2018-03-24 09:31:22 and if anything messes up i just reinstall; everything is on github anyway 2018-03-24 09:34:00 Hallo, can anyone help me? 2018-03-24 09:34:26 I try to get alpine linux running on my odroid c2 board 2018-03-24 09:34:57 u-boot is starting and also the kernel 2018-03-24 09:35:16 but then the boot process stops 2018-03-24 09:35:21 with this message: 2018-03-24 09:37:06 Begin: Running /scripts/init-bottom ... done. 2018-03-24 09:37:22 can't run '/sbin/openrc': No such file or directory 2018-03-24 09:50:06 how do i make a new package without cloning the aports repo? 2018-03-24 09:51:58 https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2018-03-24 09:52:10 you don't *need* the aports tree 2018-03-24 09:52:26 but the rest of the steps are valid 2018-03-24 10:02:21 the abuild script gets stuck at dir_has_arch_binaries() when trying to split up doc subpackage 2018-03-24 10:16:23 I used scripts from: 2018-03-24 10:19:38 https://github.com/umiddelb/z2d/tree/master/odroid-c2 2018-03-24 10:19:38 but this scripts install alpine-minirootfs 2018-03-24 10:19:39 and there is no openrc 2018-03-24 10:19:40 how to add openrc to alpine-minirootfs 2018-03-24 10:19:40 ? 2018-03-24 10:24:38 specifically it hangs on scanelf -R $dir 2018-03-24 10:28:18 gerald, I 2018-03-24 10:28:27 I'm not sure, but can't you do apk add openrc? 2018-03-24 10:30:53 also, why would you want openrc within an docker image? 2018-03-24 10:37:28 markweston: i did some research regarding your micro issue 2018-03-24 10:37:58 its already fixed in the git tip of micro (issue #20), and will be released in 1.4.1 2018-03-24 10:38:21 correction bc of algitbot: https://github.com/zyedidia/micro/issues/20 2018-03-24 10:38:41 alpine is still on 1.1.-something 2018-03-24 10:42:11 Hallo Harm133, i don't install alpine in docker 2018-03-24 10:42:29 i installed it on a SD for the odroid c2 board 2018-03-24 10:42:54 what i found out is that these scripts install alpine-minimal 2018-03-24 10:43:19 mount the rootfs and enter in a chroot 2018-03-24 10:43:44 and install busybox-initscripts in there 2018-03-24 10:44:02 the problem is, chroot is not working by me 2018-03-24 10:44:12 it tells me: cannot find sh 2018-03-24 10:44:19 liwakura: okay, i will probably have to submit a pr for upgrade 2018-03-24 10:44:43 i think i will try this scripts with the alpine-uboot image 2018-03-24 10:45:52 but there the file structure is different 2018-03-24 10:46:38 markweston: 1.4.1 upstream is an pre-release, so i would wait until its fully released 2018-03-24 10:47:23 on the other hand, i dont know why go projects need to invent a wrapper around terminfo (tcell) instead of just using terminfo 2018-03-24 10:55:55 yeah, so abuild gets stuck at scanelf -R /root/aports/main/sic/pkg/sic-doc 2018-03-24 10:56:31 gerald, I think I cant help you with that :) the only suggestion I wouldve given you is trying the Uboot image 2018-03-24 10:56:45 i don't know what to do, it worked when i wasn't root 2018-03-24 10:57:35 and scanelf doesn't get stuck when i call it standalone in shell 2018-03-24 10:57:50 i mean when i call it outside abuild 2018-03-24 10:58:46 i guess i will have to create a user; this really, really pisses me off 2018-03-24 10:59:13 markweston: tried strace'ing it? 2018-03-24 10:59:18 to see what it does? 2018-03-24 10:59:23 no 2018-03-24 11:02:22 it's stuck on wait4 2018-03-24 11:02:50 strace -f, its an forked off process then thats hanging 2018-03-24 11:03:01 -f for recursing into child processes 2018-03-24 11:08:08 it's the same 2018-03-24 11:10:22 semop(0, [{0, -1, SEM_UNDO}], 1 2018-03-24 11:11:32 has anyone tried build packages in root? 2018-03-24 11:11:48 *building packages on root 2018-03-24 11:13:18 semop is related to locking.. 2018-03-24 11:18:25 i don't think it's worth investingating it though, i think it will be quicker to just make a user 2018-03-24 11:19:19 although im dissapointed i can't make packages as root 2018-03-24 11:20:22 whats the resoning for trying to do it as root? 2018-03-24 11:22:10 when you run abuild as root you allow arbitrary code execution to the buildscripts, if anyone wants to mess with you it's much easier. also just rm-rf-ing. you should not run stuff as root generally. 2018-03-24 11:23:16 hs3dUBwdmCjy: build scripts are arbitrary code execution by design 2018-03-24 11:23:29 at least with the current "download & execute" mentality 2018-03-24 11:24:30 Thank you ... i will try the uboot image 2018-03-24 11:28:20 and how would one "mess with me"? could you provide a practical example? 2018-03-24 11:28:38 or are you reciting what you've read on stackoverflow? 2018-03-24 11:30:35 markweston: it could install stuff via npm v5.7.0 2018-03-24 11:30:46 so your system might not boot again 2018-03-24 11:31:27 or do some syslinux stuff and accidentaly write to your hosts extlinux.conf 2018-03-24 11:32:43 liwakura: i need to run that program to begin with so it could do anything 2018-03-24 11:33:05 and i don't run programs i don't know 2018-03-24 11:33:07 im just talking about unintentional bugs 2018-03-24 11:33:30 huh, i dont think anyone here has read the code of what alpine consists of 2018-03-24 11:33:47 just segments when something broke.. 2018-03-24 11:34:14 well okay, i meant it more like "i don't run programs i don't trust" 2018-03-24 11:35:00 currently trying to manually compile lvm 2018-03-24 11:35:14 and its a huge pain... 2018-03-24 11:40:58 is it standard practice on distros to have sshd running and remote login enabled? 2018-03-24 11:43:34 depends on the distro 2018-03-24 11:43:55 it makes sense for servers, but less for desktop 2018-03-24 11:59:49 markweston: easiest non-malicious example is rm-rfing /. and no i did not recite stackoverflow. i predate s.o. by quite some time ;) 2018-03-24 12:01:35 never happened for me, and if it does, i keep everything of any significance on the cloud, so i have no fear of rm -rf / 2018-03-24 12:01:54 liwakura: "build scripts are arbitrary code execution by design" - but that doesn't have to be so actually in most cases, its a very unfortunate technical debt we are carrying. 2018-03-24 12:02:28 rob pike also keeps everything on the cloud btw 2018-03-24 12:03:29 but does rob pike run stuff unnecessarily as root? 2018-03-24 12:04:40 i don't know and i don't really care - i prefer the inconvenience of having to reinstall everything (which never happened to me to this time) - to typing sudo and password all the time 2018-03-24 12:06:08 i can understand if this is your server, but geez, we are talking about personal computers 2018-03-24 12:06:52 wipe and restart in an hour at any time - anything important is already pushed on github or whatever you use 2018-03-24 12:19:07 so mark pushes his certificates and secret keys to github if i understand correctly what he writes. 2018-03-24 14:57:14 hs3dUBwdmCjy: rob pike wouldnt have a root user in the first place. namespace separation is all that's needed. 2018-03-24 15:50:26 the cloud otoh is never a solution. too many external dependencies that will result in crapy service or total meltdown. rob sadly hasn't grasped that yet. 2018-03-24 15:52:01 this is anyways only possible to huge inbalances of uplink cost and throughput. most people on earth don't have a cloud in the first place. the only ones that do have acces would be able to use alternatives, but they cannot think for themselves... 2018-03-24 16:17:22 why does alpine-sdk depend on acct? 2018-03-24 17:20:23 hi 2018-03-24 17:40:57 hi 2018-03-24 17:43:25 hi 2018-03-24 17:43:30 hi 2018-03-24 17:46:10 liwakura: i remember you said you use a text browser as a daily driver; how do you replace youtube? is it some combination of links + mpv + youtube-dl? 2018-03-24 17:49:51 actually haven't used w3m in quite a while 2018-03-24 17:50:12 but i used to peek at urls, and just click them (im still in xorg= 2018-03-24 17:50:30 urxvt then fires an custom xdg-open script, which detects the youtube url and runs mpv on it 2018-03-24 17:50:57 https://github.com/nero/scripts/blob/master/xdg-open 2018-03-24 21:25:34 I'm finding a reference to 'aterm' in aports but it's not in the repositories... is it just some are not available for x86? 2018-03-24 21:25:44 or are some of them FTBFS 2018-03-24 21:28:03 TomTheDragon: what kind of reference do you see, and on which branch of aports.git? 2018-03-24 21:28:31 jn__, just saw a reference to it here: https://git.alpinelinux.org/cgit/aports/diff/main/aterm/APKBUILD?id=5c904717322f8113870abf48cc93c3c8b00ab4e8 2018-03-24 21:28:46 looking at the date, guess it could have been pulled easily 2018-03-24 21:33:35 TomTheDragon: aterm has been removed, upstream seems dead 2018-03-24 21:34:13 in commit v2.7.0_rc2-53-g6e4a21f72c 2018-03-24 21:35:18 liwakura, ah, I see... I take it that it's fairly easy to revert, if we're willing to build our own pkgs 2018-03-24 21:36:14 however, I'm up for suggestions on very lightweight terminal emulators (all I really care about personally is keyboard-based scrolling) 2018-03-24 21:36:32 i use urxvt 2018-03-24 21:36:41 scrolling with shift + PgUp/PgDown there 2018-03-24 21:37:40 aterm website also recommends rxvt-unicode 2018-03-24 21:38:24 hmm, apparently aterm is based on rxvt? wonder why it seems to use less memory than anything else I've tried 2018-03-24 21:41:13 more like, derived from 2018-03-24 21:52:34 I'm glad that fvwm is still there... just got spoiled by its configurability, as opposed to something like xfce4 2018-03-24 22:04:18 i feel content with i3 2018-03-24 22:04:41 xfce config behaves not so well in git 2018-03-24 22:15:45 Hello. Who do I ask about whether the Alpine maintainers want a non-GNU bc? 2018-03-24 22:17:18 for what? 2018-03-24 22:19:04 As an Alpine package. And to build the kernel. 2018-03-24 22:26:17 i've been using aterm for a long time, in the belief that it's lighter than urxvt 2018-03-24 22:26:43 but i hate it so much. 2018-03-24 22:27:22 there are really only two viable terminal emulators, with slightly different things they emulate best: urxvt, xterm. everything else has lacking support. 2018-03-24 22:27:38 i prefer urxvt, because it responds better to resizing the window size. 2018-03-24 22:27:52 also because i use urxvt's history searching functionality 2018-03-24 22:41:41 I've switched over to i3 too, but that doesn't solve the "which terminal" question... currently I use mate's terminal. I've briefly used evilvte at some point, can't remember why I got rid of it 2018-03-24 22:42:10 of course on modern hardware without memory limitations, it's Konsole all the way 2018-03-24 22:43:26 Installing g++ I just got "BAD signature" 2018-03-24 22:43:48 hiro, I'd use xvt for most stuff if it just had keyboard-based scrolling 2018-03-24 22:43:56 it emulates... vt100. pretty much nothing else. 2018-03-24 22:44:28 :D 2018-03-24 22:44:54 i don't see why Konsole would be better than urxvt 2018-03-24 22:45:09 what do you miss there?! 2018-03-24 22:45:23 most of the time I don't care about colors, or anything but 7bit ASCII 2018-03-24 22:48:40 I can't afford to stick to 7 bit ASCII thanks to living in a country that depends on more than that 2018-03-24 22:49:58 true, many will require UTF8 2018-03-24 22:50:05 as for Konsole ... it's basically my preference only because during the years I've used it, it has never given me problems of any sort 2018-03-24 22:51:29 the only real problem with Konsole is, installing it anywhere drags half the world along with it... 2018-03-24 22:51:53 how accurate its emulation is, that I can't really comment on 2018-03-24 22:53:27 anyone know why g++ is borked in edge? I had a feeling using edge would cause issues like that, but just wanted an up-to-date devel environment 2018-03-24 22:54:01 a possibility it's just the mirror? 2018-03-24 22:54:02 that bad signature problem is probably just a mirror sync related thing 2018-03-24 22:54:22 seems to happen every once in a while 2018-03-24 22:54:42 something that I have to wait for it to resolve? 2018-03-24 22:55:03 my guess is, either that or changing the mirror you use 2018-03-24 22:55:39 will try if it fails again 2018-03-24 23:24:10 what metapackage similar as "base-devel" in arch or "build-essential" in debian? 2018-03-24 23:24:35 build-base 2018-03-24 23:24:42 I think 2018-03-24 23:25:05 it works, thanks 2018-03-24 23:34:45 Is there a way to download .apk files from a different OS? 2018-03-24 23:35:14 they're files on a web server; sure 2018-03-24 23:36:29 yeah, but, where are they? 2018-03-24 23:41:44 I found it 2018-03-24 23:42:08 dl-cdn.alpinelinux.org 2018-03-25 00:25:45 Roserin, it's pretty easy to setup a chroot as well. Not a ton of use for the files without being able to install them. 2018-03-25 00:40:32 I'm just curious, can we run normal Linux binary applications in Alpine? 2018-03-25 00:41:31 TomTheDragon: What do you mean by "normal"? 2018-03-25 00:42:24 Do you have an example of such a program? 2018-03-25 00:43:28 has anyone gotten synapse (matrix.org server) built into an APKBUILD yet? 2018-03-25 00:43:32 Anything linked with glibc, e.g. steam etc. 2018-03-25 00:43:57 TomTheDragon: there's a package called gcompat, which helps with that, but i don't know the details 2018-03-25 00:44:50 I haven't found anything not open source I actually care to run, but might need to at some point. 2018-03-25 00:47:17 another approach would be to put the binary and glibc in a container/chroot/etc., but that might be difficult with GUI programs 2018-03-25 01:59:21 Invalid configuration `i586-alpine-linux-musl': machine `i586-alpine-linux' not recognized 2018-03-25 01:59:26 why would I be getting that? 2018-03-25 02:00:21 TomTheDragon: what are you doing? 2018-03-25 02:00:37 abuild'ing a package 2018-03-25 02:04:45 I figure i'm missing a pkg 2018-03-25 02:10:42 what is it you're building? 2018-03-25 02:10:52 it's possible that the configure script or whatever just has a problem 2018-03-25 02:11:01 especially if it's largely glibc-centric software 2018-03-25 02:35:42 maybe the config.{guess,sub} files of that packages haven't been updated in a long time, so it doesn't know what musl is 2018-03-25 02:45:29 yeah that's plausible 2018-03-25 04:33:09 TomTheDragon: it's what jn__ and dalias said 2018-03-25 04:33:17 you need to update the config.sub/guess files 2018-03-25 04:33:22 abuild has a builtin function for that you can call: 2018-03-25 04:37:09 prepare() { 2018-03-25 04:37:14 default_prepare 2018-03-25 04:37:17 update_config_sub 2018-03-25 04:37:19 } 2018-03-25 04:37:25 try adding that to your APKBUILD (with proper indentation) 2018-03-25 04:38:17 and maybe also update_config_guess 2018-03-25 05:40:54 Shiz, thanks, trying that 2018-03-25 06:51:41 How do you install musl-dev? 2018-03-25 06:52:11 I am trying to install it offline, and if I install it it removes any packages that depend on musl 2018-03-25 06:52:20 which is almost all of them 2018-03-25 07:03:01 probably because it depends on a specific version of musl 2018-03-25 07:03:09 and whichever musl-dev you downloaded is newer/older 2018-03-25 07:03:37 Makes sense 2018-03-25 07:03:44 just a wild guess, but easy enough to verify 2018-03-25 07:04:05 so, would getting a newer version of musl fix that? 2018-03-25 07:04:32 you should check what version of musl you have now, and get the matching musl-dev 2018-03-25 07:05:00 I have the 7.3.0 release version 2018-03-25 07:05:04 iirc 2018-03-25 07:05:34 # apk info musl 2018-03-25 07:05:35 musl-1.1.18-r3 description: 2018-03-25 07:05:42 seems to be the latest on 3.7.0 2018-03-25 07:05:54 oops 3.7 2018-03-25 07:06:44 Mine has 1.1.18-r2, if it makes much of a difference 2018-03-25 07:07:01 do you know where I could get the equivalent musl? 2018-03-25 07:07:14 what arch are you on? 2018-03-25 07:07:15 x86_64? 2018-03-25 07:07:21 Yep 2018-03-25 07:09:01 well 2018-03-25 07:09:06 official mirrors dont seem to have r2 2018-03-25 07:09:11 but for some reason its on my mirror /shrug 2018-03-25 07:09:21 https://mirror.astr.al/alpine/v3.7/main/x86_64/musl-dev-1.1.18-r2.apk 2018-03-25 07:10:06 And to add a bit of confusion, the pkgs.alpinelinux.org have .19-r1 2018-03-25 07:10:25 that's on edge 2018-03-25 07:10:26 not 3.7 2018-03-25 07:10:40 Yea 2018-03-25 07:17:55 Hey it worked maybe 2018-03-25 07:24:48 So I was nearly there 2018-03-25 07:25:09 Ima finish up tomorrow 2018-03-25 07:25:13 C yall 2018-03-25 07:26:00 seeya 2018-03-25 07:27:06 Thanks for the package 2018-03-25 07:27:19 np :) 2018-03-25 10:13:20 hello 2018-03-25 10:13:29 how is using linux without the gnu parts? 2018-03-25 11:45:34 Hallo i wanted to ask how i can start a service in alpine linux without openrc 2018-03-25 11:45:47 i install the minirootfs image 2018-03-25 11:45:53 and openrc is not installed 2018-03-25 11:46:05 i don't really see how you'd run a service without something to actually manage the service 2018-03-25 11:46:15 but i have no network connection 2018-03-25 11:48:09 in the minirootfs image openrc is not installed 2018-03-25 11:48:34 but i cannot install it because i have no network connection 2018-03-25 11:50:52 gerald: so you want to setup network at first? 2018-03-25 11:51:01 yes 2018-03-25 11:51:26 i think editing /etc/network/interfaces and then ifup'ing the device could do it 2018-03-25 11:51:45 how do you run the image? 