[00:08] <cjwatson> doko: meh.  I'd sort of like to be in sync
[00:08] <cjwatson> otherwise we'll waste time diagnosing weird stuff
[00:08] <cjwatson> bleh, who NEWed linux?  it's all gone to universe
[00:09] <cjwatson> use kernel-overrides
[00:10]  * cjwatson fixes it up with security-kernel-overrides
[00:11] <cjwatson> or I would if LP hadn't apparently exploded
[00:13] <cjwatson> ... downtime, I guess
[00:14] <apw> cjwatson, yeah was some downtime
[00:14] <cjwatson> 2200-0000 UTC, so that would be now
[00:14] <apw> yeah ... bah
[00:15] <cjwatson> any luck with the problem at hand?
[00:16] <apw> i think i've sorted it out, am trying get a test build started to confirm
[00:18] <cjwatson> I'd like to try test-building something on !powerpc to make sure that the new linux-libc-dev works, but I can try that in a chroot locally
[00:20] <apw> cjwatson, good plan, before it polutes the chroots
[00:35] <cjwatson> a not very randomly chosen sample of anna and man-db still compile
[00:36] <cjwatson> which includes man-db's rather rudimentary test suite so binaries also still run
[01:01] <cjwatson> LP's back up sufficiently to run change-override, so have moved things back to main where they belong
[01:01] <apw> cjwatson, thats good news.  the build is looking plausable on ppc, not quiet sure how long it'll run yet
[01:02] <apw> am inclined to upload it so you can new, and we can sleep
[01:02] <cjwatson> sounds good
[01:03] <cjwatson> need sleep myself, have to handle the kids tomorrow morning
[01:05] <apw> building the package now
[01:10] <apw> cjwatson, ok its uploaded, should be on the queue as soon as the queue runner finds it
[01:12] <apw> ok it's there
[01:14] <cjwatson> accepted, thanks
[01:14] <cjwatson> just looking at that queuebot bug
[01:14] <apw> cjwatson, heh it'll wait till tommorrow
[01:15] <cjwatson> if I do that I'll forget
[01:15] <cjwatson> (it's a one-liner anyway)
[01:18] <cjwatson> (slightly too late to demonstrate the fix, unfortunately)
[01:24] <stgraber> auto reconnect ?
[01:24] <cjwatson> stgraber: no, code upgrade
[01:25] <cjwatson> I'll add a "don't report initial queue" option for next time
[08:43] <cjwatson> apw: hm, still broken on powerpc
[08:44] <apw> cjwatson, ok that overnighter also failed to build for a new error, i have a fix for the error it hit, but am flying blind fixing it as i cannot do a full build on the porters due to chroot issues... do we want to hope this fixes it en-toto or wait for the porters to be fixed
[08:44] <apw> (from -devel)
[08:44] <apw> flying blind testing it
[08:53] <cjwatson> hm, it got quite a distance through didn't it?
[08:53] <apw> yeah i have fixed the build failure in that flavour
[08:53] <cjwatson> I'd say upload blind but keep going on davis in parallel
[08:53] <apw> ack
[08:53] <apw> what is odd is the build seems to progress in a different order on a buildd than a full build with dpkg-buildpkg -B seems to
[08:53] <apw> cjwatson, who owns the porters is that #is ?
[13:29] <apw> cjwatson, ok porters fixed and testing done, sadly failed, another build failure in perf.  I have fixed that and have built all the components on the porter successfully
[13:30] <apw> so i believe i now have a fixed source package, a full test is building as belt-and-braces
[13:32] <apw> i beleive this should be the final version, and i suspect I should just upload it
[13:40] <cjwatson> apw: I suspect so, from the sound of it
[13:41] <cjwatson> interesting ride, this
[13:41] <apw> cjwatson, ports can be a major pain at this stage, normally we can let them drift
[13:41] <apw> i am not sure the kernel has gated the release before
[13:44] <cjwatson> I don't mind letting them drift if they've built once, but I'm concerned about having the architectures out of sync for major toolchain components right at the start - the point of splitting off ports is to make them consume less developer time, and while I realise this is a bit of extra time at the start I think that having things too out-of-sync would consume more developer time in the long run
[13:45] <cjwatson> and we're not behind schedule or anything so it seems worth spending a bit of time now for less hassle later
[13:45] <apw> cjwatson, fair enough.  i am pretty hopeful from the testing i have managed to do that this one should build fine on all arches
[15:00] <apw> cjwatson, we might need to rescore that ppc build
[15:01] <cjwatson> apw: done
[22:26] <apw> cjwatson, ok finally all those builds finished, so once powerpc is new'd we have a full set of linux-libc-dev
[23:12] <cjwatson> looks like somebody's done it
[23:12] <cjwatson> doko: do we need to wait for your new gcc-4.5?  I assume it'll take a while
[23:12] <doko> cjwatson: no
[23:13] <doko> binutils should be built however
[23:13] <cjwatson> doko: did you NEW linux/powerpc?
[23:13] <doko> cjwatson: yes. anything wrong?
[23:13] <doko> did do the same for the other archs
[23:13] <cjwatson> doko: if you could use the kernel-overrides script in future to make sure the overrides are synced up with the last ABI, that would be good
[23:13] <doko> ahh, ok
[23:14] <cjwatson> can fix it up after the fact, though not 'til it's finished publishing
[23:14] <cjwatson> anyway, not urgent, linux-libc-dev will still be in main which is what matters for now
[23:14] <doko> well, I did see that linux-libc-dev did end up in main, and that was good enough for me
[23:14] <cjwatson> (the actual kernel and such has gone to universe)
[23:15] <cjwatson> I guess not much depends on linux-headers any more
[23:15] <cjwatson> binutils has finished building and seems to be published
[23:15] <doko> ok, binutils is in the archive as well, so we can open
[23:15] <cjwatson> we need to wait for linux-libc-dev/powerpc to be published
[23:16] <doko> ahh, ok
[23:16] <doko> another 20min
[23:16] <cjwatson> ok, feel free to ask a LOSA for it, I'm off to bed :-)
[23:16] <doko> will write the announce email tomorrow
[23:16] <cjwatson> I can start an autosync pass tomorrow if you like
[23:17] <doko> cjwatson: on #is or #soyuz?
[23:17] <cjwatson> probably more likely to find somebody on #is
[23:17] <cjwatson> though there's a reasonable number of sysadmins on #soyuz too
[23:19] <doko> cjwatson: see #is
[23:21]  * apw is glad to see that lot finally build