[00:28] https://www.independent.co.uk/news/world/europe/brexit-france-theresa-may-deal-nathalie-loiseau-cat-a8828026.html [00:28] France’s EU minister names her cat ‘Brexit’ because ‘he meows loudly to be let out but won’t go through the door’ [00:48] . [04:46] whoops, this is not https://wiki.pumpingstationone.org/IRC :p [08:16] Does anyone know how this images is created? [08:16] http://cdimage.ubuntu.com/releases/18.04.2/release/ubuntu-18.04.2-server-arm64.iso [08:17] There is lots of documentation regarding how to build a customised Ubuntu ISO for x86_64 (using ISOLINUX), but none for ARM64 [08:17] Does anyone know of a working `mkisofs` command that works for the ISO image above? [08:18] .. or where the scripts that create this image reside? [08:18] lag: the code is all in debian-cd somewhere [08:18] mwhudson: Is that a package? [08:19] no, it's a branch [08:19] this one https://code.launchpad.net/~ubuntu-cdimage/debian-cd/ubuntu [08:19] arm boxes have cds?! :) [08:19] alkisg: Sure, why wouldn't they? [08:20] alkisg: Also flashing these images onto USB flash drives and SD cards work [08:20] I've only seen raspberries and the like, so no idea, I didn't imagine they'd have [08:20] fair warning, this code is not really designed for third party use [08:20] sadly [08:20] alkisg: x86_64 boxes have cds !? :) [08:20] none of mine do... [08:20] For usb/sd cards, I thought the images were .img [08:21] I.e. no need for hybrid .isos with workarounds etc [08:22] BIOS/UEFI do have code to boot from El-torito cds, I didn't know arm firmware has that code too [08:28] alkisg: You have led a sheltered life :D [08:28] :D [08:28] alkisg: There is more to (ARM) life than RPi [08:29] * alkisg would like to get sheltered from rpi as well, but has to support it in a couple of projects... [08:30] alkisg: I'm not saying it a bad platform - just that there are so many more (better) platforms out there [08:32] +1. Or maybe +1000. :D === M_hc[m] is now known as _hc === Wryhder is now known as Lucas_Gray [10:03] Does anyone know a bit about ISO files? [10:04] alkisg: You seemed pretty knowledgable in that area (/me has lived a sheltered (ISO) life) [10:04] I have 2 files - 1 made by Canonical, 1 by me: [10:04] https://paste.ubuntu.com/p/btdkvtnJDs/ [10:04] They look the same (should be ISO9660), but when they're flashed, the partition tables look totally different [10:05] Canonical: https://paste.ubuntu.com/p/btdkvtnJDs/ [10:05] Whoops - ignore that [10:05] Canonical: https://i.imgur.com/MldCw7i.png [10:05] Mine: https://i.imgur.com/uIbyFAJ.png [10:05] What gives? [10:07] CMD: xorriso -as mkisofs -r -checksum_algorithm_iso md5,sha1 -V "Bionic: AArch64 Laptops" -o laptops-ubuntu-bionic.iso -J -l -c boot/boot.cat -partition_offset 16 -append_partition 2 0xef CDDIR/boot/grub/efi.img -e boot/grub/efi.img -no-emul-boot CDDIR [10:14] Is there a good answer to https://lists.ubuntu.com/archives/ubuntu-server/2019-March/007820.html ? [10:14] (Ctrl-C during release upgrade) [10:14] I would apt -f install and fix anything up manually, or better redeploy to a new release and not do a release upgrade in the first place. [10:15] I'm not sure if there's anything better that could be recommended. === ricab is now known as ricab|brb === ricab|brb is now known as ricab|bbiab [10:25] lag: sorry, I haven't studied the hybrid .iso internals recently, I don't know that much. I did find some related pages on the internet about it in the past, explaining both ubuntu's and fedora's methods, but it was so long ago that I don't remember anything to help you. [10:28] rbasak: yeah if dpkg --configure -a and apt install -f and friends don't fix it, either you're in for a lot of fiddling or should just start again i think [10:29] mwhudson: I'm wondering: is that acceptable, or a bug that we don't have a better option? [10:29] rbasak: we should probably block C-c i guess... [10:30] i don't know, i guess apt install -f should really fix things and any way it doesn't would pedantically be a packaging