[07:50] <seb128> acheronuk, hey, could you remove the libnm-util-dev Build-Depends from kdelibs4support? it was added in 5.47.0-0ubuntu1 and isn't in Debian, it's probably not needed but it's not clear to me why it was added in the first place. That lib is deprecated and NBS now in Eoan
[07:51] <acheronuk> seb128: let me look
[07:52] <seb128> acheronuk, thx!
[07:53] <LocutusOfBorg> infinity, first, I know how to develop for devel release, but sometimes for just my packages that I source upload in Debian, and have no strange things in clean target, I do dpkg-buildpackage -S -d and dput, and meh, I'm not an Ubuntu developer in my daily job, so I have to stick with at least one LTS, because I'm not a student who can upgrade because meh new stuff. Moreover, I can figure out, and I already use chroots, but
[07:53] <LocutusOfBorg> sometimes its just an overkill when you know the stuff you are working on, so I prefer to have a never debhelper on my bionic machine, e.g. if I want to compile the latest vbox in my laptop, I have to 1) lower compat requirements or 2) use a custom built debhelper version. Right now I'm following 2, and I'm not the only one who does this. if you want to provide updates to an LTS for 10 years, I think a backport is worth the
[07:53] <LocutusOfBorg> effort.
[07:54] <amurray> seb128: do you expect evolution-data-server currently in cosmic-proposed from LP #1813268 to migrate soon?
[07:55] <seb128> amurray, hey, not unless someone test/mark the bugs verification-done, at this point I expect more that it's not going to be verified and sit there or end up being delete from proposed
[07:55] <seb128> amurray, which is sort why I made that point on the list a while ago about doing SRUs on non LTS/non-current-stable is a waste of efforts
[07:55] <LocutusOfBorg> juliank, 1) I use already chroots, but sometimes I just need to create something quick for test or ppa, the chroot solution doesn't scale efficiently on 100% of the use cases 2) upgrade to devel is a no-go for $dailyjob 3) what if I want to also test the built deb file on my laptop?
[07:56] <seb128> amurray, in practice no-one use cosmic now and we don't have people to do the verifications
[07:57] <LocutusOfBorg> and using -nc is source of troubles, I don't want the risk of something in the tarball I don't want, I prefer to not use nc and chroot when clean doesn't work (or backport tools when I find them useful in my daily life)
[07:57] <juliank> I mean, you are supposed to use -nc anyway, to avoid spurious changes from a broken clean target
[07:58] <juliank> Maybe not for devel, but certainly for SRUs
[07:58] <seb128> amurray, anyway, sorry for the sort of ranting ... did you need that update in for a specific reason? if so I guess we can find some time to do the verification and unblock you (I need to reinstall a cosmic system for that though, since I don't have one at this point)
[08:18] <acheronuk> seb128: this is why: -- The following OPTIONAL packages have not been found:
[08:18] <acheronuk>  * NetworkManager (required version >= 0.7.0), The NetworkManager headers, <http://projects.gnome.org/NetworkManager>
[08:18] <acheronuk>    Needed for kded's networkstatus module
[08:19] <acheronuk> seb128: I assume the files that were in libnm-util-dev will move to the normal -dev package?
[08:20] <seb128> acheronuk, the non deprecated ones yes
[08:21] <seb128> acheronuk, what .pc does it try to use?
[08:35] <acheronuk> seb128: not sure. its optional though, and the Qt5/KF5 kded doesn't use kdelibs4support, so I think safe to drop
[08:35] <seb128> good
[08:40] <acheronuk> seb128: uploading...
[08:40] <seb128> acheronuk, thanks!
[08:52] <LocutusOfBorg> juliank, I usually debdiff changes files to be sure about my changes :) I prefer to do that anyway, and in case cruft is there, I create a custom chroot and run clean there :)
[08:53] <LocutusOfBorg> I admit, I don't SRU that much!
[08:53] <LocutusOfBorg> in any case, a custom chroot is also useful to verify the fix, so since I have to do it anyway... better use it from the begin
[08:58] <juliank> I have not used chroots in ages
[09:01] <LocutusOfBorg> I find them really useful!
[09:02] <LocutusOfBorg> with xhost + I can easily test even graphic applications...
[09:02] <LocutusOfBorg> just for curiosity, what do you use to setup an old Ubuntu version to confirm if an SRU is good or not? docker?
[10:49] <cpaelzer> hi, I have a package which runs build time tests that need a $HOME to be sort of valid (existing dir)
[10:49] <cpaelzer> but in the build env this is set to /sbuild-nonexistent/
[10:49] <cpaelzer> is there a common best practise to that?
