[00:03] <slangasek> doko__: hrm, why does this eglibc diff appear to include a patch to *add* memcmp sse4 (a7f4cf5f1a6c0c9011c8dc0339cbe860fd272b6c)?
[00:04] <slangasek> doko__: that doesn't look like a bugfix to me
[00:33] <slangasek> hum, why has OOo/armel only been building for 10h?  I thought it had already started building before today, and gotten farther than that
[01:06] <ScottK> slangasek: It did and then today got restarted on a different buildd.  I don't know why.
[01:06] <slangasek> hmm
[03:21] <slangasek> if someone could bump the build prio on libmsn/ia64, that would help with getting NBS fully cleaned out (lets us avoid kopete being rebuilt against the old lib)
[03:22] <slangasek> actually
[03:22] <slangasek> processing the NBS and leaving libmsn-dev uninstallable on ia64 also does that
[04:58] <ScottK> slangasek: Should I be worrying about pinging about things like libperl-critic-perl MIR not done so other stuff won't build in Main?
[05:14] <ScottK> BTW, I'm looking at lucas rebuild results so the stuff you're seeing in the queue or in sync requests from me is for that
[05:15] <slangasek> ScottK: if you have the spare cycles to ping about those missing MIRs, yes please
[05:15] <slangasek> so
[05:15] <ScottK> OK.
[05:16] <slangasek> if you use irssi with a properly-configured, SSL-secured proxy
[05:16] <slangasek> don't upgrade to the security update
[05:17]  * slangasek files bug #565182
[05:17] <ubottu> Launchpad bug 565182 in irssi "security regression: SSL CN check breaks IRC proxy" [Undecided,New] https://launchpad.net/bugs/565182
[05:18] <slangasek> jdstrand, kees, mdeslaur: I've subscribed ubuntu-security to the above; anything else I should do with it?  assign it to anyone in particular?
[05:19] <slangasek> (patch will be sent shortly)
[05:23] <slangasek> ScottK: portmap and irssi in the queue are both mine; care to review those?
[05:24] <ScottK> No guarantees I'll feel qualified to decide, but I'll look.
[05:25] <mdeslaur> slangasek: thanks, I'll assign it to jdstrand since he published the update, he can reassign it as necessary
[05:25] <slangasek> mdeslaur: ok - fwiw, you can read what I'm typing, so I can vouch for the correctness of the fix ;)
[05:26] <mdeslaur> slangasek: hehe :)
[05:27] <ScottK> portmap OK.  Looking at irssi.
[05:39] <ScottK> slangasek: You may wonder why we marked Bug #565180 as RC (high and milestoned) - Kubuntu has had a very poor reputation for translations in the last few years due to this kind of thing.  The distro can't afford to leave it unaddressed.
[05:39] <ubottu> Launchpad bug 565180 in language-pack-kde-de "Translation error in Launchpad changes (KMail/kdepim)" [High,Triaged] https://launchpad.net/bugs/565180
[05:43] <ScottK> slangasek: Why does define IRSSI_VERSION_TIME 1938 change in your irssi upload?
[06:17] <slangasek> ScottK: it changed as a side-effect of doing a package build in the working tree
[06:17] <ScottK> slangasek: RC bug cound it up by four for missing MIRs.
[06:17] <ScottK> slangasek: OK.  Do we care?
[06:18] <slangasek> about that change as a side effect?  I don't
[06:18] <slangasek> if that's what you get as output of ./debian/rules build && ./debian/rules clean, IMHO that's what should get uploaded
[06:19] <slangasek> missing MIRs> good, now they're documented
[06:20] <slangasek> 565180> we should have time to fix that before the final langpack export
[06:20] <ScottK> OK.  Fair enough.
[06:20] <ScottK> Done
[06:20] <slangasek> thanks!
[06:26] <slangasek> 565180 is a little light on detail, why is the new string considered an error?  (And where does the upstream translation live?)
[06:26] <ScottK> slangasek: Upstream translation lives in kdesvn.
[06:27] <ScottK> slangasek: The agreement we had with dpm was that upstream strings wouldn't get changed.
[06:27] <ScottK> (the alternative was we would forgo Ubuntu translations entirely and just use upstream's)
[06:29] <ScottK> BTW, I just did my first "those are all cosmetic changes to the package, too late" rejection.