bug? [10:32] Maybe there's also "power loss recovery" involved, not just ctrl+c... [10:33] In theory all maintainer scripts are idempotent and so apt install -f should always be able to fix things. [10:33] But in practice do-release-upgrade exists instead of apt; doesn't that fix things up after apt is finished? [10:33] So perhaps do-release-upgrade needs a "resume" entry point. [11:01] cjwatson: thanks for merging support for the new "Commands-*" indexfiles to LP. I did a quick test on disco and command-not-found there seems to be working fine and using the data from the archive.ubuntu.com indexfiles now instead of the old command-not-found-data package. so yay! [11:03] Excellent [11:10] whoop whoop [11:15] awesome! === Wryhder is now known as Lucas_Gray [13:37] juliank, found the bug why u-u autopkgtest fails in disco when upgrading cosmic [13:38] :) [13:38] i'm rather :-\, will push the fix after lunch === ricab is now known as ricab|lunch [14:01] doko: things should be fine now that nvidia is finally in restricted (rather than in multiverse) [14:06] tseliot, without transitionals for i386 I don't see this properly working [14:07] and libnvidia-common-410 should be arch all [14:08] ricotz: let me check [14:13] ricotz: this should do it: http://paste.ubuntu.com/p/mbyPR5djWn/ [14:17] tseliot, this looks good [14:19] good [14:48] Is command-not-found information now on Ubuntu mirrors? === ricab|lunch is now known as ricab [15:23] seb128: does gnome-initial-setup still ask about automatic crash reporting? [15:23] bdmurray, it never did on Ubuntu I think [15:24] bdmurray, I don't think it knows how to talk to apport? [15:24] seb128: I thought didier did something about that a while ago [15:24] gnome-control-center yes [15:24] initial setup no [15:25] Ah, okay [17:26] seb128, juliank: I pushed the converted repo, it matches usd imports (ignoring .bzr-builddeb) https://code.launchpad.net/~ubuntu-core-dev/software-properties/+git/software-properties [17:27] tomorrow we can switch the default repo on lp and i push a pointer on top of the bzr repos [17:27] and import the few dscs missing from trusty/precise [17:28] rbalint: Awesome! [17:29] rbalint, great, thx [17:30] then I'll rebase the PackageKit patch queue from Debian on top of the matching Ubuntu tag [17:30] and then I'll rebase again on top of ubuntu/master [17:30] It [17:30] It will be fun [17:36] juliank, cool! [17:36] juliank, i see there used to be a git repository on alioth but it is not migrated [17:37] (to salsa) [17:37] was there anything useful or i should not care about it? [17:39] juliank, also are you ok with the git repo scheme? Debian uses quilt [17:39] rbalint: I have a git repo on salsa at https://salsa.debian.org/pkgutopia-team/software-properties, but it imports the ubuntu tarballs [17:39] rbalint: Yeah, it's fine. What I do in Debian is import the ubuntu .tar.xz and as an .orig.tar.xz, and then use 3.0 (quilt), so I can ship the same tarball [17:40] I'll probably merge in the Ubuntu git branches into the Debian branches soon, though [17:40] juliank, ok, ack [17:40] instead of continuing to import tarballs :) [17:41] juliank, i'm transitioning to this upstream/latest scheme, too, with my packages [18:54] bdmurray: hello, would you be able to take a look at keystone in the cosmic unapproved queue? this version would replace the current version that is in cosmic-proposed (2:14.0.1-0ubuntu2). [21:32] LocutusOfBorg: hi, I have three patches for llvm/clang, needed by intel's "NEO" opencl compute runtime. I'll test the build tomorrow and would like to get them in disco [21:33] LocutusOfBorg: all are in llvm master, but not on llvm-8 [23:05] bdmurray: i added the additional test details to bug 1815126 [23:05] bug 1815126 in python-networkx (Ubuntu Bionic) "[SRU] Bionic missing dependency for pydotplus" [Medium,Incomplete] https://launchpad.net/bugs/1815126 [23:07] coreycb: cool, thanks [23:07] bdmurray: np thank you [23:21] bdmurray: thanks for accepting parted to proposed for 1820090 :)