2018-03-25 11:52:22 i think your problem is actually some other 2018-03-25 11:52:53 thank you, ifup has done it 2018-03-25 11:53:08 i installed alpine linux on a SD 2018-03-25 11:53:18 for the odroid c2 2018-03-25 11:53:45 the regulare images don't work with the odroid c2 2018-03-25 11:53:47 how did you even boot it without openrc? 2018-03-25 11:54:50 someone from odroid wrote scripts to install alpine linux on a SD 2018-03-25 11:55:10 he make this with a chroot environment 2018-03-25 11:55:27 running on ubuntu on the odroid c2 2018-03-25 11:55:49 rc-status says: 2018-03-25 11:55:55 runlevel: sysinit 2018-03-25 11:56:16 bootmanager is u-boot 2018-03-25 11:56:29 runlevels are managed by openrc 2018-03-25 11:57:48 i deleted the apk cache file ... how can i rebuild them ... 2018-03-25 11:58:20 the /etc/apk/cache ? 2018-03-25 11:59:50 no in /var ... 2018-03-25 12:00:28 the folder /etc/apk/cache don't exists 2018-03-25 12:00:36 setup-apkcache 2018-03-25 12:01:02 setup-apkcache: not found 2018-03-25 12:01:17 apk add alpine-conf 2018-03-25 12:01:25 actually 2018-03-25 12:01:30 just do apk add alpine-base 2018-03-25 12:01:32 you shouldn 2018-03-25 12:01:48 *you should not just start with an plain minirootfs on an actual device 2018-03-25 12:02:25 yes i know ... but i know no other way to start it for odroid c2 2018-03-25 12:02:28 especially when you used a script for installing it on a SD card 2018-03-25 12:02:41 the uboot image for aarch64 don't boots 2018-03-25 12:02:58 also, how are you even able to use the minirootfs? did you boot the odroid? or an chroot? why do you need manual network then? 2018-03-25 12:04:11 also you could post an link to the script "someone from odroid wrote" 2018-03-25 12:05:42 https://github.com/umiddelb/z2d/tree/master/odroid-c2 2018-03-25 12:06:15 alpine linux only runs on the raspberry pi 2018-03-25 12:06:35 but with this scripts i can start it also on my odroid c2 2018-03-25 12:06:58 WARNING: Ignoring APKINDEX.70c88391.tar.gz: No such file or directory 2018-03-25 12:07:13 WARNING: The repository tag for world dependency 'most@testing' does not exist 2018-03-25 12:07:26 ERROR: Not committing changes due to missing repository tags. Use --force-broken-world to override. 2018-03-25 12:07:52 please use an external paste site 2018-03-25 12:09:43 gerald: so you are still running in a chroot? 2018-03-25 12:10:08 no, alpine linux is now running 2018-03-25 12:10:16 on my odroid c2 2018-03-25 12:10:29 i am connected with a serial console 2018-03-25 12:10:55 and can interact wiht alpine linux on my odroid 2018-03-25 12:11:06 what i can see 2018-03-25 12:11:20 alpine linux boots up normaly without errors 2018-03-25 12:12:34 now the network is running 2018-03-25 12:12:55 only apk package management has to get working 2018-03-25 12:13:01 the script do not setup any bootloader-related stuff up 2018-03-25 12:13:24 so i have no idea why your stuff is working at all 2018-03-25 12:14:15 alpine.00.sh is a link to ubuntu-core-00.sh 2018-03-25 12:14:43 the script writes uboot with dd to the SD 2018-03-25 12:15:26 alpine-01.sh chroots in the directory 2018-03-25 12:16:01 alpine-02.sh updates alpine 2018-03-25 12:16:25 and alpine-03.sh installs basic stuff 2018-03-25 12:16:44 it installs busybox-initscripts 2018-03-25 12:16:54 so alpine starts with it 2018-03-25 12:17:17 not sure if that guy knows about "alpine-base" 2018-03-25 12:18:08 alpine-base where i can get it? 2018-03-25 12:18:21 apk add alpine-base 2018-03-25 12:18:35 minimal for self-running systems 2018-03-25 12:19:01 he should have used something like apk add -p /path/to/chroot alpine-base 2018-03-25 12:19:12 the minirootfs is intended for chroots and containers 2018-03-25 12:19:23 i could change the scripts 2018-03-25 12:20:15 i dont even want to know why he hardcodes the disk uuid to deadbeef-dead-beef-dead-beefdeadbeef 2018-03-25 12:20:22 gerald: my miniroot install is simple 2018-03-25 12:20:43 apk update; apk add alpine-base ; apk add openrc 2018-03-25 12:20:49 im so disgusted im considering getting an C2 just for the sake for adding better support for it 2018-03-25 12:21:00 mps: apk -p makes this all work without chroot 2018-03-25 12:21:27 https://wiki.alpinelinux.org/wiki/Installing_Alpine_Linux_in_a_chroot 2018-03-25 12:21:45 yes, I know 2018-03-25 12:22:16 but this i have to make in the chroot, or? 2018-03-25 12:22:49 because apk tells me: 2018-03-25 12:23:07 ERROR: Not committing changes due to missing repository tags. Use --force-broken-world to override. 2018-03-25 12:23:13 you can, as I do. but you can do that if you booted real machine 2018-03-25 12:23:41 alpine linux is now booted on the odroid c2 2018-03-25 12:23:50 but apk is not working 2018-03-25 12:25:54 gerald: im a bit angry because you are using (bad) scripts from someone else, and now we are the ones supporting it 2018-03-25 12:26:24 hi, but i don't know a other way to get it working 2018-03-25 12:26:28 I don't remember exact steps because I did that a long time ago, but something like that was how I installed AL on armhf and aarc64 machines 2018-03-25 12:26:35 is there a other way? 2018-03-25 12:27:03 the problem is with u-boot 2018-03-25 12:27:26 i tried it with alpine-uboot and the wiki 2018-03-25 12:27:32 I remember that I used debootstrap and chroot to make initial installation and then copied to sdcard 2018-03-25 12:27:32 but it don't boots 2018-03-25 12:28:20 with the scripts uboot is working 2018-03-25 12:28:30 but i don't need the scripts 2018-03-25 12:28:39 and, ofcourse, install appropriate u-boot and wrote extelinux.conf 2018-03-25 12:28:41 if there is a other way 2018-03-25 12:29:15 i am not a expert to manage this 2018-03-25 12:29:34 it would be great if there would be a wiki for that 2018-03-25 12:29:34 first thing is to ensure that u-boot works and can find extlinux.conf 2018-03-25 12:29:44 i only found this scripts 2018-03-25 12:29:57 someone in the odroid forum pointed me to it 2018-03-25 12:30:01 maybe I should write wiki page about that 2018-03-25 12:30:50 there are wiki page on wiki.alpinelinux.org about install alpine on arm from scratch 2018-03-25 12:31:06 yes, i used them 2018-03-25 12:31:16 but i did not get uboot working 2018-03-25 12:31:29 uboot is not well documented 2018-03-25 12:31:37 it is 2018-03-25 12:31:42 where? 2018-03-25 12:32:03 you should use u-boot from odroid for your board 2018-03-25 12:32:05 Anyone got any brilliant ideas with gcompat/libc6-compat and appimage binaries? When I try to run one it complains that it doesn't look like a squashfs image, and a simple strace http://ix.io/12qc shows why. It tries to open itself by opening /proc/self/exe and doing readlink on /proc/self/fd/, which returns the path of the linker 2018-03-25 12:32:10 It comes with documentation. 2018-03-25 12:32:31 You are not using mainline u-boot though. 2018-03-25 12:32:58 yes i tried it 2018-03-25 12:33:04 i compiled it from source 2018-03-25 12:33:22 but uboot from odroid uses a boot.ini 2018-03-25 12:34:33 I built u-boot from source for bananapi (A20 SOC) and it works 2018-03-25 12:34:47 zhasha: ... what? 2018-03-25 12:35:02 ok, i will try the chroot methode in the wiki 2018-03-25 12:38:08 one last question 2018-03-25 12:38:10 zhasha: gcompat is rather wacky to begin with, you are probably better with doing an glibc chroot 2018-03-25 12:38:23 how to enable cache 2018-03-25 12:38:28 for apk 2018-03-25 12:38:30 ? 2018-03-25 12:38:37 liwakura: it's a self-mounting binary with an embedded squashfs image that it mounts using fuse. I assume it needs to pass a path to fuse or something like that, so it tries to determine its own absolute path by opening /proc/self/exe and doing a readlink on the fd file in /proc/self/fd. I'm guessing this returns the absolute path of the binary on a glibc system but on a musl system it returns the path of the linker 2018-03-25 12:38:38 gerald: setup-apkcache 2018-03-25 12:38:54 because now apk is working but only with --no-cache 2018-03-25 12:39:00 zhasha: usually not. 2018-03-25 12:39:22 maybe its an side effect of gcompat 2018-03-25 12:39:23 My guess is that the gcompat pre-linker invokes the real musl linker 2018-03-25 12:39:55 (explicitly) 2018-03-25 12:40:42 likely.. 2018-03-25 12:45:59 for what is the world file in /etc/apk 2018-03-25 12:46:01 ? 2018-03-25 12:46:16 because for some packages apk tells me: 2018-03-25 12:46:38 ERROR: unsatisfiable constraints: setup-apkcache (missing): required by: world[setup-apkcache] 2018-03-25 12:47:26 have you done apk update ? 2018-03-25 12:47:53 oh 2018-03-25 12:48:02 no, its because the package is named "alpine-conf" 2018-03-25 12:48:20 its the command that is named setup-apkcache 2018-03-25 12:49:11 but for what is the world file in /etc/apk ? 2018-03-25 12:49:30 keeping track of the manually installed packages 2018-03-25 12:49:52 no, please dont edit it until you understand what its doing 2018-03-25 14:28:17 Is iptable-persistent required in Alpine in order to save iptable rules? 2018-03-25 14:30:43 technically, no 2018-03-25 14:30:59 you can make iptables persistent using any comparable mechanism 2018-03-25 14:31:45 but i guess you want iptables-persistent using the openrc startup script 2018-03-25 14:31:54 so for you, yes 2018-03-25 14:33:41 dont forget rc-update add iptables then 2018-03-25 14:50:49 I usually write my own firewall scripts and trigger them on if-up 2018-03-25 14:53:48 ^ 2018-03-25 15:25:46 liwakura, I've already used the package in debian so it'd be easier to port if the iptable-persistent package is already available in alpine 2018-03-25 15:26:07 alpine has an own mechanism for that 2018-03-25 15:26:14 they probably work alike 2018-03-25 15:27:00 iptables-persistent creates two files for ipv4 and ipv6 in /etc/iptables-persistent and loads them 2018-03-25 16:17:34 For some reason /etc/init.d/iptables does not exist 2018-03-25 16:17:57 iptables is installed so I'm not sure what's wrong 2018-03-25 16:23:44 <_ikke_> the_drow: apk info -L iptables 2018-03-25 16:39:30 _ikke_, checking 2018-03-25 22:20:49 I'm liking how easy it is to create APK files as compared to e.g. debian packages, Portage ports. 2018-03-25 22:24:46 If we want to create pkgs for our own use what is the best place to put it in? A separate dir in aports under a different branch, a separate ports tree entirely? 2018-03-25 23:02:58 TomTheDragon: i just put them on my own branch of aports.git, but that's my personal preference, i don't know what others prefer or recommend 2018-03-25 23:13:57 TomTheDragon: generally if oyu don't intend on upstreaming those packages, a separate ports tree is better 2018-03-25 23:14:00 that's the approach I take, at least 2018-03-25 23:14:16 after all, a ports tree is just a git repo with top-level folders-as-categories 2018-03-25 23:18:32 Separate branch with internal/ folder is what I have 2018-03-26 01:29:26 Shiz, I think I'll stick with a separate ports tree since most of the stuff I'm interested will probably never be upstreamed. 2018-03-26 01:31:22 aterm, xmms1, trinity desktop (a big maybe)... gkrellm might be a candidate for upstream, not sure 2018-03-26 02:17:06 Is there some kind of "user repository" like Arch's AUR? 2018-03-26 02:17:58 Not "officially", and AUR is a trash fire and a half anyway 2018-03-26 02:18:31 To be honest, my experience with AUR is it often is a place to dump junk packages which don't do any compiling at all. 2018-03-26 02:18:50 It is full of unmaintained, broken, or downright malicious packages 2018-03-26 02:19:37 malicious as in proprietary stuff, or even more so? 2018-03-26 02:21:49 There are packages that could be interpreted as such without anybody checking, e.g. packages with patches that have little or no provenance, the packages for which exist to satisfy an extreme niche or are aimed at a very specific group 2018-03-26 02:22:22 e.g. https://aur.archlinux.org/packages/tor-browser-en-iran/ 2018-03-26 02:22:35 ^ this does not inspire confidence in the AUR 2018-03-26 02:22:51 ^ this does not inspire confidence in the AUR 2018-03-26 02:22:54 whoops 2018-03-26 02:22:59 specifically for users in iran.. wow 2018-03-26 02:23:59 when I last used it years ago, it didn't seem that bad 2018-03-26 02:28:17 https://medium.com/zendesk-engineering/making-docker-and-consul-get-along-5fceda1d52b9#22a1 2018-03-26 02:28:26 how would I do that in alpine? 2018-03-26 03:42:37 any way to normalize and make existingpatches human readable? 2018-03-26 03:42:55 not the same as what's in the wiki 2018-03-26 08:28:15 duncan^: wow, but did you check the source? what did they change for iran? 2018-03-26 13:21:17 hiro: Nobody checks the source 2018-03-26 13:21:42 I think they added languages and stuff, and then left it to rot, so putting the people in Iran in danger, if they choose to download it 2018-03-26 14:04:37 uhg 2018-03-26 14:05:05 having separate packages/binaries for a locale rather than just a pure-data message translation add-on is such an awful idea 2018-03-26 14:05:10 hi dalias, wazzap? 2018-03-26 14:05:13 _especially_ for something like tor 2018-03-26 14:05:15 hi ncopa 2018-03-26 14:05:21 did you see my msgs about gnu make? 2018-03-26 14:05:32 if it's not broken yet it will be when alpine recompiles it 2018-03-26 14:05:45 i saw the gnu make bug yes 2018-03-26 14:05:51 and i think i saw someone fixing it in alpine too 2018-03-26 14:06:18 i wonder if gnu glob implementation is really needed 2018-03-26 14:06:19 would be nice to use musl glob 2018-03-26 14:07:16 iirc make was actually the main user of gnu glob bs -- passing it custom dir funcs so it could process archives (static libs) as dirs 2018-03-26 14:07:23 but i may be mistaken and they may do that another way 2018-03-26 14:08:23 sounds ugly 2018-03-26 14:08:39 not feeling for opening that can in case its filled with worms :) 2018-03-26 14:09:57 grepping the source would show if it's used or not 2018-03-26 15:18:16 this is not exactly alpine specifc, but has anyone been playing with truecrypt more than the average user? 2018-03-26 15:19:48 I'm asking as there's an interesting little performance related feature that makes me wonder 2018-03-26 15:30:41 Isn't truecrypt abandonware and unmaintained though? 2018-03-26 15:32:49 hello 2018-03-26 15:33:06 could someone, please, help with building php with php-fpm 2018-03-26 15:33:13 ? 2018-03-26 15:33:24 it refuses connections for some reason 2018-03-26 15:33:35 yes it is duncan^, but there are still reasons for some people to use it 2018-03-26 15:33:48 but refuses them only on out IPs 2018-03-26 15:35:33 <_ikke_> Why are you building PHP yourself? 2018-03-26 15:38:10 duncan^: interestingly enough, this might also apply to VeraCrypt unless this specific thing has been altered in it 2018-03-26 15:39:19 TBB: Yeah, is there a reason you wish not to use VeraCrypt? 2018-03-26 15:39:51 audit status, mainly 2018-03-26 18:38:46 I've been struggling with doing full-disk encryption on a machine that can only boot using uefi -- right now I have gpt part 1 (/boot/efi, 2MB), part 2 (/boot and grub files, 96MB) and part 3 (/, 50GB, luks) 2018-03-26 18:39:11 my choice of grub was based on setup-alpine selecting that during the autoinstall, would extlinux be a better choice for my purpose? 2018-03-26 18:39:55 the issue I'm currently facing is grub freezing after selecting linux-hardened in the boot list, displaying the text "Booting Alpine Linux" 2018-03-26 19:20:30 trfl, maybe edit the commandline and remove quiet/add noquiet 2018-03-26 19:26:00 good suggestion, but no change actually - still just a black screen :/ 2018-03-26 19:29:28 grub, the emacs of bootloaders 2018-03-26 19:31:13 personally i hate grub and would use extlinux but i don't know if there are reasons to prefer grub for your needs 2018-03-26 19:32:11 I gave extlinux a brief try at first, but couldn't get it to do anything sensible (even though it technically booted), but I also didn't put much effort into it since I assumed there was a reason for grub being the default 2018-03-26 19:33:11 I'll give it another go, if not just to see how it failed :p 2018-03-26 20:33:12 no dice, can't make it see the extlinux boot image, just goes into the uefi gui... and grub deadlocks as soon as I try to load any initramfs 2018-03-26 20:50:36 is alpine available without binary blobs? 2018-03-26 20:51:01 and, does apk give the option of using only FLOSS repos? 2018-03-26 20:51:29 (like the ACCEPT_LICENSE option in the portage .conf file) 2018-03-26 20:51:43 <_ikke_> gnulligan: afaik, there is nothing like that 2018-03-26 20:52:00 oh... 2018-03-26 20:53:22 <_ikke_> And all repos are binary only 2018-03-26 20:53:56 <_ikke_> Depending on your usecase, you can after installing a base system use aports to build a custom repo 2018-03-26 20:54:21 <_ikke_> / installing packages from source 2018-03-26 20:54:49 <_ikke_> abuild -i would build and install a package from source 2018-03-26 20:55:04 does it have selinux? 2018-03-26 20:56:51 <_ikke_> there is libselinux in testing 2018-03-26 21:53:30 looks like my stuff might have been working after all - just noticed that even with the live usb, I get "stuck" at grub loading the kernel until it gets to the "Loading hardware drivers..." part 2018-03-26 21:53:41 so the password prompt simply wouldn't be visible regardless 2018-03-26 21:53:57 this sounds like a hassle to debug... :p 2018-03-26 23:01:35 in an attempt to get the display working, I appended a set of features (drm, fbcon, i915, asus_wmi) to /mnt/etc/mkinitfs/mkinitfs.conf -- however the created initramfs does not seem to actually get the additional modules included 2018-03-26 23:01:58 am I doing something wrong here? mkinitfs -c /mnt/etc/mkinitfs/mkinitfs.conf -b /mnt 2018-03-26 23:24:14 trfl: where did you find these features, i.e. fbcon, i915 etc 2018-03-26 23:59:22 can I somehow prevent the initramfs from being built when installing linux-vanilla? 