[06:30] <slangasek> in kdesvn?
[06:31] <slangasek> not in kdepim?
[06:31] <ScottK> Sorry, KDE svn
[06:31] <slangasek> didn't realize there was special handling for KDE upstream strings, ok
[06:31] <slangasek> hum - in that case, how is it supposed to get into the distro?
[06:32] <slangasek> does rosetta do a direct import from KDE svn?
[06:32] <ScottK> We upload the KDE language packs and they get stripped
[06:33] <slangasek> aha
[06:33] <ScottK> We know from providing PPA uploads of new versions to users with translations that providing the KDE languauge packs directly works just fine.
[06:33] <ScottK> We'd just lose the 'benifits' of Rosetta.
[06:33] <slangasek> ScottK: the translation in the kde-l10n-de source package matches the one shown at https://translations.edge.launchpad.net/ubuntu/lucid/+source/kdepim/+pots/kmail/de/991/+translate
[06:34] <ScottK> Hmmm.
[06:35] <ScottK> Maybe it got changed upstream after our last update from there.
[06:35] <ScottK> I'll see if I can get someone to look into it.
[06:35] <ScottK> For the last two cycles we've had "stop using Rosetta so our translations get better" specs.
[06:35] <slangasek> the string in language-pack-kde-de-base also matches
[06:35] <ScottK> OK.
[06:36] <ScottK> Let me put that in the bug and see what we can find out.
[06:38] <slangasek> oh - echidnaman linked to the wrong string
[06:38]  * slangasek just looked at the upstream bug
[06:38] <ScottK> Ah
[06:43] <ScottK> https://translations.edge.launchpad.net/ubuntu/lucid/+source/kdepim/+pots/kmail/de/1448/+translate
[06:44] <slangasek> yep
[06:45] <ScottK> This will be sufficient to cause the abandon Rosetta spec to get resurrected for the next UDS.
[06:45] <ScottK> It's really all pain and no visible gain.
[06:46] <ScottK> KDE already strips translations out by language so upstream has already done the most important part for us.
[06:48] <slangasek> well, if the Kubuntu community's position is to not use rosetta for KDE translations in favor of using the upstream infrastructure/community, and upstream translations are already split into packages by language, it doesn't really seem to make sense to have KDE language packs
[06:50] <ScottK> slangasek: I agree.
[06:51] <ScottK> We got a lot of pushback.
[06:51] <ScottK> I guess we'll decide again at UDS.
[06:52]  * wgrant sighs.
[06:54] <slangasek> perhaps the right compromise here is to continue importing strings and translations into rosetta, but to not strip them from kde-l10n-* at build-time, and not push them into langpacks?
[06:54] <ScottK> Perhaps.
[06:55] <slangasek> translation teams could still use rosetta for collaboration, but committing translations would have to be funneled via upstream
[06:55] <ScottK> Part of the problem, as I understand it, is that KDE has some very specific standards for translations to get continuity across the workspace and Gnome doesn't follow them.
[06:56] <ScottK> Since most Ubuntu translators have a Gnome background, they often do things that are wrong from a KDE perspective.
[06:56] <ScottK> The times exporting stuff from Rosetta has been tried with KDE, the translations have been rejected due to low quality
[06:56] <slangasek> ah.
[06:57] <ScottK> Doesn't mean it can't work, but the history isn't good.
[06:59] <ScottK> Given some of the historical sensitivities in this area the tolerance for error from Rosetta/Ubuntu translators is nil both in Kubuntu and KDE.
[07:05] <ScottK> slangasek: Time for me to crash.  Good night and good luck.
[07:08] <slangasek> ScottK: ok, 'night!
[15:10] <hyperair> Are bugfix-only uploads allowed or do I need to file an FFe?
[16:15] <Laney> allowed
[18:50]  * ScottK is done clearing the queue.
[18:50] <ScottK> The rest I'll leave for someone else.
