| Unit193 | schopin: Heya, I presume you saw your email? | 09:08 |
|---|---|---|
| schopin | Unit193: I hadn't so far, but now I did! Thanks for the NMU sponsoring :) | 09:12 |
| Unit193 | Happy to have that done. | 09:13 |
| schopin | I don't suppose there's an explicit package->lowNMU mapping available somewhere, is there? | 09:15 |
| Unit193 | 1. https://wiki.debian.org/LowThresholdNmu 2. https://tracker.debian.org/pkg/irssi-plugin-xmpp top left | 09:17 |
| schopin | I knew about the first page, but I'd never noticed the notice on the tracker, thanks! | 09:20 |
| slyon | seb128: would you mind if I push my NetworkManager-netplan integration work to https://git.launchpad.net/network-manager directly? (as a separate "ubuntu/jammy+netplan" branch, of course) – I think that's the central place where everybody could easily find (and modify) it. | 09:42 |
| slyon | It currently sits at: https://git.launchpad.net/~slyon/network-manager/log/?h=ubuntu/jammy%2bnetplan | 09:42 |
| seb128 | slyon, hey, feel free to push there, it's only a branch | 10:11 |
| slyon | thanks, will do! | 10:11 |
| === guiverc2 is now known as guiverc | ||
| rbasak | In bug 1979933 an SRU upload patches the configure script directly instead of using dh-autoreconf. Is this OK? There's future risk with maintaining such a patch in the face of eg. a future security update, but there's also regression risk in introducing autoreconf in a stable release when it wasn't there previously, as the version of autoconf being used might differ from upstream. | 11:56 |
| ubottu | Bug 1979933 in net-snmp (Ubuntu Focal) "snmptrapd Segmentation Faults When Calling my_load_defaults()" [Medium, Incomplete] https://launchpad.net/bugs/1979933 | 11:56 |
| rbasak | Opinions? | 11:56 |
| mdeslaur | rbasak: personally, I think it's ok if it also modifies the source files, so that if autoreconf is added in the future nothing will get lost...as you say, there's a risk both ways, not sure if there's a recommended way | 12:35 |
| mdeslaur | or rather, an ideal way | 12:35 |
| slyon | rbasak: o/ thanks for rebasing that protobuf-c SRU! | 12:36 |
| mdeslaur | rbasak: I assume the change to configure.d/config_os_libs2 would regenerate the configure changes, no? | 12:39 |
| rbasak | mdeslaur: you're right - I just simultaneously noticed and commented in the bug | 12:41 |
| mdeslaur | cool | 12:42 |
| rbasak | Accepted into proposed - thank you for the discussion. | 12:42 |
| bluca | slyon: any answer from rel team on repartd? | 12:50 |
| slyon | bluca: not yet (I guess most a busy with spinning the 22.04.1 release). If I do not get any reply until after the .1 release, I will escalate that email to the ubuntu-devel/ubuntu-release mailinglist (and put you in CC) | 12:55 |
| bluca | thank you | 12:55 |
| slyon | yw | 12:55 |
| bluca | on a somewhat-related topic: once you merged the latest changes from Salsa, and with a couple additional changes in d/rules which would be fine to have, it becomes possible to do i386 builds that produce only the two public libraries and the EFI binaries | 12:56 |
| bluca | with a reduced set of dependencies among other things | 12:57 |
| bluca | is that something you'd be interested in for ubuntu? | 12:57 |
| slyon | bluca: IIRC we added some EFI binary quirks in our most recent build, so yes, that sounds interesting. @enr0n might know more ^ | 13:15 |
| bluca | the main advantage is to reduce the i386 build set, instead of the full set of binary packages, only libs + efi | 13:17 |
| enr0n | slyon: the EFI build quirks were just for new linker warnings in binutils | 13:17 |
| bluca | IIRC your goal is to have libs-only i386? | 13:17 |
| bluca | (reduce here is in the context of src:systemd) | 13:18 |
| slyon | bluca: yes, indeed. the systemd binary is even uninstallable in recent Ubuntu releases on i386 (since jammy, according to https://autopkgtest.ubuntu.com/packages/systemd), so getting rid of it would be nice | 13:22 |
| bluca | ok, I'll get the missing bits in d/rules added and CC you to the MR | 13:23 |
| bluca | you'll need to keep the Architecture: list diff in your downstream tree, but that shouldn't be too much burden | 13:25 |
| slyon | cool thanks! | 13:33 |
| === tumbleweed_ is now known as tumbleweed | ||
| jawn-smith | vorlon: There's a package called hedgewars in the NBS report that depends on a package that hasn't existed since bionic. It has no reverse dependencies and is just a video game so not critical. Feel like breaking out the old archive chainsaw? | 18:26 |
| jawn-smith | oh hold up on that, I'm surprised to see it somehow built in jammy and would like to understand how it did so | 18:28 |
| vorlon | jawn-smith: which package do you say hasn't existed since bionic? | 18:33 |
| jawn-smith | I was looking at haskell-gnuidn | 18:35 |
| jawn-smith | But I think that was incorrect as it isn't actually a dependency of hedgewars | 18:36 |
| LocutusOfBorg | jawn-smith, what? | 19:37 |
| LocutusOfBorg | hedgewars does not depend on gnuidn | 19:37 |
| LocutusOfBorg | oh already answered nice | 19:38 |
| jawn-smith | Yeah I mistyped some commands to apt-cache so that was an incorrect statement | 19:39 |
| LocutusOfBorg | for sure hedgewars is a blocker for haskell and ffmpeg | 19:46 |
| LocutusOfBorg | I'm working on it | 19:46 |
| jawn-smith | Okay great, I'll focus my efforst elsewhere | 20:09 |
| === ben_ is now known as ben | ||
| === nuccitheboss1 is now known as nuccitheboss | ||
| Unit193 | mapreri: Thanks for the -backport ACKs, they're in the queue now. \o/ | 22:12 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!