[10:49] <cpaelzer> I currently think of mktemp + export $HOME in override_dh_auto_test
[10:50] <rbasak> I think that sounds reasonable
[10:50] <amurray> seb128: re e-d-s on cosmic - the version in -proposed happens to already fix CVE-2018-15587 - so if that gets out then I don't have to manually patch it
[10:51] <rbasak> cpaelzer: or something in the build tree, and then you can have an easier time writing the clean target?
[10:51] <amurray> seb128: also from what I could see that is marked verification-done in the bug report
[11:23] <seb128> amurray, bug #1815882 isn't verified, I will test that now
[11:24] <amurray> seb128: I was looking at bug #1813268
[11:26] <seb128> amurray, right, that one is verified :)
[12:52] <Odd_Bloke> LocutusOfBorg: I use lxd for testing on old Ubuntu versions; a full system container means less Docker weirdness to work around.
[13:57] <seb128> cyphermox, hey, on that new japanese era SRU bug, are you/someone from foundations looking at the desktopish packages on the list as well (poppler-data, gucharmap, gnome-characters)
[14:03] <seb128> wasn't launchpad letting you see full comments on the web without downloading them before?
[14:06] <cjwatson> Depends on the length
[14:09] <seb128> well, it has always been cutting to some length with "..."
[14:09] <seb128> but I though you had a "see full comment" or similar before
[14:09] <seb128> which would open that comment in its own page
[14:12] <cjwatson> You do
[14:12] <cjwatson> Example?
[14:12] <cjwatson> Nothing has changed there
[14:15] <seb128> cjwatson, https://bugs.launchpad.net/ubuntu/+source/xdg-dbus-proxy/+bug/1811824/comments/3
[14:15] <seb128> or maybe firefox changed, now it pop the download dialog, I'm pretty sure I was able to open those "inline" before
[14:15] <cjwatson> I think that must be a browser change
[14:16] <cjwatson> That said, we're sending Content-Disposition: attachment; ...
[14:16] <seb128> chromium does the same
[14:18] <cjwatson> Anyway, the last change to that code was nearly a year ago and that was just to quote the filename properly in the response header; before that the last change was in 2012
[14:20] <seb128> maybe my memory is wrong :)
[14:20] <seb128> thx for checking/replying cjwatson!
[14:27] <seb128> (well in fact the item to download the comment always existed, but I though there was also a link to view the full comment in a page... anyway, not important)
[14:30] <seb128> k, it's still fine on other bugs, e.g https://bugs.launchpad.net/ubuntu/+source/usbguard/+bug/1816548/comments/10 so maybe it has to do with the size after all
[14:33] <cjwatson> lib/lp/bugs/browser/bugcomment.py:        """View redirects to +download if comment is too long to render."""
[14:34] <seb128> cjwatson, that explains, thx again and sorry for the noise (and I learnt something new about launchpad today ;)
[14:34] <cjwatson> Threshold is currently 3200 bytes if I'm reading the configs right
[14:34] <cjwatson> Err characters not bytes
[14:35] <seb128> that feels a bit low, but there is probably a reason for it
[14:35] <cjwatson> I'm guessing performance but it was before my time
[18:06] <bdmurray> connor_k: bug 1828615 doesn't need fixing for the development release right?
[18:07] <connor_k> bdmurray, no, those changes should already be there for development since they were fixed for disco
[18:07] <bdmurray> connor_k: okay, I'm going to set the task to Fix Released then and a new bionic task will be added
[18:08] <connor_k> bdmurray, thank you :)
[18:12] <bdmurray> LocutusOfBorg: bug 1535045 is missing a test case and "regression potential" is supposed to be about how a regression could / would manifest, not what the chances of a regression are. Additionally, please add information to the bug about the status for disco and cosmic.
[19:59] <seb128> bdmurray, hey, could you review epiphany-browser/bionic? That's a standard GNOME bugfix update which we didn't pick for a while and got upstream angry at us, I would like to be able to get back to them with "update is in proposed ready for testing" :)
[20:12] <bdmurray> seb128: sure
[20:15] <seb128> bdmurray, thanks!
[21:41] <ryuguns> Hey, is the Ubuntu SDK still worked on? It's just a Qt Creator plugin isn't it? And is there any benefit to using it other than working on the (AFAIK discontinued) Ubuntu Touch platform?
[21:44] <sarnold> ryuguns: if you like ubuntu touch I understand there's still a community working on it https://ubports.com/
[21:45] <ryuguns> sarnold: Wow, that's awesome, do they have any official relationship with Canonical?
[21:46] <ryuguns> Just checked, looks like they don't, but that's still cool
[21:46] <ryuguns> Thanks a lot