[19:18] <jdstrand> slangasek: re irssi> thanks, I'll take care of security releases. it seems you wrote this patch yourself-- did you send it upstream? (I don't see it in the upstream svn)
[19:35] <jdstrand> slangasek: if not, I can get the patch to the right people
[20:38] <ScottK> slangasek: Did you get a chance to get some traction on the Python open.codecs question?  If we can get some guidance on evaluating/fixing, I can try to mobilize some people.
[21:31] <slangasek> jdstrand: nope, hadn't upstreamed it yet, appreciated if you would
[21:32] <slangasek> ScottK: not yet, apparently my calendar was off and yesterday wasn't actually a 36-hour day like I had planned
[21:32] <ScottK> Ah, no problem.
[21:33] <slangasek> I'll hopefully get to it today, but have to tend the yard a bit first before I leave it to its own devices for 4 weeks
[21:35] <slangasek> pitti: I'm not counting -en either! :)
[21:35] <slangasek> pitti: es xh pt de bn fr hi
[22:13] <ScottK> slangasek: Are you updating all the *-meta packages or was Xubuntu special?
[22:14] <slangasek> ScottK: all of them; netbook, kubuntu already uploaded
[22:14] <ScottK> OK.  I just accepted xubuntu
[22:14] <slangasek> ta
[22:14] <ScottK> Don't see the others yet.
[22:15] <slangasek> unfortunately we still have one more font seed change pending, so there'll be another round of uploads yet; but these uploads seem to be required in order to let livecds build, so
[22:16] <ScottK> OK.  I'll push them through as I see them....
[22:16] <ScottK> There they are
[22:18] <ScottK> Done.
[22:20] <slangasek> ScottK: I think mythbuntu / ubuntustudio will also still need uploads; still running the ./update
[22:20] <ScottK> OK.
[22:21] <ScottK> I guess it's a good thing I went ahead and New'ed the split Khmer font pacakge.
[22:22] <slangasek> the final upload is still waiting for ttf-takao-gothic, I think (or at least, a decision on same)
[22:25]  * ScottK nods
[22:26] <nhandler> Do you think we could squeeze in a ftbfs-fix for libgtk2-perl ?
[22:27] <slangasek> there is generally room for ftbfs fixes, yes
[22:33] <nhandler> slangasek: Will a sync request get processed? Or should we get the minimal patch applied in Ubuntu and uploaded, and then sync with Debian for 10.10 ?
[22:35] <Laney> if it won't get processed by the AAs, you can upload it with the syncpackge script
[22:35] <ScottK> nhandler: At this point a minimal fix for the FTBFS would be preferred.
[22:36] <ScottK> nhandler: I did a number of these yesterday and uploaded with a build1 revision so the sync in Maverick is automatic.
[22:36] <nhandler> ScottK: Have time to sponsor the patch? It is in main.
[22:37] <ScottK> nhandler: Yes.
[22:37] <nhandler> ScottK: Check out Bug #564070
[22:37] <ubottu> Launchpad bug 564070 in libgtk2-perl "Please sync libgtk2-perl from Debian (fixes FTBFS)" [Wishlist,In progress] https://launchpad.net/bugs/564070
[22:37] <ScottK> nhandler: -->#ubuntu-devel
[22:37] <nhandler> k
[22:38] <jdstrand> slangasek: done
[22:45] <slangasek> jdstrand: thanks!
[22:48] <slangasek> ScottK: {mythbuntu,ubuntustudio}-meta don't need updated
[22:48] <slangasek> (apparently these don't derive from platform/desktop-common ?)
[22:48] <ScottK> OK, well that simplifies things.
[22:48] <ScottK> Weird.
[22:49] <ScottK> That may be a larger bug that needs investigation.
[23:17] <doko> slangasek: re-uploaded eglibc, the strcmp/memcpy fixes now confirmed, disabling the x86_64 memcpy in the series file
[23:17] <ScottK> doko: Want me to reject the old one?
[23:18] <doko> don't mind
[23:18] <slangasek> doko: great, thanks - will pick that up shortly
[23:18] <ScottK> Done
[23:21] <slangasek> and accepted
[23:56] <slangasek> ScottK, pitti: prepromoted the MIRs needed for getting packages built and up-to-date
[23:57] <ScottK> slangasek: Sounds good.  I'll give them a manual retry after the next publisher run finishes since AFAIK automatic is still broken.
[23:58] <slangasek> right