2018-03-27 00:22:02 anyone? 2018-03-27 00:31:14 Would there be any demand for a patch to add NFS boot support? 2018-03-27 00:31:28 I'm honestly kind of surprised it doesn't already exist in the default init script 2018-03-27 00:32:09 nfs is totally insecure beyond all hope 2018-03-27 00:32:36 dalias: you can use kerberos with nfs, but it's broken in alpine 2018-03-27 00:33:10 dalias: also it's hard to set up. 2018-03-27 00:40:35 kerberos is kind of a nightmare... I was honestly hoping some day an alternative would come about 2018-03-27 00:41:34 i don't think using kerberos actually helps, does it? 2018-03-27 00:42:34 afaik the transport is always insecure (clear) 2018-03-27 00:43:35 dalias: as fas as I know you can choose between kerberos authentication and authentication+encryption 2018-03-27 00:44:31 according to the nfs manpage: "The krb5p security flavor encrypts every RPC request to prevent data exposure during network transit" 2018-03-27 00:49:32 but last time I tried using kerberos with nfs on alpine rpc.gsssvcd segfaulted everytime I connected. To be fair it wasn't even enabled in the APKBUILD, even though the gentoo init script is shipped 2018-03-27 00:50:01 *rpc.svcgssd 2018-03-27 00:53:48 I'm having some trouble getting it to load the ethernet interface on boot... is there a way to tell mkinitfs to load the module, which is in the initfs? 2018-03-27 00:57:13 TomTheDragon: as far as I can tell it loads modules from a list in features.d/network.modules, which should happen if network is defined in mkinitfs.conf. but i have no experience with mkinitfs, as i usually use my own initramfs. 2018-03-27 00:58:27 ah, I see 2018-03-27 00:58:40 it does appear that it's ignoring the module, will try again 2018-03-27 01:00:23 it looks like default init script checks the modules= kernel command line but doesn't appear to have changed a thing 2018-03-27 01:02:33 hah... I know where I went wrong now. Testing in VirtualBox because I got sick of rebooting, but of course the driver is not my native one 2018-03-27 01:14:21 didn't seem to help either. I might just give up, I can probably roll a workable initscript 2018-03-27 01:15:33 looks like I do need the modules= line 2018-03-27 01:23:15 TomTheDragon: https://github.com/alpinelinux/mkinitfs/blob/master/initramfs-init.in#L351 looks like it's loading everything from cmdline modules="" and /etc/modules 2018-03-27 01:28:37 got that working, thanks.. now I just have to figure out how to get the nfs root mounted and let it behave like a normal disk mount 2018-03-27 01:34:24 looks like it was failing because I don't quite understand the shell substrings yet 2018-03-27 01:44:45 yay, finally got boot (though it's 13MB with nothing but getty's... maybe the tmpfs uses up some) 2018-03-27 01:47:08 is that normal? 2018-03-27 01:58:30 TomTheDragon: 13mb isn't much, I think it's just the modules 2018-03-27 02:00:16 could just be the modules... I saw around 8 when I tried a standard disk install in VM, but I didn't add any modules or anything 2018-03-27 02:04:14 TomTheDragon: well you can take a look inside 2018-03-27 02:04:27 gunzip < initramfs | cpio -i 2018-03-27 02:04:36 should extract it to teh current folder 2018-03-27 02:05:43 what exactly am I looking for? 2018-03-27 02:11:00 TomTheDragon: You could do `du -hd1` to find the largest directory enter it and do it again to see whats so big 2018-03-27 02:12:18 oh... so initrd does get persisted even after pivoting the root 2018-03-27 02:15:33 I'm not terribly worried, just one system I plan to run it's 1/12 the memory 2018-03-27 02:17:38 > initrd does get persisted even after pivoting the root 2018-03-27 02:17:48 I actually never thought about that 2018-03-27 02:18:18 there's no way... I should be using 67MB 2018-03-27 02:18:21 but I would expect it to get removed from memory after switching root 2018-03-27 02:19:30 unless it only unpacks what it needs, or something rather strange that I don't see the init script doing 2018-03-27 02:20:03 it must be getting removed... perhaps nfs is just creating some bigger slab objects, and slabtop doesn't work with a vanilla kernel 2018-03-27 02:25:53 that's not to say I wouldn't like a smaller initrd... it's a bit big for tftp. 2018-03-27 02:29:19 I mean... backlight module in initial booting environment? 2018-03-27 02:34:56 TomTheDragon: the kernel unpacks everything thats in the initramfs into memory, the decision what's needed must be done by mkinitfs. I guess you have to make sure it includes nfs-utils for nfs mounting to work. 2018-03-27 02:36:00 regarding nfs, I didn't have to include anything special 2018-03-27 02:36:08 then again I might be using some kernel fallbacks 2018-03-27 07:10:23 how reliable is the /dev/usbdisk symlink? can I expect it to work if I have a FAT formatted stick inserted at all times, even if it is replaced at some point? 2018-03-27 07:10:48 this is on a RPi3 running 3.7 2018-03-27 07:51:37 mps: lsmod after booting the live usb, just trying the modules i suspect to be the "hardware driver" which powers on my display 2018-03-27 07:58:50 trfl: 'features' in mkinitfs.conf are files under /etc/mkinitfs/features.d/ which contains list of modules (or something else) to be added to initram 2018-03-27 07:59:15 not the kernel modules 2018-03-27 08:00:52 for example, if you add ext4 feature to mkinitfs.conf it will add modules listed in /etc/mkinitfs/features.d/ext4.modules 2018-03-27 08:04:35 ohh alright gotcha 2018-03-27 08:07:59 so if I rely on i915 for example, i'd have to add it to /etc/modules instead? going on a hunch from something that was said earlier 2018-03-27 08:08:10 or would the initfs module kms pull that in? 2018-03-27 08:09:52 if you want modules loaded in initram you should add them in mkinitfs features, add to existing one or create your own 2018-03-27 08:10:31 if it is enough to be loaded after initram you could add them in /etc/modules 2018-03-27 08:11:49 some kind of the 'rule of thumb' is to add modules needed to boot in initram and other in the /etc/modules 2018-03-27 08:12:08 alright, so definitely initram then 2018-03-27 08:12:42 i'll try to make an i915 module, thanks o/ 2018-03-27 08:13:32 look at examples in /etc/mkinitfs/features.d files 2018-03-27 09:33:33 Is there a php7.2 repo? 2018-03-27 09:37:49 And percona MySQL? :D 2018-03-27 09:41:22 you mean xtradb? 2018-03-27 09:59:02 scv: Jepp 2018-03-27 09:59:16 it's the default engine in mariadb 2018-03-27 09:59:17 iirc 2018-03-27 09:59:39 Hmm, i like more the percona fork, but mariadb should also do the work =) 2018-03-27 10:09:09 hello 2018-03-27 10:11:15 i'm trying to config an app on OpenShift with using Docker alpine containers and I ran into some permissions issues 2018-03-27 10:11:26 anyone around who could give me some hints? 2018-03-27 10:14:55 DanP: "some permissions issue" is too broad 2018-03-27 10:15:18 unlikely that someone figures out what problem you actually have 2018-03-27 10:15:31 well..just starting conversation..:) 2018-03-27 10:15:43 Just start right away 2018-03-27 10:16:01 Most people are rather busy and only read the chat 2018-03-27 10:16:17 If you start explaining a problem, and someone recognizes it, they'll become active 2018-03-27 10:16:42 basically I install packages and tools needed inside the Docker image under root; this includes Wildfly app server 2018-03-27 10:18:14 I add a jboss user with "adduser -D -h $HOME -u 1001 -G root jboss" (I specify UID 1001 because I use it to switch to it inside the layered Dockerfiles) 2018-03-27 10:19:21 I change ownership to the wildfly folder so that user jboss can start the server "chown -R 1001:0 $WILDFLY_HOME" 2018-03-27 10:19:51 when I start the server I get several exceptions from server regarding permissions 2018-03-27 10:20:06 the confusing part is when I ssh into the container 2018-03-27 10:20:37 the whoami shows: "whoami: unknown uid 1000100000" 2018-03-27 10:21:04 how does the passwd line of jboss look like? `grep jboss /etc/passwd` 2018-03-27 10:21:07 th id shows: "id=1000100000 gid=0(root) groups=1000100000" 2018-03-27 10:21:29 something failed there 2018-03-27 10:21:33 it looks good...:)....jboss:x:1001:0:Linux User,,,:/opt/java-apps:/sbin/halt 2018-03-27 10:22:08 aha 2018-03-27 10:22:15 by the way, probably a bad idea to use root 2018-03-27 10:22:16 and /etc/group is: root:x:0:root,jboss 2018-03-27 10:22:57 DanP: i think docker run is given an wrong -u option 2018-03-27 10:23:33 I have no idea where that awkward ids stem from 2018-03-27 10:24:09 i see liwakura ...so you suggest this isn't alpine-related 2018-03-27 10:24:23 if the server starts without existing immediately, you could do an `ps faux` on the host an look what uuid it seems to run as from the host 2018-03-27 10:24:27 *exiting 2018-03-27 10:25:16 it does exit immediatly, but I can run the container in debug mode and that's where I ran the whoami and id 2018-03-27 10:25:20 and observed this 2018-03-27 10:25:46 DanP: how are the container deployed? the issue is likely there 2018-03-27 10:25:54 debug mode= switch the actual CMD with "sleep 3600" 2018-03-27 10:28:18 they are deployed/ran through OpenShift....and I have no idea how it passes argumnts to (what probably is) the docker run command 2018-03-27 10:28:20 hmmm... 2018-03-27 10:31:21 liwakura: ty for your time!...I guess I'll have to look more into the OShift/Docker aspects of this 2018-03-27 10:32:20 what made me come here (after several tests) is that the same config works fine under CentOS....so I thought this could be OS-specific 2018-03-27 10:41:31 03:21 < DanP> th id shows: "id=1000100000 gid=0(root) groups=1000100000" 2018-03-27 10:41:34 smells like docker fuckery 2018-03-27 10:51:24 yes scv...I'm switching to this premise as well....I'm using the USER Docker command with ids not names and I somehow think that's the issue 2018-03-27 11:28:54 DanP: you might want to take a look at /etc/subuid and /etc/subgid 2018-03-27 11:30:24 Hello. 2018-03-27 11:30:39 Is there an easy way to build custum armhf kernel? 2018-03-27 11:46:08 do you have an suitable kconfig? 2018-03-27 11:49:51 I will use default config and need only more driver. 2018-03-27 11:51:06 then you will get away with copying the existing APKBUILD and modifying the config with `abuild menuconfig` 2018-03-27 11:52:15 Doing like: https://wiki.alpinelinux.org/wiki/Custom_Kernel 2018-03-27 11:52:41 oh, didn't know there is an wiki article 2018-03-27 11:52:54 thats probably as easiest as possible when building an custom kernel 2018-03-27 11:53:01 But this is to much and compilcat. 2018-03-27 11:55:00 I think I will build a real custom kernel. 2018-03-27 11:55:45 abuild helps with the dependencies, and the APKBUILD does handle some things for you 2018-03-27 11:55:53 but if you feel like, you can also do it completely manually 2018-03-27 11:58:48 I tried yióur way: 2018-03-27 11:58:58 >>> ERROR: : Could not find ./APKBUILD (PWD=/home/skuhne) 2018-03-27 11:59:18 did you get the current APKBUILD ? 2018-03-27 11:59:30 How? 2018-03-27 11:59:34 git? 2018-03-27 11:59:56 https://github.com/alpinelinux/aports/tree/master/main/linux-vanilla 2018-03-27 12:00:10 default configs are there too 2018-03-27 12:00:17 change vanilla for whichever variant you use 2018-03-27 12:04:23 I can't clone it. 2018-03-27 12:44:45 Hi folks. This is my first IRC "experience". Is it like global chat among #alpine-linux subscribers? 2018-03-27 12:45:57 timfayz: not quite. there is no clear concept of subscription in IRC 2018-03-27 12:46:06 you just join a channel and idle 2018-03-27 12:46:59 people who expect quick answers in an idle channel and then leave too early to see an answer are frowned upon 2018-03-27 12:47:42 meta-questions like "may i ask a question", "i have a problem" or "hi" are often ignored 2018-03-27 12:47:53 timfayz: that said, have fun on irc :) 2018-03-27 12:48:11 in general: ask a consise and accurate question, and wait for an answer 2018-03-27 12:48:51 I hope I got it - thank you all for quick respose :) 2018-03-27 13:03:41 or just discuss things, of course :) 2018-03-27 13:03:44 it's not just support here 2018-03-27 13:35:58 I've never used "mailing lists" and found several ones on https://alpinelinux.org/community/. Each item has (archive | subscribe | unsubscribe | help). First three is quite clear but what the difference, let's say, sending messages to alpine-user and alpine-user+help@lists.alpinelinux.org? 2018-03-27 13:46:10 there were redmine package for AL, what was the reason for removal 2018-03-27 14:06:08 Regarding my previous question - figured out, thanks :) 2018-03-27 16:50:36 i borrowed an C2, looking forward to write an guide how to install alpine on it 2018-03-27 16:52:44 liwakura: I'd be more than happy to proofread it for you 2018-03-27 16:53:53 hah, i know you have an C2 yourself :P 2018-03-27 16:54:35 yeah, and you surely know that I was too lazy to figure it out myself :/ 2018-03-27 16:55:12 i hope you have an valid excuse for that 2018-03-27 17:00:01 is there even a valid excuse for being lazy? 2018-03-27 17:01:39 anemia, diabetes 2018-03-27 17:02:15 luckily i have none of that 2018-03-27 17:02:20 ACTION knocks on the table 2018-03-27 17:27:52 liwakura: I also plan to install AL on bananapi from the begining, and document process, just need to find empty sdcard and litle time 2018-03-27 17:28:33 alhough I have AL running on that board 2018-03-27 17:29:25 I plan to bouild u-boot from mainline 2018-03-27 17:29:36 s/bouild/build/ 2018-03-27 17:37:16 thanks a bunch mps, adding the kms module was all it took :) 2018-03-27 17:38:15 trfl: you are welcome :) 2018-03-27 18:21:12 I'm starting to convert our jenkins agents into alpine linux based docker containers. One of our jobs has a couple of docker commands and I am wonder how to go about getting those commands to work. Googled a bit and found a few workarounds/hacks but nothing that seems to work. As of now I just get "Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running" when trying to run a 2018-03-27 18:21:12 docker command. 2018-03-27 18:21:41 is that socket bind-mounted into the container? 2018-03-27 18:21:52 that question is not really specific to alpine linux 2018-03-27 18:22:25 liwakura: Hmm I came across something about that but wasn't sure what to do. I assume in my dockerfile I need to put something like VOLUME /var/run/docker.sock ? 2018-03-27 18:22:49 liwakura: yeah, I was hoping others using alpine (since it seems to be popular for docker) have run into it 2018-03-27 18:23:49 what about #docker ? 2018-03-27 18:23:56 like, you need to make an bind mount 2018-03-27 18:24:09 its -v with docker run, but since you are using jenkins, its probably differently 2018-03-27 18:24:23 jenkins has some plugins on its own.. 2018-03-27 18:24:50 I'll check around that channel and see what's up. Maybe just need to narrow my search a bit. 2018-03-27 18:24:56 so vague I'm all over the place :) 2018-03-27 18:56:06 I'm using Alpine for a Run-From-Ram system. I would like to mount an extra drive on boot (using fstab), but - I need the directory created first... What is the recommended way to do that on a run-from-ram system ? 2018-03-27 18:57:19 <_ikke_> GlemSom: Not entirely sure, but you should be able to save the directory using lbu 2018-03-27 18:57:28 <_ikke_> Then it would already be present 2018-03-27 18:57:39 _ikke_: lbu only saved /etc by default 2018-03-27 18:57:51 you can add other directories to it 2018-03-27 18:58:10 Is that the recommended way of doing it ? 2018-03-27 18:58:11 <_ikke_> lbu add /path/to/directory 2018-03-27 18:58:46 _ikke_: wouldn't that also add its contents? 2018-03-27 18:58:53 And, when I then mount something to the directory, and run "lbu ci" again - it will try and commit the content of the folder... 2018-03-27 18:59:51 I guess I can create a custom script in rc.local... But, that seems like an odd method of solving a mounting issue :/' 2018-03-27 19:00:56 GlemSom: you can also add an startup script to /etc/inittab with ::sysinit: 2018-03-27 19:47:24 How do I make "lbu ci" remember executable bits on a file ? 2018-03-27 19:47:53 <_ikke_> doesn't it do that? 2018-03-27 19:48:31 _ikke_: No 2018-03-27 19:48:49 Just tested with a script in /etc/local.d 2018-03-27 19:49:51 The local init-script says the files need to end on .start, and need to be executable... I did that, and the script works fine... but - when saving the changes, and rebooting, I noticed the executable bit was not remembered by "lbu ci" :/ 2018-03-27 19:50:26 Which is kinda a show-stopper for using the builtin "local" service for starting stuff at boot :/ 2018-03-27 19:54:47 Ahh, I found the issue.. The lbu script uses "cmp" for comparing files - which does not look file permissions 2018-03-27 19:55:28 So any changes to file-permissions are not found by lbu (And I commited the script first, and then in second stage changed executable bit...) 2018-03-27 20:55:11 Anyway, created issue #8743 with proposed fix to lbu 2018-03-27 20:55:29 In case others run into the same issue as me :) 2018-03-27 21:04:18 How do you get the alpine miniroot to boot in systemd-nspawn without complaining about missing ttys? 2018-03-27 21:05:22 probably removing the entries for them from inittab 2018-03-27 21:05:47 I meant, how to properly create them. 2018-03-27 21:06:01 Do you know what the problem is? 2018-03-27 21:06:02 well, they don't exist inside a container 2018-03-27 21:06:05 that's the problem 2018-03-27 21:06:10 kc_: you should elaborate more on your problem 2018-03-27 21:06:28 it's a typical container issue liwakura 2018-03-27 21:06:32 regardless of the tech used 2018-03-27 21:06:34 I am actually trying to create an initramfs using the miniroot 2018-03-27 21:06:43 /dev/tty* isnt emulated usually 2018-03-27 21:07:01 scv: why should an minirootfs execute inittab? 