[08:29] <infinity> henrix_: The quantal/armhf linux build failed with a SEGV (probably cosmic rays).  I've retried it, but it won't be done for another ~8h.  Just a heads-up.
[09:39] <henrix> infinity: ack, thanks.
[12:39] <BenC> So is the intent to go with 3.8 for raring?
[12:52]  * henrix -> lunch
[12:54] <xnox> BenC: yes. As far as I remember that was announced by kernel team at the end of UDS.
[13:06] <rtg> BenC, definitely
[13:11] <rtg> cking, henrix: rebooting gomeisa for kernel update
[13:12] <henrix> rtg: ack
[13:15] <rtg> ppisati_, rebooting tangerine for kernel update
[13:15] <ppisati_> rtg: ack
[14:14] <janimo> apw, anything I can do on the nexus7 kernel config syncup WI?
[14:14] <janimo> no hurry, just to make sure we're still on the same page wrt the scripted headstart you can give it
[14:33]  * ppisati_ goes out for a bit
[15:52]  * ogasawara back in 20
[17:12]  * ppisati_ -> gym
[18:28]  * rtg -> lunch
[20:47]  * rtg -> EOD
[22:33] <infinity> bjf / henrix: Is it an assumption in the bot, or human error, that prepare-package-signed was set to Invalid for a non-ABI bump kernel?
[22:34] <bjf> infinity, probably an invalid assumption in the bot
[22:34] <infinity> (For 3.5.0 in quantal and lts-quantal, no signed, which means they'll be out of sync)
[22:34] <bjf> ugh
[22:35] <infinity> If you'd like to whip up a pair of signed packages for me and throw them at the PPA, you can fix the bot later. :)
[22:36] <bjf> infinity, will probably do that
[22:36] <infinity> (Since I already promoted linux/quantal before noticing the error)
[22:40] <bjf> infinity, i'm working on it
[22:40] <infinity> bjf: Oh, wait.  Were all the reverts in modules?
[22:41] <bjf> infinity, looking, but i don't think so
[22:41] <infinity> Actually, you know what?  While what I was about to say was technically true, it's probably just confusing to make the distinction and branch the workflow.
[22:41] <bjf> infinity, it's in the fs subsystem
[22:42] <infinity> Much simpler to just say "if linux revs, linux-signed also must" than to say the more-true-but-more-error-prone "if linux revs, but the kernel image stays the same, linux-signed can be left alone".
[22:42] <infinity> bjf: Yeah, check.  We need a new -signed regardless, then. :P
[22:42] <infinity> bjf: Either way, like I said, it's probably just confusing to bother making the distinction anyway.
[22:43] <bjf> infinity, agreed, let's not try to optimize it
[22:57] <bjf> infinity, they are building
[23:05] <infinity> bjf: My hero.  I'll fix up the bugs manually, if you want to make sure the bot gets this right next time.
[23:06] <herton> infinity, bjf: it's fixed now. Our script to create tracking bugs was mistakenly marking it as invalid, as it did a substring check for 'prepare-package-', as previously we didn't have the -signed task, now it should behave properly
[23:08] <infinity> herton: Oh, shiny.  Thanks for the quick fix.
[23:35] <maxb> Hi. https://help.ubuntu.com/community/AsusZenbookPrime suggests using the kernel parameter acpi_osi= to make backlight control keys work. How concerned should I be about this breaking other unrelated things?
[23:37] <maxb> (I have enough understanding of what this does to know that it's causing the kernel to lie rather a lot to the ACPI stuff)