[17:10] <cjwatson> So does anyone want to express opinions about Perl 5.14 (see ubuntu-devel/ubuntu-release) or should I just do it?  I'd probably start after the daily CD builds tomorrow if nobody objects.
[17:12]  * ogra_ would just do it, now seems the right time for such a change
[17:14] <cjwatson> I'm at least going to wait until tomorrow anyway, to minimise chances of disrupting daily builds
[17:16] <cjwatson> infinity: if I do this, would you be able to make sure that armhf starts out with perl-base 5.14, so that we don't have to rebuild everything twice?
[17:17]  * cjwatson adds a transition tracker page
[17:28] <infinity> cjwatson: Yup.
[17:29] <infinity> cjwatson: Give me a second ping on the subject when you upload?  I'm not watching -changes like a hawk while I'm busy with Other Things.
[17:29] <Laney> is it much more than no-change rebuilds?
[17:29] <cjwatson> I'd advise also making sure that  liblocale-gettext-perl libtext-iconv-perl libtext-wrapi18n-perl libtext-charwidth-perl  are in sync.
[17:29] <cjwatson> Laney: that's about it; possibly a few merges of packages that have required explicit fixes for Perl 5.14, but most of those have already been taken care of AFAIK
[17:30] <Laney> yeah, seemed well organised in debian
[17:30] <cjwatson> there are eight open blocking bugs in Debian, but nothing particularly core
[17:31] <Laney> some random ftbfs IIRC
[17:31] <cjwatson> yeah, at least a couple were FTBFSing anyway
[17:31] <cjwatson> infinity: will do
[17:33] <infinity> cjwatson: Admit it, when you type "armhf", you now say "arewoof" in your head, don't you?
[17:33]  * infinity owes slangasek some reciprocal pain.
[17:34] <cjwatson> (Scrap libtext-wrapi18n-perl from the above list; it's not ABI-dependent)
[17:34] <cjwatson> infinity: totally
[17:35]  * ogra_ thinks armhf is the wrong naming anyway ... i mean, we never stopped using eabi and also still build little endian ... should theoretically be armelhf
[17:36] <infinity> ogra_: Don't go there. :P
[17:36] <ogra_> *g*
[17:36] <infinity> ogra_: It'll get more confusing when we do arm64, which should be armelhf64, right? :P
[17:36] <ogra_> indeed !
[17:36]  * infinity vomits a little.
[17:36] <ogra_> heh
[17:39] <cjwatson> just use pwgen for your new architecture names
[17:39]  * infinity adds a --arm switch to pwgen.
[17:40] <ogra_> lol
[17:52] <slangasek> one of those perl-5.14 FTBFSes is mine, which I'll sort out just as soon as I get a new version of freetds to the archive
[17:54] <cjwatson> slangasek: which one is that?
[17:54] <slangasek> libdbd-sybase-perl
[17:54] <cjwatson> right
[17:54] <cjwatson> it's at dependency level 3 anyway
[18:13] <jibel> there is a problem with libtasn1-3 and multiarch which break amd64 desktop installation
[18:13] <jibel> bug 890338
[18:13] <ubot4> Launchpad bug 890338 in libtasn1-3 (Debian) (and 2 other projects) "package libtasn1-3 2.10-1 failed to install/upgrade: './usr/share/doc/libtasn1-3/NEWS.gz' is different from the same file on the system (affects: 3) (dups: 2) (heat: 24)" [Unknown,Unknown] https://launchpad.net/bugs/890338
[18:14] <slangasek> jibel: how do you mean, "Breaks amd64 desktop installation"?  There should be no i386 packages pulled into a default desktop
[18:14] <slangasek> also, didn't I see that pitti worked around this by dropping compression of NEWS.Debian in the package?