2018-03-27 21:07:10 with no openrc installed? 2018-03-27 21:07:16 I installed openrc 2018-03-27 21:07:40 being systemd-nspawn i'd expect it to launch init/pretend to be a real system 2018-03-27 21:07:53 The point is that a miniroot download is ~2MB, but an iso download is much larger. 2018-03-27 21:08:02 (no experience with nspawn in particular but executing init is typical) 2018-03-27 21:08:09 Systemd just runs the /sbin/init (unless you tell it otherwise) 2018-03-27 21:08:16 kc_: what scv said, you should remove the getty processes 2018-03-27 21:08:43 there is no reasonable expectation to have /dev/tty{1,...} in a container 2018-03-27 21:08:47 I only ask about systemd-nspawn because it is the same error as in the initramfs, and I remember seeing something about this in the past, but can't find it. 2018-03-27 21:09:12 well, openvz does have them :p 2018-03-27 21:09:14 but that's openvz 2018-03-27 21:09:22 i dont think any other container tech provides them 2018-03-27 21:09:30 kc_: "no job control" can have alot of causes 2018-03-27 21:09:31 Lot's of containers have them. 2018-03-27 21:09:36 missing terminal is one of it 2018-03-27 21:09:55 if it's init->sh you don't get job control FWIW 2018-03-27 21:10:01 kc_: which types? 2018-03-27 21:10:08 duncan^: because its not the controlling terminal, i know 2018-03-27 21:10:09 (some people make sh an init script, that's not how it works) 2018-03-27 21:10:26 yeah,. you can have agetty do nologin on a given tty, then you get sh and job control 2018-03-27 21:10:34 busybox getty doesn't support autologin 2018-03-27 21:10:41 sorry s/nologin/autologin/ 2018-03-27 21:10:57 docker doesn't have it period afaik, lxc has emulation but it's not transparent (different major/minor for the devices) 2018-03-27 21:11:04 duncan^: it has, -l 2018-03-27 21:11:04 and it's not stock either 2018-03-27 21:11:33 liwakura: There was still some reason you can't quite have it go straight to the shell, I forget what it was 2018-03-27 21:11:46 getty needs to set the controlling terminal first 2018-03-27 21:11:54 but then can safely exec to /bin/sh 2018-03-27 21:12:12 So maybe I shouldn't have asked about systemd-nspawn. 2018-03-27 21:12:52 kc_: just remove the getty processes if they point to devices that wont be available in a container 2018-03-27 21:13:10 missing job control sounds more like missing devpts tho 2018-03-27 21:13:15 What would cause this when using the alpine miniroot for the initramfs filesystem? My purpose is to create an initramfs that can run from ram so I can wipe the actual root hard drive. 2018-03-27 21:13:45 liwakura how to you check for devpts 2018-03-27 21:14:00 i think i missed the part about job control being the issue? 2018-03-27 21:14:01 kc_: shell not set as terminal controlling process in initramfs 2018-03-27 21:14:36 kc_: `exec getty -l /bin/sh` fixes the initramfs issue, probably 2018-03-27 21:15:02 the nspawn issue is probably caused by having /dev/pts not bind mounted into the container 2018-03-27 21:15:08 how would that look in the inittab file? 2018-03-27 21:15:10 therefore the terminal device isn't available there 2018-03-27 21:15:51 kc_: look into the inittab file... shouldn't be too difficult 2018-03-27 21:15:52 liwakura 2018-03-27 21:16:01 nspawn creates ptmx/pts on its own 2018-03-27 21:16:07 regardless of whether you have it or not 2018-03-27 21:16:23 kc_: do you have your nspawn container ready? 2018-03-27 21:16:26 (yes, that's awful, but it's systemd) 2018-03-27 21:16:54 can you do `stat $(tty)` ? 2018-03-27 21:17:05 but kc_ hasn't confirmed the actual error message, i believe it's related to tty* missing but lets confirm what the actual error is ;P 2018-03-27 21:17:34 scv: its two issues actually, no job control in nspawn, and no job control in initramfs 2018-03-27 21:17:37 with different causes 2018-03-27 21:17:52 i don't see where job control was mentioned besides when you brought it up though 2018-03-27 21:18:11 > I only ask about systemd-nspawn because it is the same error as in the initramfs, and I remember seeing something about this in the past,... 2018-03-27 21:18:35 ah 2018-03-27 21:18:41 But my issue is not systemd (I was only using systemd-nspawn to try and troubleshoot the problem). The actual problem is qemu-system-x86_64 -kernel /boot/vmlinuz-linux-hardened -initrd hijackerramfs.img -nographic -append console=ttyS0 2018-03-27 21:19:01 I want to sell my 2017 iMac 5K (Maxed)... what would be a good replacement? Lenovo ThinkPad P70 or Fujitsu Celsius H760? 2018-03-27 21:19:20 Should be workstation class laptop 2018-03-27 21:19:20 kc_: do `stat $(tty)` and post output 2018-03-27 21:19:27 that should give some relevant information.. 2018-03-27 21:19:44 if it's just a matter of getting a shell in the initrd 2018-03-27 21:19:45 i want to use alpine linux on it xD 2018-03-27 21:19:45 The line above launches a qemu vm using the kernel of the running machine and an initramfs which is just the alpine miniroot with the openrc and openssh packages installed 2018-03-27 21:20:11 kc_: how is this related to nspawn? 2018-03-27 21:20:25 precreating the dev nodes + having it execute getty -l /bin/sh or w/e similar solution should be sufficient 2018-03-27 21:20:44 i actually don't recall running into any job control or missing tty errors last time i crammed stuff into an initrd though 2018-03-27 21:20:46 oddly enough 2018-03-27 21:20:50 It's not (except that nspawn gives the same complaint) 2018-03-27 21:21:08 kc_: you are probably better off just using the default alpine installer with loading the modloop and apkovl from the web 2018-03-27 21:21:21 seems like a lot of overkill :< 2018-03-27 21:21:34 I need to be able to create this initramfs on the fly. miniroot is 2MB and the iso is much larger. 2018-03-27 21:22:05 kc_: just for the sake of validating the issue, can you confirm the exact error message you're trying to solve 2018-03-27 21:22:20 kc_: do you realize that you wont have any kernel modules then? 2018-03-27 21:22:25 I don't have a problem getting to a terminal in the initramfs. My problem is that the openrc init scripts (which I wanted to use since they will set-up the network) complain about this and then hang. 2018-03-27 21:22:53 kc_: so you dont get an terminal at all? 2018-03-27 21:22:54 liwakura: there's no reason why you can't modify it to load modules 2018-03-27 21:23:16 scv: im still saying he essentially wants alpine diskless 2018-03-27 21:23:24 yeah 2018-03-27 21:23:28 i've done similar configs 2018-03-27 21:23:45 not like, diskless install, but just an single diskless boot 2018-03-27 21:24:02 OpenRC 0.24.1.a941ee4a0b is starting up Linux 4.15.12-1-hardened (x86_64) 2018-03-27 21:24:03 * Mounting /proc ... [ ok ] 2018-03-27 21:24:03 * Mounting /run ... * /run/openrc: creating directory 2018-03-27 21:24:04 * /run/lock: creating directory 2018-03-27 21:24:06 like the installer already does per default.. (except modloop, but you can fetch that later via http) 2018-03-27 21:24:06 * /run/lock: correcting owner 2018-03-27 21:24:08 Service `hwdrivers' needs non existent service `dev' 2018-03-27 21:24:10 * Caching service dependencies ... [ ok ] 2018-03-27 21:24:12 can't open /dev/tty1: No such file or directory 2018-03-27 21:24:14 can't open /dev/tty2: No such file or directory 2018-03-27 21:24:16 can't open /dev/tty3: No such file or directory 2018-03-27 21:24:33 liwakura: right, in my case i just jammed my application into an initrd 2018-03-27 21:24:44 kc_: if you do this again, you will be kicked from the netword due to excess flood 2018-03-27 21:24:47 please use a pasting service 2018-03-27 21:24:53 sorry 2018-03-27 21:25:19 so anyway 2018-03-27 21:25:21 it looks like /dev is not mounted at all 2018-03-27 21:25:23 the error *is* what i expected 2018-03-27 21:25:33 and not job control :P 2018-03-27 21:26:09 kc_ you need to prepopulate your image with a /dev, or get the `dev` service it mentions into your image 2018-03-27 21:26:11 either way is suitable 2018-03-27 21:26:30 This is with the openrc init scripts. If I boot directly into "ash -l" and then run "mdev -s" I can see the devices in /dev, but I still get the same tty complaints. 2018-03-27 21:27:10 I can't populate dev into the miniroot, because the openrc scripts mount some sort of devfs over whatever I do. 2018-03-27 21:27:46 kc_: mdev creates the devices nodes in whatever is currently in /dev 2018-03-27 21:28:10 I checked around and can't find what the 'dev' service is. I installed Alpine to disk to see if it had some sort of 'dev' service but couldn't find it. 2018-03-27 21:28:12 kc_: you are still trying to recreate what works almost out of the box with alpine 2018-03-27 21:28:27 kc_: /etc/init.d/devfs 2018-03-27 21:28:40 The "mdev -s" command created tty0-tty64, but they still did not work correctly. 2018-03-27 21:28:54 kc_: due to which error? 2018-03-27 21:29:14 I am confused, I expected out of the box would be to install the openrc package, and it would work. 2018-03-27 21:29:32 well, this is well outside of a standard configuration 2018-03-27 21:29:42 no, just use the standard initramfs 2018-03-27 21:30:16 take the kernel, initramfs, give an http location for an apkovl and modloop, and you get an full alpine linux system in RAM 2018-03-27 21:30:21 It is a custom initramfs that gets recreated each time it is needed. 2018-03-27 21:30:22 without touching disk 2018-03-27 21:30:38 kc_: why recreated? from what? 2018-03-27 21:30:42 The point is to eliminate the http location. 2018-03-27 21:31:14 I basically need an initramfs that can phone home via "ssh". Each one needs to be created since it has the ssh keys baked into it. 2018-03-27 21:31:44 Once it boots into ram it phones home via ssh to get an image file to write to the hard drive (to hijack an existing VPS). 2018-03-27 21:31:49 personally i wouldn't even be bothering with openrc in this case 2018-03-27 21:32:03 i'd just make a simple, dumb init replacement to handle what needs to be done 2018-03-27 21:32:32 if the use case is that rigid then there's no need for a full blown system 2018-03-27 21:32:36 kc_: fun fact: you can add any files to any pre-existing cpio archive 2018-03-27 21:32:44 which is what initramfs' are 2018-03-27 21:32:47 I just thought that openrc could initialize the network better than I could manually. I would rather just let openrc set it up than messing with udhcpc, etc. 2018-03-27 21:33:01 it's just gonna do the same thing you'd do manually 2018-03-27 21:33:55 kc_: where do you define the init executable, btw? 2018-03-27 21:34:16 ok, thanks for the advice. I guess I will pound out a script. 2018-03-27 21:35:36 The kernel loads the initramfs and then runs /init, which is a script I can do stuff before calling /sbin/init. That is where openrc takes over, but since it mounts something over /dev anything I do in advance is blown away. Whatever the issue is (which seems to be related to the /dev directory) causes the init scripts to hang. 2018-03-27 21:36:55 liwakura: it defaults to a list of paths 2018-03-27 21:37:04 /linuxrc, /init, /bin/init and so on 2018-03-27 21:37:10 scv: not in the initramfs 2018-03-27 21:37:17 ? 2018-03-27 21:37:22 what do you mean "not in the initramfs" 2018-03-27 21:37:29 rootfs and initramfs have different init behavior 2018-03-27 21:37:37 what 2018-03-27 21:37:40 initramfs for example never falls back to /bin/sh 2018-03-27 21:37:45 like, different code 2018-03-27 21:37:57 there is no rootfs here 2018-03-27 21:38:13 therefore no /linuxrc 2018-03-27 21:38:25 yes, no rootfs and no /linuxrc 2018-03-27 21:39:01 initramfs only does /init, and afaik nothing else 2018-03-27 21:39:52 liwakura: https://elixir.bootlin.com/linux/latest/source/init/do_mounts_initrd.c#L37 2018-03-27 21:40:17 My understanding is that the kernel makes no distinction between an initramfs and anything else. If it finds an /init on the initramfs it executes it and completely hands over control. 2018-03-27 21:40:21 scv: ... thats the initrd code 2018-03-27 21:40:25 so, looks like i'm remembering it wrong 2018-03-27 21:40:26 :( 2018-03-27 21:40:36 initrd != initramfs 2018-03-27 21:40:55 yeah 2018-03-27 21:41:58 I know I could "phone home" with i simpler script, but thought it would be a nice perk to have a fully bootable initramfs for troubleshooting (when needed). 2018-03-27 21:42:48 kc_: why are you trying to optimize the http out when you phone home anyways? 2018-03-27 21:43:01 If I do end up using a simple script, I think busybox already has everything that is required, so Alpine wouldn't be needed at all. 2018-03-27 21:43:07 honk 2018-03-27 21:43:20 kc_: what about ssh server? or do you do netcat shell? 2018-03-27 21:44:09 we don't transmit data or configuration over http; only ssh (and I don't want to tunnel http in ssh, etc.) 2018-03-27 21:44:24 what do you mean ssh server? 2018-03-27 21:44:39 how do you get the ssh server into the initramfs if you only have busybox? 2018-03-27 21:45:21 if i'm following, it's ssh outbound to retrieve an image? 2018-03-27 21:47:47 the initramfs is only the client part. If CoreOS is installed on a VPS that you have root access to, my utility will create an image file with a EFI partition containing only the kernel and initramfs. You can ssh to the VPS and use dd to copy the image to the boot device, and after you reboot, the initramfs uses ssh to connect back to the original host that created the image which has an ssh server running that will "cat" back a full 800+MB image 2018-03-27 21:47:47 that can be written to the hard drive (since it is currently running from an initramfs) 2018-03-27 21:48:52 kc_: netboot is no option? 2018-03-27 21:49:30 This should only happen once, won't netboot try and boot this way every time? 2018-03-27 21:49:43 This is only because some VPS provides won't let you upload custom images. 2018-03-27 21:50:30 Do you think this is something worth exploring with netboot? 2018-03-27 21:51:36 if its on VPS where you dont have control over netboot, then no 2018-03-27 21:54:47 thanks for the help 2018-03-27 23:41:38 what does flagged mean on https://pkgs.alpinelinux.org ? 2018-03-27 23:42:04 flagged as out-of-date 2018-03-27 23:42:07 It means somebody flagged that package (perhaps erroneously) as being out of date 2018-03-28 03:17:59 what are "textrels" and how can they be removed? 2018-03-28 03:18:46 a textrel is what happens when you link code that's not position-independent (has absolute address references in it) into a shared library or PIE executable 2018-03-28 03:19:30 they can only happen if (a) some of the object files linked in were compiled without -fPIC/-fPIE, or (b) some of the object files were built from asm, and the asm was not written to be position-independent 2018-03-28 03:19:58 how can it be tracked down to given symbols? chances are it's just missing -fPIC/-fPIE 2018-03-28 03:21:08 alpine's gcc defaults to pie 2018-03-28 03:21:31 so to get a non-pie object file the build would have to be doing -fno-pic/-fno-pie 2018-03-28 03:22:20 eww 2018-03-28 03:22:31 that means it's not a simple fix 2018-03-28 03:23:06 will definitely need some way of tracking the symbols. 2018-03-28 03:24:55 or, despite how much I don't like doing it... skipping the textrel check 2018-03-28 03:33:20 dalias: I used scanelf, says something about has textrel marking but doesn't appear to have any real textrels 2018-03-28 03:33:41 dalias: do I need some postprocessing to strip the marking? 2018-03-28 03:56:05 http://termbin.com/vdhk 2018-03-28 04:02:06 I saw a reference to I386_ASSEM... looks like it's doing some kind of assembly stuff that I was unaware of. should have figured, as a multimedia program. 2018-03-28 13:01:03 Error: Error loading shared library ld-linux-x86-64.so.2: No such file or directory 2018-03-28 13:01:51 (needed by /access-watch/node_modules/appmetrics/appmetrics.node) 2018-03-28 13:01:54 how do i fix this? 2018-03-28 13:03:20 https://pkgs.alpinelinux.org/contents?file=ld-linux-x86-64.so.2&path=&name=&branch=&repo=&arch= 2018-03-28 13:06:32 how does one decide on this? ^ 2018-03-28 13:09:40 realies: you are running an program compiled for glibc under musl 2018-03-28 13:09:55 well apparently /access-watch/node_modules/appmetrics/appmetrics.node needs it 2018-03-28 13:11:20 *shrugs* 2018-03-28 13:11:31 gcompat can help with this, but it does not work reliable for all programs 2018-03-28 13:13:35 sometimes i wonder if people even understand what "musl-based distro" means 2018-03-28 13:13:57 some people don't understand what a libc even is 2018-03-28 13:14:04 so i wouldn't be suprised 2018-03-28 13:17:35 i get that it's a really twisted thing to do 2018-03-28 13:17:46 was confused between which two of those to go with 2018-03-28 13:18:05 it seems it's working all good with libc6-compat 2018-03-28 13:18:15 "dont buy a diesel car if you want to feed it petrol" 2018-03-28 13:18:50 ok, that aside, gcompat vs libc6-compat? 