=== yofel_ is now known as yofel [12:08] lamont, infinity, could someone take a look at celbalrai ? it seems to be completely dead [12:08] 11:16 cjwatson: Hmm, it's not happy - lots of I/O errors. I think it may be out of action for a while. [12:08] FWIW [12:08] (yesterday) [12:09] oh, well, i thinnk we should replace the HW [12:09] the last builds that actually produced logs had very weird gzip decompression errors [12:09] mm [12:12] oh, and on a sidenote.... i would REALLY appreciate if some archive admin could let linux-nexus7 out of NEW, it is blocking images building and stuck since nearly a week now (though in the current situation with the builder it might be less pressing) [12:12] *image [12:13] i think infinity had agreed to give it a review, but he seemingly didnt get to it === mmrazik is now known as mmrazik|lunch === mmrazik|lunch is now known as mmrazik === doko__ is now known as doko === mmrazik is now known as mmrazik|afk_for_ [14:26] so glew got reverted, but the britney block is still in place. [14:27] xnox, which is not so much of an issue because the revert is basically the raring version with an extra changelog [14:27] true =) [14:59] oh I see, block doesn't actually take a version === Guest8350 is now known as balloons [15:20] I'll be doing a fair bit of code shuffling in a branch of ubuntu-archive-tools and the cdimage branch to rebase all the tools on the "official" qatracker python module and make sure our code supports dealing with multiple trackers (to support the chinese builds). [15:20] I'll be testing some of the changes directly on nusakan and will trigger a few manual builds [15:30] seb128: I've dropped the glew block now, anyway [15:36] infinity: Any chance you could have a look over my secure boot stuff in precise-proposed/{NEW,UNAPPROVED}? === mmrazik is now known as mmrazik|otp === mmrazik|otp is now known as mmrazik [16:56] cjwatson: Can do. [16:57] cjwatson: As soon as me three laptops settle down and turn into one again. [16:57] infinity, jani claimed you wanted to look at linux-nexus7 too [16:57] would be helpful to have that in the archive soon [16:57] ogra-cb: Same caveat. Let me find some coffee or something. ;) [16:58] yeah, well, as long as the live builder is broken its less urgent :) [16:58] celbalrai.buildd still isnt reachable from nusakan, so i guess it will still take a while [16:59] Is IS aware that it's sad? [16:59] Yeah, apparently it's producing I/O errors and failing to get past the initramfs. [16:59] Oh, "fun". [16:59] ChrisS had a brief look for me yesterday. [16:59] Before that it was just randomly corrupting builds. [17:00] well, as i said above it also produced pretty weird gzip errors over the last week or so [17:00] I wonder if it's loose or in the box. [17:00] i would blame memory or disk and suggest to just replace it [17:00] ogra-cb: I'd be more inclined to blame power, actually. [17:01] http://www.embedded.rs/products/ti-omap5430-pico-itx-sbc [17:01] with one of these :) [17:01] I wonder if, going back through history, we'd discover that every one of the weird corruption errors we've seen on Pandas has been on the ones in the PandaBox (which celbalrai is, I just looked), and that PSU just isn't delivering enough power. [17:01] (it claims software compatibility with the panda, i wonder if thats true for kernel and bootloader) [17:02] hmm [17:02] ogra-cb: It's an OMAP5, it can't possibly be compatible at the bootloader and kernel level. [17:02] it started pretty much out of the blue though [17:03] infinity, yeah, thats why i was wondering about that SW compatibility claim [17:03] I suspect they just mean userspace, since not all hobbyist dev boards are A-series cores. [17:03] yeah [17:03] And I'm betting some people have been burned by switching between A and M and realising nothing works. [17:04] well, in any case it looks like a chromebook with sata :) [17:04] though the usb 3.0 is breezing fast on the cb :) [17:08] does a regression test for an SRU for bug 938670 seem sufficient to anyone else? [17:08] Launchpad bug 938670 in Unity Videos Lens "Remote scope should use libsoup" [Undecided,Fix released] https://launchpad.net/bugs/938670 [17:08] er, bug 936870 [17:08] Launchpad bug 936870 in unattended-upgrades (Ubuntu Precise) "unattended upgrade failing due to incorrectly detected conffile prompt due to "newconffile" line" [Medium,Triaged] https://launchpad.net/bugs/936870 [18:20] cjwatson: All accepted. I goofed the overrides on the NEW stuff, but fixed as of :33 [18:22] Great, thanks [18:23] cjwatson: I take it we've opted to just hand-wave past the fact that all the -signed stuff probably actually belongs in restricted, not main? [18:24] I had a discussion with slangasek about that a while back. I forget our exact justification. [18:24] I think as long as the user can produce a signed version that would fit if they installed their own keys on the firmware, we decided that was good enough for us [18:25] Sure, the source is absolutely free, but the binaries aren't reproducible, which is the icky part. But I'm not going to be picky about it either. [18:27] If we used signed .debs, I wouldn't consider those non-free [18:27] e.g. signed with the archive key [18:27] shim-signed is skating on the edge a bit, I'll admit [18:28] cjwatson: Yeah, I'm not sure that's quite the same thing, but I can see the extension of the argument. [18:28] cjwatson: More convincing if we can reliably strip the signatures and replace them (can we?) [18:28] I believe that's even tested, at least in shim-signed [18:28] Shiny. [18:28] sbattach does it [18:29] I thought there were some issues with that not returning one to a pristine unsigned binary, but maybe that was resolved. [18:29] AIUI that was the case but was fixed === Ursinha_ is now known as Ursinha === rsalveti_ is now known as rsalveti [20:27] cjwatson: any chance you're still around? :) [20:28] I submitted https://code.launchpad.net/~stgraber/ubuntu-archive-tools/various-qatracker-improvements/+merge/134192 for inclusion in ubuntu-archive-tools [20:30] ^ or anyone in ~ubuntu-archive-tools who has a few minutes to review the merge proposal [20:31] the diff is pretty massive as that's porting all the code to using the upstream python-qatracker module. Rest of the changes are mostly cleaning up, python3 support and a bunch of bugfixes I found while testing [20:32] the cdimage code will need a couple of matching changes once that one lands but those are trivial and I have commit access :) [20:53] Note to nusakan users: I merged that branch manually in the copy that's on nusakan so I can land the cdimage bits [20:53] (running a test now) [20:57] infinity: even if it doesn't return you to the pristine unsigned binary, it returns a functional equivalent, which is about all we get from package rebuilds anyway (if we're lucky ;) [20:57] slangasek: Yeah, I thought there may have been concerns at one point that it wasn't even quite doing that. [21:00] ah, nope [21:13] looking at ubuntu-dev-tools in the precise-proposed queue there is an undocumented (in the changelog) change to requestbackport [21:13] http://launchpadlibrarian.net/122313290/ubuntu-dev-tools_0.141_0.141ubuntu0.1.diff.gz [21:13] I'd approve it but wanted a second opinion [21:14] infinity, slangasek? [21:15] doesn't look like a change that needs SRU validation, so yeah, I'd accept as-is [21:55] and success! we now have the Chinese builds automatically posted on http://localized-iso.qa.ubuntu.com [21:56] infinity: while playing on nusakan I noticed there's a non-commited change to bin/wait-for-package is that one of yours? [21:56] I also reverted your release time etc/default-arches change and updated the edubuntu entry for raring while I was at it [21:57] stgraber: Not mine, no. I've been ruthlessly hunting and committing production changes when I see them. [23:32] rejected nexus7-dev-tools at mterry's request [23:32] (uploader)