[08:07]  * cjwatson enables daily builds for raring
[08:12] <infinity> cjwatson: Did you get webops to make livefs chroots?  I may have missed that in the excitement of our unusual opening.
[08:37] <cjwatson> infinity: *cough* no
[08:37] <cjwatson> infinity: wanna check? :)
[08:38] <infinity> cjwatson: Yeah, so, that might not work. ;)
[08:38] <infinity> cjwatson: I need a webops who isn't about to go to bed, but I'll get it looked after.
[08:38] <infinity> Every six months, this really reminds me that livefs-in-soyuz has other knock-on benefits.
[08:39] <infinity> Like not proxying people to create chroots by hand...
[08:41] <stgraber> cjwatson: I guess you noticed the grub2 FTBFS?
[08:42] <infinity> Hahaha.
[08:42] <infinity> More gets().
[08:43] <infinity> Did no one get the memo 13 years ago?
[08:44] <infinity> (At least, I think it was first deprecated in C99...)
[08:47] <stgraber> infinity: marked obsolescent in 2008 and was removed from the spec in C11 according to the manpage
[08:48] <infinity> stgraber: That manpage skips the C99 reference.
[08:53] <infinity> stgraber: Oh, heh.  Nevermind.  C99 drafts happened much later than 1999. :P
[08:53] <infinity> Either way, the appearance of gets() in anything other than a CompSci project teaching people about buffer overflows is always a bug.
[09:30] <infinity> cjwatson: qapt is probably a much more interesting test-case for britney out-of-date migration issues, since it's also snagging a transition.
[09:31] <infinity> cjwatson: Likely a better end-to-end test-case anyway.
[09:44] <jbicha> I'm curious about ptlib getting stuck in the raring proposed transition too
[09:47] <infinity> jbicha: Same NBS issue, probably.  I'll look into it in a sec/
[09:49] <jbicha> yeah, it adds new libraries
[10:07] <cjwatson> stgraber: thanks, I'll fix that
[10:08] <cjwatson> infinity: it's not actually using gets directly - it's an obscure thing inside gnulib that's trying to use typeof on it to warn about its misuse
[10:08] <infinity> cjwatson: Oh, the irony.
[10:09] <cjwatson> yeah.  fixed upstream a while back but it takes a while to filter down.
[10:40] <infinity> tumbleweed: Hrm, reverse-depends(1) doesn't do pockets?
[10:40] <tumbleweed> it probably merges pockets
[10:41] <infinity> tumbleweed: I'm going to go with no, it doesn't.
[10:41] <tumbleweed> yeah, that's about it
[10:41] <infinity> (And even if it did, the data would always be slightly incorrect, then, as I wouldn't know which version of packages it's looking at)
[10:41] <tumbleweed> I have been fairly tempted to rewrite it over the last day or two
[10:42] <infinity> But no, it doesn't merge.  If it did, it would find rdeps for libpt2.10.7
[10:42] <infinity> (Which is only in proposed)
[10:42] <tumbleweed> it currently regenerates the reverse-index from source a couple of times a day, but if it were iterative, it could try and follow publisher cycles
[10:43] <tumbleweed> yup, release pocket only. No merging
[10:44] <infinity> Check.  Please to add proposed/updates/security targets? :)
[10:44] <infinity> (And t-p-u, s-p-u, etc)
[10:44] <infinity> Not that I care as much about the latter ones, but someone might.
[10:44] <tumbleweed> it'll bloat the DB a lot, so I probably have to re-write it first
[11:20] <cjwatson> infinity: care to approve that grub2 binary for me when you have a minute?