2018-03-28 13:21:18 probably gcompat 2018-03-28 13:22:38 ...why? 2018-03-28 13:23:49 i think libc6-compat is just symlinks to musl, while gcompat is a separate library, which implements more than is supported in musl 2018-03-28 13:28:13 ok so it makes sense to have more than the symlinks 2018-03-28 13:31:10 hi 2018-03-28 13:31:35 is anyone using nftables successfully with the raspi build of alpine? 2018-03-28 13:33:15 i think i'll write an short FAQ on meta questions because this happens way too often 2018-03-28 13:33:32 tostr: you have better chances of getting an answer by explaining your problem 2018-03-28 13:34:30 anyways, im off catching a train 2018-03-28 13:35:30 sure, I just tought if anyone uses it successfully then it's me, not alpine ;) 2018-03-28 13:44:55 basically I can't get nftables to work in any way on the raspi build (3.7) 2018-03-28 13:45:43 I get the error "Error: Could not add table: Invalid argument" 2018-03-28 13:45:50 or "Error: Operation not supported" 2018-03-28 13:45:57 for the most basic of commands 2018-03-28 13:47:21 I have found a similiar bug in raspbian: https://www.raspberrypi.org/forums/viewtopic.php?t=191819 2018-03-28 13:47:37 # modinfo nf_tables 2018-03-28 13:47:38 modinfo: ERROR: Module nf_tables not found. 2018-03-28 13:47:45 same result on alpine 2018-03-28 13:49:45 tostr: did you tried 'modprobe nf_tables' 2018-03-28 13:50:20 which kernel you run 2018-03-28 13:50:46 modprobe: FATAL: Module nf_tables not found in directory /lib/modules/4.9.65-0-rpi2 2018-03-28 13:51:20 is this kernel from Alpine or you build it 2018-03-28 13:51:26 original kernel 2018-03-28 13:52:00 any idea when the main cdn will be back online? i keep getting errors on docker builds 2018-03-28 13:52:31 ls /lib/modules/4.9.65-0-rpi2/kernel/net/netfilter 2018-03-28 13:52:43 are there any nf_* file 2018-03-28 13:53:39 some nf_conntrack* some nf_nat* and nf_log_common.ko 2018-03-28 13:54:18 no nf_tables* in any way 2018-03-28 13:54:35 if not, you have to build kernel with nft modules enabled and install it 2018-03-28 13:55:32 hmm ok 2018-03-28 13:55:48 is this considered a bug? should I report this anywhere? 2018-03-28 13:56:37 hey there 2018-03-28 13:56:40 well, don't know, but I think nf_tables should be in distro kernel 2018-03-28 13:57:38 they are in vanilla kernel on the Alpine for x86_64, at least 2018-03-28 13:57:39 is it possible to extract an apk file ? 2018-03-28 13:58:39 jeanjack: apk is gzipped file, so you can gunzip it 2018-03-28 13:58:52 mps: ok thanks :) 2018-03-28 13:59:24 mps: I'll report a bug, thanks 2018-03-28 14:00:05 jeanjack: after gunzip you can un-tar it because it is tar archive 2018-03-28 14:00:38 or, in single step just use tar 2018-03-28 14:01:13 tar tvf file-name-ver.apk to see what it contains 2018-03-28 15:47:37 is there a way to install linux-vanilla without installing mkinitfs, I really don't want to build an initramfs. 2018-03-28 15:48:10 jnt, install from the vanilla iso image? 2018-03-28 15:50:44 I installed by extracting minirootfs, i then did apk add linux-vanilla, but it depends on mkinitfs and builds one, I have my own initramfs. as far as i can tell, overlays won't work without mkinitfs's initramfs, but i don't use those anyway. 2018-03-28 15:51:10 I essentially want to blacklist that package 2018-03-28 15:54:36 on gentoo it looks like i can do this by adding a package to package.provided, does alpine have something similar? 2018-03-28 15:57:06 mm, you can spoof it with "apk add -t mkinitfs" 2018-03-28 16:00:52 hey so I'm trying to decide between alpine and hardened gentoo 2018-03-28 16:01:17 I was going to set up a musl hardened gentoo install, until I saw that alpine is basically that already ootb 2018-03-28 16:01:37 However, the install needs to be blob-free 2018-03-28 16:02:40 on gentoo you can just put ACCEPT_LICENSE="@FREE" in protage/make.conf 2018-03-28 16:02:48 is there something similar for alpine? 2018-03-28 16:06:42 http://dl-cdn.alpinelinux.org/alpine/v3.7/main: temporary error (try again later) 2018-03-28 16:06:58 it's been 24 hours... how long is this usually out? 2018-03-28 16:07:47 fabled: I just tried it, but it still installs the package, but for some reason it works for packages that don't exist, i can install `apk add -t pizzaistasty`, when i try an existent package it just gets installed though 2018-03-28 16:09:23 cornfeedhobo: works for me 2018-03-28 16:09:40 I understand that the packages are distributed as binaries in apk 2018-03-28 16:09:48 that's not really an issue 2018-03-28 16:09:51 for me 2018-03-28 16:10:07 I'm interested in what license the packages/components are 2018-03-28 16:10:23 I found https://forum.alpinelinux.org/forum/general-discussion/does-alpine-linux-install-non-free-blobs-without-letting-user-kow 2018-03-28 16:10:44 but that guy didn't really know what he was talking about 2018-03-28 16:11:01 oh dear god i'm embarrassed. it was my firewall! my routes for the docker daemon got nuked somehow! 2018-03-28 16:11:08 jnt: thanks! 2018-03-28 16:11:18 the openbsd rebuttal is true... but I happen to have a system that doesn't need blobs 2018-03-28 16:11:42 so openbsd for example would install without blobs on this system 2018-03-28 16:12:44 so I would like to know if there are any non-free kernel components, and if there's a way to specify for apk to not install anything unless its gpl 2018-03-28 16:13:02 or other open-source licenses 2018-03-28 16:13:04 gnulligan: I think you're gonna have a hard time with a completely libre blobfree alpine. 2018-03-28 16:13:32 jnt: so there's nothing like the ACCEPT_LICENSES option in portage? 2018-03-28 16:13:53 gnulligan: linux-vanilla definitely depends on linux-firmware and installs multiple megabytes of blobs 2018-03-28 16:13:53 <_ikke_> no 2018-03-28 16:14:07 oh... you use linux-vanilla... 2018-03-28 16:14:15 is it possible to use linux-libre? 2018-03-28 16:14:39 <_ikke_> You'd probably have to compile / package it yourself 2018-03-28 16:14:39 gnulligan: should be easy to create an APKBUILD for it 2018-03-28 16:15:06 gnulligan: you can only check package licenses and their dependencies and then only install those without proprietary dependencies 2018-03-28 16:15:45 gnulligan: It's not in the repos as far as i know, and blobs aren't in a separate repo. 2018-03-28 16:16:02 but i had no problems running alpine with my own kernel in the past 2018-03-28 16:16:33 jnt: how did you do that? 2018-03-28 16:17:03 I've used mkinitcpio to install arch with a libre kernel 2018-03-28 16:17:17 but that's the extent of my experience on that xD 2018-03-28 16:19:24 gnulligan: I didn't use linux libre, but i Installed the minirootfs, chrooted into it, compiled upstream linux, when unsure about an option I looked at https://git.alpinelinux.org/cgit/aports/plain/main/linux-vanilla/config-vanilla.x86_64, then I built my own initramfs by using another alpine minirootfs and installed extlinux as bootloader. 2018-03-28 16:20:21 oh, and I also needed to install openrc, but I am a longtime gentoo users, so all of this was a pretty straightforward process for me 2018-03-28 16:20:55 I see... it sounds easier to just use hardened musl gentoo and edit /portage/make.conf tbh 2018-03-28 16:21:19 which kinda sucks because alpine really sounds great 2018-03-28 16:21:55 gnulligan: you will certainly get more documentation, when choosing gentoo. Wouldn't say it'll be easier though 2018-03-28 16:24:12 oh well. It would be nice if there was a blob-free version. Some people consider blobs to be a security risk so it might be something to consider in a security-oriented distro 2018-03-28 16:24:20 thanks jnt 2018-03-28 16:25:22 gnulligan: if you never installed gentoo i'd start with glibc though, or at least in a VM, when I did it the first time it took me three days to get to a fully working system. 2018-03-28 16:25:53 jnt: well I 2018-03-28 16:25:59 I've installed arch before 2018-03-28 16:26:01 through cli 2018-03-28 16:26:10 and gentoo looks similar 2018-03-28 16:27:57 gnulligan: they're remotely similar, but gentoo compiles everything from source, so it takes a while, if you install it on desktop. 2018-03-28 16:31:20 gnulligan: about blob-free alpine, I am sure there are other's who would like to see that, personally I'd love to run a blob free, but my hardware needs blobs. 2018-03-28 20:33:13 Hi, i tried to install alpine on a rpi 1 b+, though it boots and i can run setup-alpine, i have problems writing on the sd card, using this tutorial https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi . I get some I/O Errors, sometimes the mountpoint changes from rw to ro, or he can't access the device anymore. 2018-03-28 20:34:11 My first thought was that the sd card is corrupted, so i tested the card with f3 on my desktop machine and the cards seems to be ok. Maybe someone experienced the same problems? 2018-03-28 20:46:27 We build python images based on alpine and find the builds really slow - I think alpine doesn't make use of the binary wheels. Any experienced this? 2018-03-28 20:48:14 As discussed here -https://github.com/docker-library/docs/issues/904 2018-03-28 21:09:47 hey, on pkgs.alpinelinux.org is tmux-2.6 flagged as out-of-date, with 2.7 being the new version, however it's just a -rc 2018-03-28 21:10:40 it's not worth the hassle to update it to 2.7-rc, right? 2018-03-28 21:15:43 probably not if you have no problems with tmux-2.6 2018-03-28 21:16:25 not really, just saw it flagged and thought I could do something useful every now and then 2018-03-28 21:20:15 https://github.com/tmux/tmux/blob/master/CHANGES 2018-03-28 21:55:39 if I install alpine and run the linux-libre scripts, that should make it blob-free, right? 2018-03-28 22:01:00 jnt: 2018-03-28 22:22:45 for some reason I thought linux-libre was only distributed as a kernel 2018-03-28 23:26:36 gnulligan: Well it will [nominally] make the kernel blob-free, but you'll be ignoring the blobs in linux-firmware 2018-03-29 01:32:00 I was going to try out lxde, it appears there are some packages but there doesn't seem to be a convenient metapackage 2018-03-29 01:32:12 is it perhaps an incomplete port? 2018-03-29 01:38:13 TomTheDragon: All you really need is lxsession, lxpanel, openbox, lxappearance 2018-03-29 01:38:22 it should work 2018-03-29 01:39:43 ah, I see. will try that out, see if it's slightly lighter than xfce4. 2018-03-29 01:40:23 I'm having some issues with firefox-esr, doesn't even display anything. I do have netsurf so I have basic web browsing, but some sites just don't like it. 2018-03-29 01:42:57 Hey, I'm back to ask some questions about the beaglebone black 2018-03-29 01:42:57 I've gotten a non-alpine kernel to boot using the alpine initramfs, however I cannot get a shell. It either says `can't open /dev/ttyO0: no such file or directory` or, after changing some kernel CONFIG_ s, `getty: bad speed: 115200n8` 2018-03-29 01:44:10 Well maybe /dev/ttyO0 does not exist 2018-03-29 01:44:15 My end goal is to port postmarketOS, but I'm trying to get alpine to boot first. 2018-03-29 01:44:16 Here's a github issue with my progress: 2018-03-29 01:44:16 https://github.com/postmarketOS/pmbootstrap/issues/1347 2018-03-29 01:45:10 Did you try Alpine's kernel+dtb (it does work) 2018-03-29 01:46:12 duncan^: I'm pretty sure it didn't/doesn't. The "newer" serial platform changes it from `/dev/ttyO0` -> `/dev/ttyS0`, but my config modifactions reverted it to the classic OMAP style of just `/dev/ttyO0` 2018-03-29 01:47:15 duncan^: Yeah, I tried multiple times. It simply hangs after u-boot exits. irregardless of my `console=...` options. 2018-03-29 05:06:20 so my initramfs is corrupt, how might i rebuild this? 2018-03-29 05:11:22 chroot into your system from a livecd/usb - and run mkinitfs 2018-03-29 05:12:18 not familiar with chrooting a good system 2018-03-29 05:12:30 to a bad one 2018-03-29 05:13:42 initramfs is just one tiny file in /boot. unless you have massive corruption odds are high that the rest of the system are useable, and probably even intact 2018-03-29 05:13:54 how did your initramfs get corrupt? 2018-03-29 05:14:11 thats a great question i will be sure to investigate once i finish figuring out how to chroot into this box 2018-03-29 05:15:24 okay. figure out where it is. mount it, then mount /dev, /proc and /sys to the respective directories under where you mounted it 2018-03-29 05:15:35 and once that's done you can just chroot into it 2018-03-29 05:25:47 extlinux: cannot open device /dev/vda1 2018-03-29 05:25:49 :| 2018-03-29 05:27:19 I'm getting some crazy blurry/distorted images in Seamonkey 2018-03-29 05:27:35 then again it's in testing, and it wasn't even compiled for i386 2018-03-29 05:34:52 no text either, though that could be i18n 2018-03-29 09:45:29 Where's the best place to find vulnerability infomation about alpine packages? Redmine? 2018-03-29 12:42:40 xaocon, valuable information? :D 2018-03-29 12:44:03 Well, yes, that would be for the best. :) 2018-03-29 12:44:19 I was hoping that there might be something like this that I just hadn't seen yet: https://security-tracker.debian.org/tracker/status/release/stable 2018-03-29 13:01:46 xaocon, there's http://pkgs.alpinelinux.org/ 2018-03-29 13:19:23 realies: I found alpine-secdb which is closer to what i was looking for. Thanks for the help. 2018-03-29 13:22:49 has anyone encountered any problems with init on edge branch? 2018-03-29 13:24:52 works fine for me 2018-03-29 13:27:02 what was the last time you did a reboot? 2018-03-29 13:33:43 today 2018-03-29 13:34:13 last upgrade was yesterday 2018-03-29 13:44:29 hey there, I have a situation, I need would be happy about some advise: there is an alpine-package (https://pkgs.alpinelinux.org/package/v3.7/main/x86_64/ruby) that needs a security release (from ruby 2.4.3 to 2.4.4). 1) I tried the flag button, but it is not working 2) I wrote the maintainer (https://git.alpinelinux.org/cgit/aports/tree/main/ruby/APKBUILD?h=3.7-stable), but no anwser yet. is there any other possible way I could 2018-03-29 13:46:31 how long since you wrote them? 2018-03-29 13:46:58 also, i think its better to just send it to the mailing list, and Cc the maintainer 2018-03-29 13:47:54 only some hours ago... I wonder if there is a way to contribute or if it is up to the maintainer to fix this :) 2018-03-29 13:48:40 you are aware of timezones? 2018-03-29 13:49:09 in addition to that, some people even have family or even worse, a job 2018-03-29 13:49:28 or sleep ;) 2018-03-29 13:51:07 i am aware of all those constraints and totally fine to wait for an anwser from the maintainer :) but: I ask myself how to contribute or help to get this done even without the maintainer :) 2018-03-29 13:53:31 maybe i should open a bug and attach a patch? how about that? 2018-03-29 13:55:44 robbl: you can create a bug on bugs.alpinelinux.org 2018-03-29 13:55:48 mark target 3.7 2018-03-29 13:57:03 i'wll. you are the maintainer, right? :) 2018-03-29 16:12:16 I'm having trouble booting my Alpine Linux install today. 2018-03-29 16:12:26 It ends up in a kernel panic: 2018-03-29 16:13:09 "Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentation/init.txt for guidance." 2018-03-29 16:13:21 I tried passing init=/bin/sh, but that results in: 2018-03-29 16:13:59 <_ikke_> azarus: what kind of setup is it? 2018-03-29 16:14:11 "Kernel panic - not syncing: Requested init /bin/sh failed (error -2)." 2018-03-29 16:14:16 _ikke_: a disk install 2018-03-29 16:15:27 is the disk OK? 2018-03-29 16:15:37 yup 2018-03-29 16:15:38 maybe /bin/sh really doesn't exist? 2018-03-29 16:15:54 i'll do a zpool scrub just to make sure 2018-03-29 16:16:05 What does smartctl say? 2018-03-29 16:18:32 SMART looks OK to me 2018-03-29 16:18:53 That's good 2018-03-29 16:19:25 I'll now do a scrub. I do doubt that's the issue tough 2018-03-29 16:19:31 though* 2018-03-29 16:21:18 ZFS reports no known data errors. 2018-03-29 16:22:24 /bin/sh exists. 2018-03-29 16:22:49 I tried booting both linux-vanilla and linux-hardened, both show the same behaviour. 2018-03-29 16:22:58 <_ikke_> is the initramfs ok? 2018-03-29 16:23:17 _ikke_: Recreated it just now, should be OK. 2018-03-29 16:31:19 any pointer to url or docs about remastering AL iso image 2018-03-29 16:34:47 its volume levels are pretty good as they are. pretty avant garde tho, mostly bleeps and buzzes. 2018-03-29 16:37:33 I recreated the initramfs and it still fails to boot :/ 2018-03-29 16:56:27 Usually, it asks me for the passphrase for the LUKS partition, but it panics before that 2018-03-29 17:11:20 Any other clues? I'm kinda at a loss :( 2018-03-29 17:12:43 hey 2018-03-29 17:13:28 do standard, extended and miniroot include the grsec patch? 2018-03-29 17:22:33 no? 2018-03-29 17:34:59 Evening 2018-03-29 17:36:30 I’m installed AL on my 2017 5K iMac 2018-03-29 17:36:54 Having some troubles with Xorg not being able to find the screens. 