[18:14] <cjwatson> jibel: we know about the problem in any case
[18:14] <slangasek> "if the user selects to install flash" - right
[18:14] <jibel> slangasek, when flash is installed during the initial installation it install the i386 version of this package
[18:14] <cjwatson> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=647522
[18:14] <ubot4> Debian bug 647522 in gzip "gzip -9n is not deterministic" [Normal,Open]
[18:16] <cjwatson> I will move my debugging of that a bit further up my list (though I'm by no means taking an exclusive lock on this problem)
[18:17] <cjwatson> I've marked jibel's bug as a dup of the one pitti fixed this morning
[18:18] <jibel> cjwatson, found bug 889303. thanks
[18:18] <ubot4> Launchpad bug 889303 in libtasn1-3 (Debian) (and 5 other projects) "gzip -9n sometimes generates a different output file on 64 bit (affects: 19) (dups: 5) (heat: 102)" [Unknown,Unknown] https://launchpad.net/bugs/889303
[18:18] <slangasek> how long ago was the fix published?  I'm not seeing the new libtasn1-3 on us.archive.u.c
[18:19] <cjwatson> source uploaded 10 hours ago
[18:19] <slangasek> perhaps there's a mirroring issue
[18:19] <cjwatson> -rw-r--r-- 1 lp_publish lp_publish 43136 Nov 14 10:20 /home/lp_archive/ubuntu/pool/main/libt/libtasn1-3/libtasn1-3_2.10-1ubuntu1_amd64.deb
[18:19] <cjwatson> -rw-r--r-- 1 lp_publish lp_publish 36982 Nov 14 10:20 /home/lp_archive/ubuntu/pool/main/libt/libtasn1-3/libtasn1-3_2.10-1ubuntu1_armel.deb
[18:19] <cjwatson> -rw-r--r-- 1 lp_publish lp_publish 43668 Nov 14 10:21 /home/lp_archive/ubuntu/pool/main/libt/libtasn1-3/libtasn1-3_2.10-1ubuntu1_i386.deb
[18:19] <cjwatson> -rw-r--r-- 1 lp_publish lp_publish 42498 Nov 14 10:21 /home/lp_archive/ubuntu/pool/main/libt/libtasn1-3/libtasn1-3_2.10-1ubuntu1_powerpc.deb
[18:19] <cjwatson> so eight hours ago
[18:19]  * slangasek nods
[18:19] <cjwatson> (for binaries)
[18:20] <slangasek> I just hit the bug on my daily dist-upgrade here
[18:20] <cjwatson> add this one to the list of install/upgrade-breaking bugs for which the strategy of reversion would be useless
[18:20] <slangasek> where does that list live? :)
[18:20] <cjwatson> in my head
[18:21] <cjwatson> I suspect it will be arbitrarily long and it would be quicker to construct a list of bugs where reversion is the correct strategy
[18:22] <slangasek> jibel: which mirror are you using when you see this?  We should probably track back the mirroring problem
[18:22] <slangasek> oh, or I could go straight to archive.ubuntu.com and see that it's missing there :P
[18:23] <cjwatson> wait, isn't us.archive internal?
[18:23] <slangasek> dunno
[18:23] <cjwatson> as in IP addresses in our datacentres.  (it is.)
[18:24] <jibel> slangasek, fr.archive.ubuntu.com, duplicates use il and at respectively
[18:24]  * slangasek nods
[18:24] <cjwatson> https://launchpad.net/ubuntu/+archivemirrors not looking terribly healthy right now
[18:25] <cjwatson> I wish it were easier to see on that list which ones are country primaries
[18:25] <slangasek> I'm told +archivemirrors is fundamentally broken
[18:30] <cjwatson> anyway, IS discussion indicates syncproxy is busted and undergoing maintenance
[18:31] <cjwatson> so we can still do builds but can't effectively publish anything new to users right now
[18:31] <cjwatson> (which is obviously bad but out of our direct control ...)
[19:34] <infinity> cjwatson: It's come to my attention that we've had no armel builds of anything since the 12th.  Perhaps relating to the machine cutover?
[19:35] <infinity> cjwatson: (Don't really have the time to look right now, but I can tomorrow if you don't beat me to it)
[19:35] <infinity> cjwatson: (image builds, that is)
[19:43] <infinity> cjwatson: Err, nevermind.  "nothing building" was a log issue.
[19:43] <infinity> cjwatson: It's just core (I think) that isn't.
[19:45] <infinity> cjwatson: NM.  Dead buildd.
[20:26] <cjwatson> Yeah.  I'd noticed that and was going to investigate tomorrow if it persisted.  There was apparently some power work in the DC over the weekend (or at least that was wgrant's guess) and I thought it could be related.
[20:26] <cjwatson> Do you know what's wrong and if we should be switching buildds or something?
[21:30] <wgrant> cjwatson, infinity: It could also just be buildd-manager ignoring them.
[21:45] <cjwatson> wgrant: buildd-manager doesn't do image builds.
[21:51] <wgrant> Oh. Image builds.
[21:51] <wgrant> I see.
[21:56] <cjwatson> sorry, I did bring up your name - the livefs buildds are on the same segment as the LP ones AFAIK
[21:56] <cjwatson> so I figured they might be affected by similar power work
[21:56] <cjwatson> (if any)
[22:16] <infinity> cjwatson: apache was dead on annonaecaeaeaeaeeea.  lamont's rescued it.
[22:16] <infinity> cjwatson: No idea why, and not a whole lot of care factor.  That host's days are numbered.
[23:44] <GrueMaster> infinity: Not sure who to ping, but the recently tested and released omap4 kernel for oneiric is no longer showing up in http://ports.ubuntu.com/pool/main/l/linux-ti-omap4/ .  Source is there, meta is there.
[23:45] <GrueMaster> bug 885466 shows it as fix-released.  And it was there last week.
[23:45] <ubot4> Launchpad bug 885466 in linux-ti-omap4 (Ubuntu Oneiric) (and 12 other projects) "linux-ti-omap4: 3.0.0-1206.12 -proposed tracker (affects: 2) (dups: 1) (heat: 20)" [Undecided,Fix released] https://launchpad.net/bugs/885466