2018-03-29 17:37:16 Installed the amdgpu and the intel video 2018-03-29 17:59:19 and I just wanted to help id4rk... ;_; 2018-03-29 17:59:25 but he's gone! 2018-03-29 18:39:26 Still not getting any further. init=/bin/ash doesn't work either 2018-03-29 18:45:52 Same here. 2018-03-29 18:49:19 anyone can tell me parameters for xorriso to build bootable AL iso image 2018-03-29 19:02:56 nevermind, found it 2018-03-29 19:06:15 I believe the newest mkinitfs broke zfs for me 2018-03-29 19:29:40 yay... tmux in 5k 2018-03-29 19:29:49 <_ikke_> 5k? 2018-03-29 19:30:29 well i dont really know... but at least AL works on the 2017 5k imac, except for xorg 2018-03-29 19:30:56 but yeah... it looks damn small, im breaking my eyes reading 2018-03-29 19:33:26 even wireless works 2018-03-29 19:33:48 out of the box 2018-03-29 19:33:56 that was not the case with debian 2018-03-29 19:36:43 azarus: kernel panic with init not found means that either the initramfs has no +w /init or failed to load 2018-03-29 19:37:04 it cannot happen after the initramfs already ran 2018-03-29 19:53:07 liwakura: okay then 2018-03-29 19:53:34 liwakura: where do you suggest I check next? 2018-03-29 19:56:32 Can anyone guide me trough configuring xorg, even if it was only vesa. It tells me that i do not have any screens, i googled it for about 2 or 3 hours and testing and not able to get it to work and i am not sure what i am doing wrong 2018-03-29 19:57:22 id4rk: how are you running xorg 2018-03-29 19:58:26 first, i used setup-xorg-base, then installed xfce4 2018-03-29 19:58:35 i tried to startx 2018-03-29 19:58:58 id4rk: what's in your .xinitrc 2018-03-29 19:59:22 azarus: dmesg 2018-03-29 19:59:41 liwakura: how can I provide a dmesg if i can't boot= 2018-03-29 19:59:42 ?* 2018-03-29 19:59:58 oh, i mean the kernel log 2018-03-29 20:00:11 do yo have "quiet" in your kernel command line? if yes, remove it 2018-03-29 20:00:30 yeah i didn't see anything special when I removed quiet 2018-03-29 20:00:43 "Unable to open initial console" ? 2018-03-29 20:00:43 I can provide screenshots tough 2018-03-29 20:00:44 azarus: you mean in the home directory? 2018-03-29 20:00:50 id4rk: yes 2018-03-29 20:00:57 liwakura: no 2018-03-29 20:01:09 azarus: i dont have any 2018-03-29 20:01:27 id4rk: well you need one 2018-03-29 20:01:51 azarus: does you initramfs contain secret information? 2018-03-29 20:02:01 otherwise, you could upload the kernel and initramfs 2018-03-29 20:02:01 liwakura: nope, not really 2018-03-29 20:02:08 OK, will try to do that 2018-03-29 20:02:17 and i could put it into qemu and look for myself 2018-03-29 20:02:32 liwakura: be warned, it's a LUKS+ZFS setupü 2018-03-29 20:02:33 setup* 2018-03-29 20:02:38 azarus: ok created it 2018-03-29 20:03:06 i wont be able to boot it, but i'll be able to take more detailed look in what happens before the luks comes into play 2018-03-29 20:03:21 liwakura: OK, a sec 2018-03-29 20:03:30 id4rk: what did you put into it? 2018-03-29 20:04:28 Hopefully a quick question. Where do I configure the docker daemon running on alpine, for a proxy? Tried env vars, conf.d/docker and env vars in the rc script... 2018-03-29 20:05:06 nothing yet lol 2018-03-29 20:05:24 i dont understand, if startx does not work, how can i use it in xinit 2018-03-29 20:05:42 id4rk: you need something like "exec xfce4-session" in there 2018-03-29 20:05:45 then use startx 2018-03-29 20:07:30 ok, sorry for being sceptical, i still cannot see how it will find the screen (does not detect my monitor), but ok lets try 2018-03-29 20:07:51 If I run dockerd myself, it's getting the proxy settings (probably from the env?) and works. 2018-03-29 20:09:56 Trondd-w: the dockerd init file has an config script in /etc/conf.d 2018-03-29 20:10:05 which might or might not exist yet 2018-03-29 20:10:16 i think you can specify the proxy there 2018-03-29 20:10:17 It exists. Doens't seem to be using it 2018-03-29 20:10:41 did you read the manpage on it? 2018-03-29 20:11:27 or some documentation? 2018-03-29 20:11:31 does not work, and i expected it to not work, if i cannot even get the xorg server to work it will not run and desktop environment 2018-03-29 20:11:41 because i think just putting it there might not do it 2018-03-29 20:11:42 and = any 2018-03-29 20:11:56 Which man page? for dockerd? 2018-03-29 20:12:15 id4rk: did you the xorg via setup-xorg-base ? 2018-03-29 20:12:26 on my laptop i also start with startx without having an rcinit file 2018-03-29 20:12:26 Trondd-w: on openrc 2018-03-29 20:12:37 id4rk: *setup 2018-03-29 20:12:37 liwakura: yes 2018-03-29 20:12:47 liwakura: https://azarus.ch/files/initramfs-vanilla 2018-03-29 20:12:53 https://azarus.ch/files/vmlinuz-vanilla 2018-03-29 20:12:57 initramfs and kernel image 2018-03-29 20:13:05 (thanks for your effort) 2018-03-29 20:17:29 liwakura: Looking... 2018-03-29 20:19:31 azarus: does this look like yours? https://w1r3.net/Fkuwra.txt 2018-03-29 20:19:57 liwakura: exactly. 2018-03-29 20:20:05 1 to 1 what I'm experiencing 2018-03-29 20:20:18 wasn't experiencing it two days ago 2018-03-29 20:20:37 (or even yesterday, for what it matters) 2018-03-29 20:20:49 i got it 2018-03-29 20:20:57 Oh? 2018-03-29 20:21:03 "/init" is a script by "/bin/sh" interpreter 2018-03-29 20:21:12 but /bin only has a single busybox interpreter, nothing more 2018-03-29 20:21:34 so an error with the initramfs generator 2018-03-29 20:21:36 So they broke init? 2018-03-29 20:21:37 yay 2018-03-29 20:21:40 were you using mkinitfs ? 2018-03-29 20:21:43 Yup 2018-03-29 20:22:02 Can I downgrade my mkinitfs version? 2018-03-29 20:22:33 a thought: why do other users do not experience your issue? 2018-03-29 20:22:51 ... i'm on edge, might be the first one to experience it 2018-03-29 20:23:37 Thank you very much for helping me diagnose it. 2018-03-29 20:23:53 oh crap 2018-03-29 20:23:58 i can reproduce from scratch 2018-03-29 20:24:02 uh oh 2018-03-29 20:24:17 certainly worth a bug report? 2018-03-29 20:25:49 ping ncopa 2018-03-29 20:25:55 i think 35d9bb7bde58c09fada46448d7964ccd4c75dbc9 broke mkinitfs 2018-03-29 20:25:58 vgacon disables amdgpu kernel modesetting ... 2018-03-29 20:26:13 azarus: yeah, definitely worth some bug report 2018-03-29 20:26:32 liwakura: i'll write it, want to be a "witness"? :P 2018-03-29 20:26:51 just do it 2018-03-29 20:27:13 if you wait a second, i'll check if the commit in question is indeed the culprit 2018-03-29 20:27:22 yeah, that'd be preferable 2018-03-29 20:27:49 Hey @ all 2018-03-29 20:28:13 rebooting to unload some modules 2018-03-29 20:28:14 Do you have the Boot problem on edge? 2018-03-29 20:28:24 feuerteufel: yes 2018-03-29 20:28:34 liwakura: Ok, with -d I see openrc read the conf.d/docker file and sets the proxy environment variables but they don't get passed to teh process environment according to /proc 2018-03-29 20:28:57 then I think I have the same here 2018-03-29 20:29:05 Trondd-w: well debugged! 2018-03-29 20:29:15 i suspect they need to be explicitly export'ed 2018-03-29 20:29:19 rc_env_allow? 2018-03-29 20:29:25 oh, that could also be it 2018-03-29 20:29:35 liwakura: i can't see any commit named '35d9bb7bde58c09fada46448d7964ccd4c75dbc9', which repo are we talking about? 2018-03-29 20:29:47 what is this /etc/init.d/httpd script? doesn't seem to reference any real httpd 2018-03-29 20:29:50 you just read the documentation, so you probably have an better idea than i do 2018-03-29 20:29:53 azarus: aports 2018-03-29 20:30:02 ah, ok 2018-03-29 20:30:32 I'm just not familiar with openrc so I don't know what kind of scrubbing/isolation it does. Learning outloud. :) 2018-03-29 20:30:39 that commit causing that error would make total sense to me 2018-03-29 20:31:43 mh, how do i force a reinstall in apk? 2018-03-29 20:32:28 My system is running as a DomU with the Virt kernel 2018-03-29 20:32:36 apk del && apk add 2018-03-29 20:32:37 ? 2018-03-29 20:32:49 liwakura: That did it. 2018-03-29 20:32:56 Trondd-w: nice! 2018-03-29 20:33:16 Thanks for the prodding 2018-03-29 20:33:31 azarus: cant uninstall due to dependencies.. 2018-03-29 20:33:38 :/ 2018-03-29 20:33:48 --force-broken-world ? 2018-03-29 20:35:01 i just removed the dependencies 2018-03-29 20:35:19 ah, that works fine too I suppose 2018-03-29 20:35:40 i fear that the busybox upgrade might have done it.. 2018-03-29 20:36:13 liwakura: somebody already had done it: https://bugs.alpinelinux.org/issues/8751 2018-03-29 20:36:56 alright 2018-03-29 20:37:11 now lets figure out how to do downgrades then -.- 2018-03-29 20:37:29 yeah... :/ 2018-03-29 20:37:40 also; dropping a line that it broke our boot won't hurt 2018-03-29 20:39:20 not sure how to write an comment there 2018-03-29 20:39:42 I already did ;) 2018-03-29 20:44:11 I've also dropped a pointer to the suspected bad commit 2018-03-29 20:44:19 damn finally... 2018-03-29 20:44:23 its working 2018-03-29 20:44:28 id4rk: what was the cause? 2018-03-29 20:45:02 ACTION is also curious 2018-03-29 20:45:46 i removed all the modules i added to the modules file, removed all isntalled drivers.... followed the warnings in the xorg dev log installed the dependancies... it did not work with amdgpu nor ati, im using it now with vesa and fbdev... but at least i got gui.... 2018-03-29 20:46:22 so how did you get the system started? 2018-03-29 20:46:33 feuerteufel: i didn't 2018-03-29 20:46:38 it stopped booting 2018-03-29 20:46:57 but then again, read the bug, a way is explained to get the system booting again 2018-03-29 20:47:20 ACTION has an idea for an quick fix 2018-03-29 20:47:25 Oh? 2018-03-29 20:47:58 you can extend cpio archives by just cat'ing another cpio to the end of it 2018-03-29 20:48:07 its ridiculous fast... 2018-03-29 20:48:12 ridiculous 2018-03-29 20:48:12 Didn't know that! :o 2018-03-29 20:48:54 even better idea 2018-03-29 20:49:16 azarus: if you manage to recreate the initramfs, edit /usr/share/mkinitfs/initramfs-init 2018-03-29 20:49:26 and change the shebang line from #!/bin/sh to #!/bin/busybox sh 2018-03-29 20:49:36 Will try. 2018-03-29 20:49:38 A sec. 2018-03-29 20:50:00 the script later setups all the busybox symlinks on its own 2018-03-29 20:50:37 when i think about it, thats actually a rather good fix for the issue 2018-03-29 20:51:46 liwakura: doesn't quite work for me 2018-03-29 20:52:00 ACTION wonders 2018-03-29 20:52:25 liwakura: ix.io/14eh 2018-03-29 20:53:08 ah, did you chroot in from another boot medium? 2018-03-29 20:53:12 yes 2018-03-29 20:53:22 you need to specify kernel version then 2018-03-29 20:53:27 I did 2018-03-29 20:53:53 like `mkinitfs 4.9.65` ? 2018-03-29 20:54:04 I ran "mkinitfs 4.14.30-0-vanilla 2018-03-29 20:54:06 " 2018-03-29 20:54:20 no, the kernel version that is installed in the system you chrooted to 2018-03-29 20:54:30 yes 2018-03-29 20:54:36 oh 2018-03-29 20:54:40 my system has 4.14.30-0-vanilla installed 2018-03-29 20:55:07 (my boot medium has 4.14.15 fyi) 2018-03-29 20:55:43 string should be 4.14.30 then 2018-03-29 20:56:19 doesn't work either: ix.io/14el 2018-03-29 20:56:49 can you check /lib/modules whats in there? 2018-03-29 20:56:51 in the wiki they say you should execute mkinitfs like this: 2018-03-29 20:56:59 mkinitfs $(ls /lib/modules) 2018-03-29 20:57:04 which resolves to: 2018-03-29 20:57:12 mkinitfs 4.14.30-0-vanilla 2018-03-29 20:57:18 shit 2018-03-29 20:57:32 so that fix doesn't work 2018-03-29 20:58:09 it usually should, i suspect your error is some other 2018-03-29 20:58:20 ACTION shrugs 2018-03-29 20:58:32 I don't see why it wouldn't work for me 2018-03-29 20:59:48 i can confirm your string is correct this way 2018-03-29 21:01:05 ... can you also confirm it doesn't work for you either? 2018-03-29 21:01:09 can you run sh -x /sbin/mkinitfs 4.14.30-0-vanilla > mkinitfs.log 2>&1 ? 2018-03-29 21:01:11 it works for me 2018-03-29 21:01:23 wait 2018-03-29 21:01:28 without the edit 2018-03-29 21:01:36 huh what? 2018-03-29 21:01:53 so I should do it with or without the edit? 2018-03-29 21:02:25 i think it doesn't matter 2018-03-29 21:02:31 it works for me with or without the edit 2018-03-29 21:02:59 <_ikke_> Anyone got a clue? Trying to run uwsgi in an lxc container, but it fails on "cap_set_proc(): operation not permitted" (http://tpaste.us/jDyV). Config (http://tpaste.us/Q4Zw). Anyone knows what's happening/ 2018-03-29 21:03:03 ix.io/14eq 2018-03-29 21:03:27 (the same cp error still happens) 2018-03-29 21:05:46 yeah... >rm -rf /tmp/mkinitfs.GNCFcP/lib/modules 2018-03-29 21:05:57 happens before the modules are copied in 2018-03-29 21:09:54 azarus: what busybox version do you have? 2018-03-29 21:12:22 because the cpio -pdm should have copied the directory back in.. 2018-03-29 21:13:59 liwakura: busybox 1.28.2 2018-03-29 21:14:12 is there a way to see on which driver xorg is executed_ 2018-03-29 21:14:17 ? 2018-03-29 21:14:24 id4rk: look at your Xorg.0.log 2018-03-29 21:14:32 azarus: no idea what triggers it that it fails on your machine but works on mine 2018-03-29 21:14:47 ACTION shrugs 2018-03-29 21:14:57 I think I'll take a bath and think about this a bit 2018-03-29 21:16:45 what an nice idea! 2018-03-29 21:17:06 Thanks at this time ... got to go to slep ;) 2018-03-29 21:17:10 I reek and had a hard day at work and am now battling boot issues since yestarday :P 2018-03-29 21:17:23 Will see tomorrow ... 2018-03-29 21:17:24 I think I deserve a bath! 2018-03-29 21:17:43 GN @ all 2018-03-29 21:17:52 Goodnight, feuerteufel 2018-03-29 21:22:34 the screen is too bright, burning my eyes... man all this setting up... thats why after finishing setting up your linux you should never change back to anything else... too much efford lol 2018-03-29 21:22:49 now find out on how to do that.. 2018-03-29 21:33:23 azarus: all GPUs are supported by the AMDGPU kernel 2018-03-29 21:33:34 i guess it means that it works 2018-03-29 21:33:54 id4rk: not every single GPU on the planet is supported by AMDGPU :p 2018-03-29 21:35:41 azarus: lol, file a bug at xorg and complain about it! tell them to fix the message to "All GPU's on this system are..." 2018-03-29 21:35:55 that you found it very traumatizing 2018-03-29 21:36:05 "fix or gtfo" 2018-03-29 21:36:23 very arrogant these amd guys... 2018-03-29 21:36:38 ok, i'm off 2018-03-29 21:36:44 gnite sir 2018-03-29 21:37:29 tmux detach 2018-03-29 21:37:33 whops 2018-03-29 21:43:17 are there any games on the AL repo_ 2018-03-29 23:22:52 hello... so I was doing some more research and I'd really like to stick with alpine 2018-03-29 23:23:13 but I do need the kernel to be deblobbed 2018-03-29 23:23:47 I'm looking through the wiki right now but I don't see anything about using a custom kernel 2018-03-29 23:26:24 https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2018-03-29 23:26:37 or clone aports and just create your own variant config for any of the linux-* packages 2018-03-29 23:34:54 Linux-libre is just a giant patch so just apply that 2018-03-29 23:35:39 thanks scv! duncan^ I've been getting different answers on if that will work 2018-03-29 23:35:52 like, with the linux-firmware components in the kernel 2018-03-29 23:36:16 also, one more question... what's the difference between the standard, extended and virtual images? 2018-03-29 23:36:33 could it be that the virtual image doesn't include drivers? 2018-03-29 23:36:54 also duncan^ if I'm wrong on linux-libre please correct me 2018-03-29 23:38:26 duncan^: I might have been wrong if it is supposed to work on ubuntu: https://askubuntu.com/questions/123145/how-do-i-install-the-linux-libre-kernel 2018-03-29 23:49:28 Whats the difference between the standard, extended and virtual images? :) 2018-03-30 00:14:45 ok, I'll try digging through that 2018-03-30 00:14:54 but there really should be documentation of this somewhere 2018-03-30 00:15:11 There is 2018-03-30 00:15:44 The wiki has a page about custom ISO images, which points users to the scripts that build the installer images 2018-03-30 00:16:22 and if you are unhappy with the docs... the only credible reponse is "well-volunteered" 2018-03-30 00:18:56 wait... the last change to https://git.alpinelinux.org/cgit/alpine-iso/ was 20 months ago 2018-03-30 00:22:15 aports/scripts 2018-03-30 00:22:24 is what you should look at 2018-03-30 00:23:37 nice 2018-03-30 00:23:41 premium ^ 2018-03-30 00:24:13 very high quality 2018-03-30 00:24:25 ? 2018-03-30 00:24:53 just wondering why the freenode matrix relay still hasn't been moved to a server with something better than 9600 baud 2018-03-30 00:25:03 considering that happens on the regular 2018-03-30 00:25:51 There have been many netsplits other than Matrix tonight 2018-03-30 00:26:05 just a pity it isn't federated is all 2018-03-30 00:26:58 duncan^: https://git.alpinelinux.org/cgit/aports/scripts/ doesn't exist 2018-03-30 00:28:46 It is a directory in the aports git repository. 2018-03-30 00:28:58 Clone the git repository. 2018-03-30 00:29:18 I don't know how gitweb works 2018-03-30 00:29:40 Maybe you are meant to look through the tree first or something 2018-03-30 00:59:31 aports/scripts just has bootstrap.sh mkimage-yaml.sh mkimg.minirootfs.sh genapkovl-dhcp.sh mkimage.sh mkimg.standard.sh genapkovl-xen.sh mkimg.arm.sh mkimg.xen.sh genrootfs.sh mkimg.base.sh 2018-03-30 00:59:53 nothing about thr virtual image or the extended image in there 2018-03-30 01:03:29 also does alpine support kernel updates? 2018-03-30 01:04:08 gnulligan: as in while the kernel is still running or kernel updates in general? 2018-03-30 01:04:38 well, I'd like to use a custom kernel on a running system 2018-03-30 01:05:04 so not necessarily while it's still running, so long as it's possible 2018-03-30 01:13:03 gnulligan: Take a peak at the actual files for a change 2018-03-30 01:13:20 They build all of the images you see in Downloads 2018-03-30 01:13:35 Ugh. Life is too difficult for some people. 2018-03-30 01:16:10 duncan^: I am looking through them, I'm not seeing anything on the virtual image 2018-03-30 01:17:17 mkimage.standard.sh 2018-03-30 01:17:28 It is fairly obvious 2018-03-30 01:17:48 all it takes is to actually open the file 2018-03-30 01:18:24 zzzz 2018-03-30 01:19:11 kernel_flavors="virt" 2018-03-30 01:19:26 ^ what does that mean? 2018-03-30 01:20:18 what does that change? 2018-03-30 01:20:35 it pulls in the virt kernel flavor 2018-03-30 01:20:57 what's the difference between the virt kernel and the standard kernel? 2018-03-30 01:21:29 This is in the kernel package 2018-03-30 01:21:54 pkgs.alpinelinux.org will let you find the package in question. 2018-03-30 01:21:59 Consult the apkbuild. 2018-03-30 01:22:24 It is the case that the virt package is smaller. 2018-03-30 01:23:14 Are you telling me I need to literally dig through the entire project source code to find out what the difference is between the virtual and standard images? 2018-03-30 01:23:33 No 2018-03-30 01:23:43 You have already been told the difference. 2018-03-30 01:23:50 no I haven't 2018-03-30 01:23:59 saying "one is for virtual, the other isnt" 2018-03-30 01:24:01 The virt package is smaller and does nkt include all the things in standard 2018-03-30 01:24:12 that means nothing 2018-03-30 01:24:12 This has been said a number of times 2018-03-30 01:24:24 then why do you keep repeating it? 2018-03-30 01:24:26 Then look t 2018-03-30 01:24:55 Because that is what it is. There are things in standard a virtual system oes 2018-03-30 01:24:59 does not need 2018-03-30 01:25:01 etc 2018-03-30 01:25:08 yes, I understand 2018-03-30 01:25:19 what things were removed? 2018-03-30 01:25:23 It is fairly broad. 2018-03-30 01:25:40 Mostly platform specific drivers. 2018-03-30 01:25:54 See, that's what I'm trying to find out 2018-03-30 01:26:11 You did not ask pertinent questions, just vague ones 2018-03-30 01:26:16 What drivers were removed? Which ones are still remaining 2018-03-30 01:26:28 I asked what the difference was 2018-03-30 01:26:39 There is a lot, so you need to look at things on a case by case basis 2018-03-30 01:26:46 You have the source, go crazy 2018-03-30 01:27:04 Or pay someone a consulting fee to mail you a PDF or something 2018-03-30 01:34:00 is there a list of packages removed somewhere? 2018-03-30 01:35:35 diff --side-by-side config.vanilla.x86_64 config.virt.x86_64 2018-03-30 01:40:05 i have a question about buildrepo 2018-03-30 01:40:49 i want to use it to build my own aports, that require dependencies from alpine's main 2018-03-30 01:41:22 there the buildrepo -r switch, but i can't find out what it expects, or anything in the sourcecode that brings it to life 2018-03-30 04:15:50 virtual is intended for VMs 2018-03-30 04:27:32 duncan^: so yeah, about that matrix stability 2018-03-30 07:22:20 did anything yet happen about https://bugs.alpinelinux.org/issues/8751 ? 2018-03-30 07:35:44 This is strange. newly created initramfs's created my mkinitfs don't have *anything* in /bin ? 2018-03-30 07:35:59 azarus, this happend onlyh when updating bb or when creating initramfs? 2018-03-30 07:36:18 clandmeter: creating initramfs with an updated bb 2018-03-30 07:36:34 seems to work on boot.a.o 2018-03-30 07:37:21 i just booted in vbox and the initramfs is fine. 2018-03-30 07:37:46 (i did modify mkinitfs.conf tough) 2018-03-30 07:37:54 features="ata base ide scsi usb virtio ext4 cryptsetup zfs" 2018-03-30 07:38:36 i think it has most of them except cryptsetup 2018-03-30 07:38:45 ... and zfs, probably 2018-03-30 07:38:54 netboot supports zfs 2018-03-30 07:39:00 Oh, nice. 2018-03-30 07:39:17 well i didnt test it, but its included :) 2018-03-30 07:39:58 Anyway. Do you want to take a look at the initramfs created with that config? 2018-03-30 07:40:17 (that doesn't boot) 2018-03-30 07:47:49 azarus, are you on stable or edge? 2018-03-30 07:47:54 clandmeter: edge 2018-03-30 08:00:04 azarus, you get any error when craeting initramfs? 2018-03-30 08:00:10 clandmeter: no 2018-03-30 08:02:08 which kernel? 2018-03-30 08:08:56 clandmeter: tried with -vanilla and -hardened, same result with either 2018-03-30 08:20:10 azarus, i can reproduce 2018-03-30 08:20:25 Great :) 2018-03-30 08:44:39 ncopa, can you look into this issue? seems related to busybox. 2018-03-30 08:47:44 wtf, irc bridge is totally broken now? 2018-03-30 13:24:53 figured out my seamonkey issue with the blurry text, there was a bug report related to it... also found a workaround for blurry images in mozilla browsers (for those who experience it, seems to be limited to ia32) 2018-03-30 14:05:44 where does setup-alpine put the grub.cfg file after a lvm installation? 2018-03-30 14:07:42 fdisk -l gives 2 partitions on my main drive, one of which is bootable 2018-03-30 14:08:02 but when I tried mounting it and looking inside, I didn't see any grub.cfg file... 2018-03-30 14:08:13 I need to put it in a custom location to boot 2018-03-30 14:08:30 gnulligan: whats in the partition you tried to mount? 2018-03-30 14:10:15 System.map-hardened config-hardened initramfs-hardened ldlinux.sys libutil.c32 mboot.c32 vesamenu.c32 boot/ extlinux.conf ldlinux.c32 libcom32.c32 lost+found menu.c32 vmlinuz-hardened 2018-03-30 14:10:29 thats using the syslinux bootloader 2018-03-30 14:10:37 your boot config is in extlinux.conf then 2018-03-30 14:10:59 oh... I have grub installed in my mobo rom 2018-03-30 14:11:13 as the coreboot payload 2018-03-30 14:12:02 then it wont boot, i guess 2018-03-30 14:12:26 you need to place the grub config there, then 2018-03-30 14:12:43 i think best way would be to remove syslinux, install grub, and then make gru generaste the config 2018-03-30 14:12:47 *grub generate 2018-03-30 14:12:55 yeah... 2018-03-30 14:14:10 lol weird bug... typing halt in the preinstall environment doesn't turn it off... the last line of text is "reboot: System halted" 2018-03-30 14:14:26 no, thats what halt does 2018-03-30 14:14:32 i think you want poweroff 2018-03-30 14:14:47 oh... well halt behaves differently on arch then 2018-03-30 14:15:22 halt is literally for getting the OS to sync and stop operating, so one can shut the PSU off 2018-03-30 14:15:37 maybe you remember this: http://i0.kym-cdn.com/photos/images/original/001/286/950/e05.jpg 2018-03-30 14:16:30 lol yeah, I guess other OSes use halt differently now and I was used to that 2018-03-30 14:16:33 its actually a noteworthy distinction, because stuff like the RPi cant poweroff, just halt 2018-03-30 14:18:37 I see 2018-03-30 14:20:27 well I just tried running setup-alpine with disks: none and store configs: none 2018-03-30 14:21:07 when it asked to enter apk cache directory I left it at /media/usb/cache, the default, and mount failed 2018-03-30 14:21:32 gave error mount: mounting /dev/usbdisk on /media/usb failed: Resource busy 2018-03-30 14:21:44 already mounted, i guess 2018-03-30 14:22:10 will that impact me installing grub in the preinstall environment? 2018-03-30 14:22:25 dont think so 2018-03-30 14:24:11 liwakura: got any news regarding the mkinitfs problem? 2018-03-30 14:24:22 no, i was sleeping, mostly 2018-03-30 14:24:26 OK 2018-03-30 14:24:47 and since it doesn't affect me, im not looking much except aiding other people in debugging 2018-03-30 14:25:11 ah, fair enough 2018-03-30 14:25:20 (im neither a maintainer nor some other important person) 2018-03-30 14:25:22 cool, looks like apk worked fine 2018-03-30 14:26:09 liwakura: ah, ncopa issued a fix 2018-03-30 14:26:28 *thumbs up* 2018-03-30 14:26:58 <_ikke_> azarus: does it have to do with unsafe symlinks and busybox? 2018-03-30 14:27:25 _ikke_: yes 2018-03-30 14:27:29 https://bugs.alpinelinux.org/issues/8751 2018-03-30 14:28:36 Yay, system boots again :D 2018-03-30 14:28:44 ACTION does the happy dance 2018-03-30 14:28:49 this looks nicely handled 2018-03-30 14:28:59 <_ikke_> azarus: was about to propose that earlier, but had to do other things 2018-03-30 14:29:01 Yeah, very smooth. Worked instantly. 2018-03-30 14:44:54 if I install clang from apk and then try to compile hello world, I get linker errors 2018-03-30 14:45:16 https://clbin.com/skzPX 2018-03-30 14:47:00 you need to install the header files 2018-03-30 14:47:03 musl-dev for lib 2018-03-30 14:47:05 *libc 2018-03-30 14:47:39 and *-dev packages in general for everything you want to link against 2018-03-30 14:49:00 now it's just libgcc: https://clbin.com/nPCr5 2018-03-30 14:49:05 even though I have libgcc instnalled 2018-03-30 14:50:50 hm. 2018-03-30 14:50:56 maybe thats because you are using clang 2018-03-30 14:51:09 but thats above my head, linker stuff like that 2018-03-30 14:51:15 using clang with musl is my use case for alpine linux 2018-03-30 14:51:24 anyone at all run Seamonkey in Alpine? 2018-03-30 14:52:20 "have libgcc installed" is probably juse the useless libgcc_s.so 2018-03-30 14:52:33 you need whatever devel package provides libgcc.a 2018-03-30 14:53:10 apk search gcc: https://clbin.com/17ZAb 2018-03-30 14:53:13 I'll try gcc itself 2018-03-30 14:53:38 that solved the problem 2018-03-30 14:54:34 if clang also depends on it, libgcc.a should probably be factored out from gcc package 2018-03-30 14:54:52 e.g. a gcc-target-libs-dev package or something 2018-03-30 14:57:19 agreed 2018-03-30 15:02:29 Who do I talk to about getting a piece of software I wrote as a package in alpine? 2018-03-30 15:03:11 gdh: send a mail to the mailing lists or something 2018-03-30 15:03:18 preferably with a patch to aports 2018-03-30 15:03:34 gdh: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2018-03-30 15:03:43 creating one is much easier than getting it upstream 2018-03-30 15:04:12 liwakura, thank you. I need that, although my package should probably replace another as well. 2018-03-30 15:04:26 afaik APKBUILD has an replaces mechanism 2018-03-30 15:05:03 Oh cool. 2018-03-30 15:05:26 I will make a package and then submit a post to the mailing list. 2018-03-30 15:12:35 Regarding https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Code_review 2018-03-30 15:13:01 Is the code style meant for the source code of the package scripts or the actual source code of the software in the package? 2018-03-30 15:13:04 hello 2018-03-30 15:13:18 is anyone using alpine on laptops with nvidia ? 2018-03-30 15:13:47 gdh: the source code of the package scripts, mostly :) 2018-03-30 15:14:05 i mean, did someone suceeded to run optimus/primusrun on alpine 2018-03-30 15:14:11 azarus, thank goodness; my source would not pass the style. :) 2018-03-30 15:14:40 gdh: could I take a peek at it? 2018-03-30 15:14:49 https://github.com/gavinhoward/bc 2018-03-30 15:14:54 Meant to replace the GNU bc. 2018-03-30 15:14:54 Thanks. 2018-03-30 15:15:17 Ooh, cool 2018-03-30 15:15:31 I am about to do a release in an hour. 2018-03-30 15:15:47 It's been fully fuzzed and the fuzzer couldn't crash it. 2018-03-30 15:15:55 azarus: fun fact: your issue also broke my initramfs 2018-03-30 15:16:14 liwakura: hehe, your own initramfs? 2018-03-30 15:16:27 even if im using some custom thing instead of mkinitfs / cpio 2018-03-30 15:16:29 gdh: Well that's nice :) 2018-03-30 15:16:47 liwakura: argh, what a shame 2018-03-30 15:17:40 ok, so I did apk add zfs 2018-03-30 15:17:51 but the modules aren't loaded 2018-03-30 15:18:06 gnulligan: modprobe fzs 2018-03-30 15:18:07 zfs* 2018-03-30 15:18:10 and modprobe zfs returns "module zfs not found in modules.dep" 2018-03-30 15:18:20 are you in a chroot? 2018-03-30 15:18:31 nope, pre-install environment 2018-03-30 15:18:48 yeah that probably won't work 2018-03-30 15:18:57 /lib/modules is read only 2018-03-30 15:19:02 oh... :( 2018-03-30 15:19:11 install alpine to a usb stick or something 2018-03-30 15:19:17 mount -o remount /lib/modules -rw? 2018-03-30 15:19:18 and then you'll be able to do it 2018-03-30 15:19:26 not sure that would work, gnulligan 2018-03-30 15:19:27 why are there so many arch="x86_64" packages? 2018-03-30 15:19:47 ok azarus how would I do it with a second usb stick? 2018-03-30 15:20:00 install to usb stick... then what? 2018-03-30 15:21:18 gnulligan: install alpine from that usb stick 2018-03-30 15:21:24 treat it as your installation environment 2018-03-30 15:21:44 maybe follow my installation guide 2018-03-30 15:21:53 azarus: where? 2018-03-30 15:21:58 on the wiki 2018-03-30 15:22:05 a link will be here in a sec 2018-03-30 15:22:32 https://wiki.alpinelinux.org/wiki/Setting_up_ZFS_on_LUKS 2018-03-30 15:22:48 just disregard the LUKS bit 2018-03-30 15:23:06 and use GRUB as your bootloader if you don't want a separate /boot filesystem 2018-03-30 15:23:41 azarus: I already have grub as my bootloader :) 2018-03-30 15:23:51 but... can grub read encrypted zfs? 2018-03-30 15:23:57 nope 2018-03-30 15:24:03 read my installation guide 2018-03-30 15:24:21 Note that you must install the /boot/ directory on an unecrypted partition to boot correctly. 2018-03-30 15:25:35 azarus: that part might be different on my system 2018-03-30 15:25:51 I have coreboot with a just-grub payload 2018-03-30 15:26:06 i have coreboot with a SeaBIOS payload ;) 2018-03-30 15:26:07 so encrypting the /boot partition is fine on say, luks+lvm 2018-03-30 15:26:33 yeah, had a GRUB payload once too, but decrypting anything took like 1 whole minute 2018-03-30 15:26:37 so i gave up on it 2018-03-30 15:26:52 huh, on my thinkpad its pretty fast 2018-03-30 15:27:06 hmm, maybe there aren't that many x86_64 only package.. just a large number of ones I want 2018-03-30 15:33:37 azarus: I see that you are busy, but would you mind to tell how did you solved entering umlauts in libreoffice 2018-03-30 15:33:50 mps: didn't do anything much 2018-03-30 15:33:57 mps: are you on edge or 3.7? 2018-03-30 15:34:09 mixed :/ 2018-03-30 15:34:17 works for me on edge 2018-03-30 15:34:59 I upgraded it from edge, but still doesn't work 2018-03-30 15:35:12 maybe some strange dependency breaks it 2018-03-30 15:36:00 I thought so. nevermind, I will try to see what could be problem 2018-03-30 15:36:52 Guys, when will py-zmq package updated? version 17 is out since many days... 2018-03-30 15:37:30 btw, maybe I should upgrade everything to edge versions 2018-03-30 15:41:22 <_ikke_> shreyansh_k: alpine is currently preparing a new release 2018-03-30 15:45:07 _ikke_: in march?? 2018-03-30 15:45:31 or an sub-relese (lower than minor) ? 2018-03-30 15:45:43 Hey there, I'm trying to build an immutable image for PXE booting, I know there's some guidance on PXE booting but it's rather old, I'd also like to integrate it into a build pipeline. Anyone got any experience? 2018-03-30 15:46:15 i had some experiments with alpine and PXE 2018-03-30 15:46:42 Hey guys, I am trying to get 3.7 ARM to boot on PiZeroW and see that has a bug of "Can't open /dev/ttyS0" on initial boot. The recommendation was enable_uart=1 but that doesn't seem to work. It also says disable bluetooth but I need bluetooth. Any ideas? 2018-03-30 15:47:56 Ah nvm...pebkac. I see a fix on the blog. I will try that now. 2018-03-30 15:49:40 liwakura: how'd that go for you? 2018-03-30 15:50:20 also would it be better to ask this in #alpine-devel you reckon? 2018-03-30 15:50:38 <_ikke_> liwakura: ncopa anounced it 2018-03-30 15:50:42 I've been looking at the netboot project and it's using undocumented scripts I can't make sense of :( 2018-03-30 15:55:35 _ikke_: So, that means that the update will be released with version, right? I think edge is bleeding, so it should have the latest version, isn't it? 2018-03-30 15:55:59 *with next version 2018-03-30 15:57:08 <_ikke_> shreyansh_k: It means that the focus is shifting from making sure every package is up-to-date to making sure everything is working/stable 2018-03-30 15:58:12 _ikke_: Oh, cool. best guess of eta to new release? week, month, quarter... 2018-03-30 16:00:32 <_ikke_> I wouldn't know 2018-03-30 16:02:25 _ikke_: okay, thank you. 2018-03-30 16:02:57 <_ikke_> shreyansh_k: You might get the maintainer to update the package in time 2018-03-30 16:14:16 benw2: learned alot about the alpine initrd, but never had something production-ready due to lack of use 2018-03-30 16:14:32 _ikke_: on the ML? im not really subscribed to anything 2018-03-30 16:14:44 <_ikke_> liwakura: No, in #alpine-devel 2018-03-30 16:15:00 ACTION sighs 2018-03-30 16:15:06 <_ikke_> @ncopa │ from next week i will start set up builders for v3.8 2018-03-30 16:15:08 liwakura: well wish me luck! Hoping to use in production and keep things simple (once I've worked out the complicated stuff) 2018-03-30 16:17:39 benw2: it should not be complicated to make netboot image. yesterday I remastered iso to boot it over http on HP DL360 2018-03-30 16:18:09 it worked, with some small quuirks 2018-03-30 16:18:11 benw2: i remember booting a device via pxe, but then loading the rest of the system via USB device 2018-03-30 16:18:24 because the BIOS could not boot from USB itself and it had no cdrom slot 2018-03-30 16:20:54 haha jeez that sounds messy 2018-03-30 16:21:25 so I figure, generate rootfs via https://github.com/jirutka/alpine-make-rootfs 2018-03-30 16:21:32 grab kernel with apk 2018-03-30 16:21:41 just to do a mkinitfs now 2018-03-30 16:23:54 libxml2-dev gives me a .so - is there a way to get libxml2.a instead? 2018-03-30 16:24:27 hmm looks like mkinitfs requires the modules to be present in the rootfs to create it.. back to drawing board! 2018-03-30 16:33:24 why is --disable-static in the libxml2 APKBUILD? 2018-03-30 16:39:00 <_ikke_> andrewrk: https://git.alpinelinux.org/cgit/aports/commit/?id=67ffa41bb78 2018-03-30 16:40:32 _ikke_, I think --disable-static was a mistake 2018-03-30 16:40:40 the static build does not depend on python 2018-03-30 16:41:32 the docker use case would benefit from the static lib 2018-03-30 16:41:45 alpine is optimized for small size 2018-03-30 16:41:54 static libs are good for small size 2018-03-30 16:41:58 so disabling static libs seems reasonable forme 2018-03-30 16:42:11 if you build static then you can delete all the unused functions 2018-03-30 16:42:15 andrewrk: not when libressl or openssl are remotely included 2018-03-30 16:42:29 I don't understand what you are saying 2018-03-30 16:42:34 what does remotely included mean? 2018-03-30 16:42:46 dependency from dependency from dependency or so 2018-03-30 16:43:29 having only a single binary in the system is like, the only situation where you save space using static linking 2018-03-30 16:43:46 it's pretty clear what the correct thing to do is: provide dynamic and static libraries. let users decide which ones they want. docker users have tools to remove unused libraries 2018-03-30 16:44:13 you are free to create your own APKBUILD's 2018-03-30 16:44:28 oh I did not know that was made easy 2018-03-30 16:44:42 you can just copy and edit the existing ones 2018-03-30 16:44:47 so I can do something like cp the upstream APKBUILD, edit, and then use that with apk? 2018-03-30 16:45:11 at least thats how i would modify packages 2018-03-30 16:45:19 ok if it's that easy then I am happy 2018-03-30 16:58:26 I am guessing that depending on the C standard library and POSIX does not need to be listed under depends when building a package. Is that correct? 2018-03-30 16:59:58 <_ikke_> No, and abuild can trace most dependencies anyway 2018-03-30 17:01:06 Okay. 2018-03-30 17:05:47 I am getting an error trying to run abuild checksum: 2018-03-30 17:05:48 https://bpaste.net/show/c90762ac52ac 2018-03-30 17:05:56 What did I do wrong? 2018-03-30 17:09:10 <_ikke_> How does your APKBUILD look like? 2018-03-30 17:10:08 https://bpaste.net/show/cb945927e305 2018-03-30 17:12:04 <_ikke_> Works for me 2018-03-30 17:12:31 <_ikke_> Seems like a version mismatch 2018-03-30 17:13:12 So I should do: "apk update; apk upgrade" and try again? 2018-03-30 17:13:59 <_ikke_> You could try 2018-03-30 17:14:58 Worked. 2018-03-30 17:15:03 <_ikke_> alright 2018-03-30 17:22:51 how to force apk to show errors instead of just number of errors 2018-03-30 17:24:07 mps: had this issue today, fixed them with `apk fix` 2018-03-30 17:24:15 -vvv didn't seem to do anything.. 2018-03-30 17:25:01 liwakura: what do you mean with 'apk fix'? run it as such? 2018-03-30 17:25:55 yeah 2018-03-30 17:27:24 didn't help to me :( 2018-03-30 19:15:13 Who's the resident rpi bcrm module wizard? I can't get this thing figured out and very little docs that I can find 2018-03-30 19:20:20 Isn't there a b43-firmware package ? 2018-03-30 19:27:40 was there a change to kernels? 2018-03-30 19:49:03 Hey @ all 2018-03-30 19:50:14 About this issue from yesterday (or this early morning) 2018-03-30 19:50:28 https://bugs.alpinelinux.org/issues/8751 2018-03-30 19:51:42 I'm not qiet shure, what I have to do, tobring the system back on line 2018-03-30 19:52:30 <_ikke_> feuerteufel: first try this: apk add cpio && apk fix mkinitfs 2018-03-30 19:55:38 In a chroot? 2018-03-30 19:55:46 <_ikke_> most likely, yes 2018-03-30 19:56:40 Mom, I'll try 2018-03-30 20:16:45 I'm still on ... 2018-03-30 20:17:25 i have to mount in in a lv and I'm not prepared yet :-( 2018-03-30 21:09:07 well boys I give up. would have been nice to use alpine for a new hotspot os 2018-03-30 21:11:34 _ikke_: Ok, it's done ... 2018-03-30 21:11:59 _ikke_: And now try to boot? 2018-03-30 21:22:05 Does anyone know why vmhgfs-fuse will mount my VMware shared folders, but mount -t vmhgfs is failing? 2018-03-30 21:22:16 vmhgfs-fuse .host:/Test /mnt/hgfs : working 2018-03-30 21:22:26 mount -t vmhgfs .host:/Test /mnt/hgfs : failing 2018-03-30 21:24:48 _ikke_: Now the tty is not runnig 2018-03-30 21:25:05 "xenconsole: Could not read tty from store: Success" 2018-03-30 21:31:01 blf: mount -t does not use the fuse subsystem 2018-03-30 21:32:58 is the setup-webconf still available? 2018-03-30 21:33:13 doesn't seem so 2018-03-30 21:33:43 then it would be nice if somone updates https://wiki.alpinelinux.org/wiki/Experiences_with_OpenVPN-client_on_ALIX.2D3 2018-03-30 21:34:13 im trying to get that done... but im not sure if my skills are good enough on alpine 2018-03-30 21:34:21 on openbsd it was easier to do 2018-03-30 21:36:00 id4rk: i think its named setup-acf now 2018-03-30 21:36:13 liwakura: the documentation for mount.vmhgfs suggest using mount -t vmhgfs 2018-03-30 21:36:39 link on that please? 2018-03-30 21:36:56 mh 2018-03-30 21:37:09 liwakura: check Examples: https://manpages.debian.org/jessie/open-vm-tools/mount.vmhgfs.8.en.html 2018-03-30 21:37:13 mount has an functionality to call external wrappers instead 2018-03-30 21:37:57 liwakura: right, I thought that's what was supposed to happen when I run: mount -t vmhgfs. But it fails :/ 2018-03-30 21:38:57 i dont know if busybox even has that functionality 2018-03-30 21:39:48 mh, it has, and its enabled in alpine 2018-03-30 21:40:03 blf: can you strace the failing mount command and paste the log? 2018-03-30 21:40:31 strace mount -t vmhgfs .host:/Test /mnt/hgfs 2>&1 | curl -F upload=@- https://w1r3.net 2018-03-30 21:42:28 sure can. one minute 2018-03-30 21:43:31 liwakura: i will try thanks 2018-03-30 21:44:29 liwakura: could not resolve host wlr3.net 2018-03-30 21:44:53 its an 1, not n l 2018-03-30 21:45:03 oh 2018-03-30 21:45:03 :) 2018-03-30 21:45:16 its 1337speak of wire 2018-03-30 21:45:36 https://w1r3.net/qcrMWe.txt 2018-03-30 21:46:35 oh, can you repeat with strace -f mount ... ? 2018-03-30 21:49:09 one sec. here is the strace for working vmhgfs-fuse: https://w1r3.net/aoNKmF.txt 2018-03-30 21:50:14 with strace -f mount...: https://w1r3.net/pmuuPD.txt 2018-03-30 21:52:25 liwakura: you were right 2018-03-30 21:52:41 i guess im going to format now and start from scratch 2018-03-30 21:58:47 liwakura: I guess it looks like mount.vmhgfs might be broken? 2018-03-30 21:59:55 probably 2018-03-30 21:59:57 I guess I could try installing and running the "Standard" alpine distribution, rather than the "Virtual" distribution and see if it makes any difference 2018-03-30 22:00:07 it doesn't give in the fuse. string into the mount syscall.. 2018-03-30 22:01:07 liwakura: well, thanks for taking a look anyways :) 2018-03-30 22:04:07 does anyone know if this is fixed with more current versions of openrc https://bugs.alpinelinux.org/issues/4810 ? 2018-03-30 22:06:07 Invader_Bork: from the comments, i dont think so 2018-03-30 22:06:09 is there some magic trick to installing packages with a "+" in the name? I can see crypto++-dev in the online package listing but when I go to apk add it in my Dockerfile I see "ERROR: unsatisfiable constraints: crypto++-dev (missing): ..." 2018-03-30 22:07:19 mattcode: is edge/testing repo enabled? 2018-03-30 22:07:21 liwakura, I've definitely found that Alpine has one of the most intuitive ports/packages systems. 2018-03-30 22:07:45 TomTheDragon: it definitely sucks less than debians 2018-03-30 22:07:56 liwakura: i'm doing "FROM alpine:edge" at the top of my Dockerfile. do I need to do anything else to enable the edge repository? 2018-03-30 22:08:46 liwakura, last comment was from more than two years ago, so I was wondering if maybe somebody tested a more recent version out of curiosity. we are on 0.24.1 and the latest is 0.35.5 2018-03-30 22:09:03 mattcode: i can reproduce, im investigating. 2018-03-30 22:11:21 liwakura, the ability to "build dependencies" as opposed to install from packages is pretty nice. especially if I end up needing to support i586 after it's dropped. 2018-03-30 22:14:20 mattcode: seems to be indeed a bug, but it only affects the docker container, but not the host 2018-03-30 22:14:31 even if both run the same apk version 2018-03-30 22:18:50 but things like websocket++ work 2018-03-30 22:23:23 i think one should file a bug on b.a.o 2018-03-30 22:24:34 Ok, I give up for now with that one https://bugs.alpinelinux.org/issues/8751 2018-03-30 22:24:48 liwakura: i added the edge/testing repository and it worked :$ 2018-03-30 22:25:45 See tomorrow if I'm able to find a solution 2018-03-30 22:26:14 mattcode: ah... 2018-03-30 22:26:19 GN @ all 2018-03-30 22:37:03 my alpine install is panicking at boot https://imgur.com/a/Wj1cE, how do i fix this? 2018-03-30 22:40:18 Adran: likely https://bugs.alpinelinux.org/issues/8751 2018-03-30 22:40:51 oh man 2018-03-30 22:50:32 i did that temporary fix now i'm getting a different error 2018-03-30 22:53:11 liwakura: https://imgur.com/a/DzLij 2018-03-30 22:54:17 looks like your storage driver is not loaded 2018-03-30 22:54:40 jn__: then the initramfs would fail to mount the block device 2018-03-30 22:54:52 = initramfs error, and kernel panic bc of killed init 2018-03-30 22:55:05 this actually like no initramfs is loaded 2018-03-30 22:55:16 and the kernel itself tries to mount root, but fails 2018-03-30 22:55:31 good point 2018-03-30 23:07:37 liwakura: how can fix it? 2018-03-30 23:08:15 Adran: I dont know, but i would look at /boot/initramfs-... if that is okay 2018-03-30 23:08:37 and check if the bootloader config also loads it 2018-03-31 01:02:27 hello 2018-03-31 01:04:08 i am trying to install Nvidia drivers on laptop. can someone let me know what might cause this error: "./NVIDIA-Linux-x86_64-375.82.run: line 1085: ./nvidia-installer: not found" 2018-03-31 01:23:12 The last update I did broke the init system. kernel panic - not syncing: no working init found. Im assuming I can reinstall the init system from the live cd right? 2018-03-31 01:25:01 https://bugs.alpinelinux.org/issues/8751 2018-03-31 01:25:41 Thanks for the Link. I never thought to look in the bugs. 2018-03-31 01:26:12 seems to be a greater annoyance to several people 2018-03-31 01:26:57 It doesnt really bother me as the machine im running is only for long term storage/photo backups. 2018-03-31 01:29:45 Forgive me as I am new to linux on this level, but is there any good place to look on how to repack intramfs 2018-03-31 01:30:15 cpio does this 2018-03-31 01:30:35 it works a bit like tar, except its an different format 2018-03-31 01:31:14 you can use mkinitfs 2018-03-31 01:31:53 like: mkinitfs -o initramfs-4.9.91-1.0 4.9.91-1.0 2018-03-31 01:32:09 so just chroot into my os hdd and run that command in the terminal 2018-03-31 01:32:45 I think after tonight I am going to disable edge repos 2018-03-31 01:32:51 hahaha 2018-03-31 09:55:00 I do apk add mediainfo 2018-03-31 09:55:09 then I can't execute mediainfo 2018-03-31 09:55:51 http://dpaste.com/3QMPD7Z 2018-03-31 09:59:02 seems like the install went wrong 2018-03-31 09:59:35 https://www.filebot.net/forums/viewtopic.php?t=4211 2018-03-31 11:14:49 qswz: i cant reproduce that 2018-03-31 11:17:30 https://www.youtube.com/watch?v=V_Ac-SVYI48 2018-03-31 13:44:28 liwakura: you've java installed? 2018-03-31 13:44:46 I should say I'm running alpine on docker 2018-03-31 13:45:14 https://hub.docker.com/_/node/ node:carbon-alpine 2018-03-31 14:11:06 hey liwakura, any progress on alpine on c2? 2018-03-31 14:16:13 kahiru: yeah, strfry is working on it 2018-03-31 14:16:16 :P 2018-03-31 14:17:18 good strategy :) 2018-03-31 14:18:07 Hello @ all 2018-03-31 14:18:57 kahiru: he was working on the bootloader actually after i lost interest yesterday 2018-03-31 14:19:33 My edge system also run's again (https://bugs.alpinelinux.org/issues/8751) 2018-03-31 14:20:13 There was a second Prob on XEN ... Sorry 2018-03-31 14:22:30 cool, cool. looking forward to hearing more about it 2018-03-31 14:24:26 kahiru: You want to heare mor from me? 2018-03-31 14:26:40 feuerteufel: it was more about strfry working on odroid c2 support 2018-03-31 14:27:07 liwakura: i didn't do nothing, the computer was working on it ;P 2018-03-31 14:27:15 Oh, Ok never mind! 2018-03-31 14:29:11 i'm having a weird problem with abuild-apk 2018-03-31 14:29:26 a simple "abuild-apk add" returns: "1 error; 657 MiB in 396 packages" 2018-03-31 14:29:31 feuerteufel: I'd be glad to hear it but I'm afraid I wouldn't be able to help you with problems on xen 2018-03-31 14:29:44 how can i figure out what's going on? it causes all my builds to fail :( 2018-03-31 14:30:09 <_ikke_> strfry: try apk fix 2018-03-31 14:32:29 kahiru: Oh, the help was quit easy! 2018-03-31 14:32:30 _ikke_: ah, thanks 2018-03-31 14:33:30 I only rebooted the XEN Virt :-) 2018-03-31 14:34:29 one of the rare cases where "turning it off and back on" actually worked, huh? 2018-03-31 14:37:38 Yes 2018-03-31 14:43:45 In german, we have a corny Joke: 2018-03-31 14:44:21 "Reboot am morgen vertreibt Kummer und Sorgen!" 2018-03-31 14:45:13 Sorry, I don't Know how to translate that one in english 2018-03-31 14:45:29 google translate does a pretty good job on this one I'd say 2018-03-31 14:48:42 Yea, but the rhyme ist gone ;) 2018-03-31 14:56:25 <_ikke_> 9works in dutch as well 2018-03-31 14:59:54 "A reboot a day keeps your worries at bay", perhaps 2018-03-31 15:01:53 liwakura: what does "which mediainfo" does for you? 2018-03-31 15:01:58 please 2018-03-31 15:03:24 qswz: nothing, because i dont have it installed 2018-03-31 15:03:39 13:14 < liwakura> qswz: i cant reproduce that 2018-03-31 15:03:45 what did you mean then? 2018-03-31 15:04:20 i could install and use mediawiki like normal 2018-03-31 15:04:40 ah, on docker? because you don't have it now if I understand 2018-03-31 15:04:46 ye 2018-03-31 15:05:03 ok thanks 2018-03-31 15:05:23 we wre not using the same image I bet, I'll try with the pure alpine image, thx 2018-03-31 15:20:39 has anyone build plan9port on alpine? having a bit of difficulty with the latest git 2018-03-31 15:26:41 Ok mediainfo worked 2018-03-31 15:27:08 Can I ask the diff between apk add --no-cache mediainfo (worked fine) and apk add mediainfo (not worked) 2018-03-31 15:28:10 <_ikke_> qswz: what didn't work in the last case? 2018-03-31 15:29:10 http://dpaste.com/1RGRHWZ 2018-03-31 15:29:20 this 2018-03-31 15:54:33 weird: http://dpaste.com/3BD2RSX 2018-03-31 15:55:46 <_ikke_> qswz: try running apk update beforehand 2018-03-31 15:56:03 thx ok 2018-03-31 15:58:20 http://dpaste.com/2QYDT6R added at the end 2018-03-31 15:58:23 same 2018-03-31 16:00:24 interesting they are not same version 2018-03-31 16:00:36 http://dl-cdn.alpinelinux.org/alpine/v3.7 for alpine 2018-03-31 16:03:58 I'll install nodejs on alpine, but I'd curious to understand the difference 2018-03-31 16:04:20 <_ikke_> qswz: difference between what exactly? 2018-03-31 16:04:56 the #### 1 and #### 2 2018-03-31 16:05:05 why mediainfo can't be installed in 2 2018-03-31 16:05:29 apart he fact one is alpine 3.4 the other 3. 2018-03-31 16:05:33 7* 2018-03-31 16:05:50 <_ikke_> Well, that can make a big difference 2018-03-31 16:05:53 sorry node:carbon-alpine is 3.6 alpine is 3.7 2018-03-31 16:06:12 <_ikke_> 3.6 does not have mediainfo as a package 2018-03-31 16:06:14 <_ikke_> https://pkgs.alpinelinux.org/packages?name=mediainfo&branch=v3.6 2018-03-31 16:06:17 v3.6.2-295-g8c0f8e4449 vs v3.7.0-140-g86561a5df4 2018-03-31 16:06:26 Oh! 2018-03-31 16:06:28 thanks 2018-03-31 16:21:52 Is #busybox good for asking questions related to busybox since there only 35 people inn there 2018-03-31 16:22:14 <_ikke_> rosa_: Perhpaps just try 2018-03-31 16:22:44 I did 2018-03-31 16:24:22 hey 2018-03-31 16:24:48 is veracrypt available on alpine i searched the packages but nothing appeared 2018-03-31 16:25:55 <_ikke_> apparently not 2018-03-31 16:28:00 so what do you do when some package isn't available you just try to compile it and hop eit workS>? 2018-03-31 16:28:51 and do standard, extended and miniroot include the grsec patch? or is grsec totally deprecated and only linux-hardened is used by default in all alpine images? 2018-03-31 16:29:28 <_ikke_> linux-grsec was renamed to linux-hardened 2018-03-31 16:29:40 <_ikke_> But that's deprecated as you stated 2018-03-31 16:29:54 oh is it included by default in all the images? 2018-03-31 16:40:10 <_ikke_> Most iso's should still include hardened, but it will be phased out soon (if not already happened) 2018-03-31 16:40:47 <_ikke_> regarding veracrypt, you can try to create a package for yourself 2018-03-31 16:41:17 what about v3.7? and phased out to be replaced with what exactly? 2018-03-31 16:42:12 _ikke_: how do you build one on your onw? 2018-03-31 16:42:31 <_ikke_> m00n_urn: https://lists.alpinelinux.org/alpine-devel/6022.html 2018-03-31 16:46:00 _ikke_: okay 2018-03-31 16:46:03 thanks 2018-03-31 16:48:26 <_ikke_> m00n_urn: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2018-03-31 16:49:43 thank you 2018-03-31 16:51:15 _ikke_: and is selinux installed by default in all alpine images? 2018-03-31 16:56:30 <_ikke_> no 2018-03-31 16:57:55 i can't find a package too 2018-03-31 16:59:40 <_ikke_> libselinux is still in testing 2018-03-31 21:24:53 Hi, how can i install mongodb in version 3.6 i get a dependency error? https://nopaste.xyz/?a8ed294271215e86#kWX/PqKoHM7LBN71PWuISGXxqNlq92mPayY4d2ZwBOU= 2018-03-31 22:50:49 hey isn't selinux available under alpine packages too? 2018-03-31 23:03:35 hello 2018-03-31 23:36:47 hey isn't selinux available under alpine packages too?