[00:42] <stgraber> micahg, Laney: uploaded a lxd backport for trusty and a cgmanager and lxc backport for precise. The former is just to keep up to date with xenial. The latter is because of how outdated and unsupported the current backport is, getting in sync with trusty in precise should make everyone's life much easier (at the cost of having to get cgmanager backported too).
[00:42] <stgraber> no rush on any of those
[01:05] <micahg> stgraber: thanks for trying to keep things updated
[10:47] <LocutusOfBorg1> hi please accept libwebsockets from the queue, the reverse dependency (mosquitto) has been uploaded in debian a few minutes ago
[11:57] <cjwatson> LocutusOfBorg1: no need to ask for ones with the same version in Debian, we do those in bulk
[13:15] <LocutusOfBorg1> cjwatson, actually I was trying to avoid a mosquitto sync with the old libwebsockets, and then the need for a rebuild
[13:15] <LocutusOfBorg1> was that case possible?
[13:28] <cjwatson> LocutusOfBorg1: oh right, that would certainly have been possible, fair enough then
[13:31] <LocutusOfBorg1> so when a package is built but not accepted is not used by reverse dependencies
[13:41] <cjwatson> LocutusOfBorg1: indeed
[13:42] <LocutusOfBorg1> I hope to become a MOTU soon, to be able to binNMU them by myself :)
[14:08] <mapreri> LocutusOfBorg1: you don't binNMU stuff in ubuntu :P
[14:08] <mapreri> though I've to understand how transitions are done so quickly without binNMUs....
[14:09] <mapreri> takes time to download, dch, debuild -S, debsign, dput every package, umh
[14:09] <cjwatson> scripts to do no-change rebuilds
[14:09] <cjwatson> it does indeed take some time but not as much as you might think
[14:10] <cjwatson> I have horrible hacky messes like http://paste.ubuntu.com/14137855/
[14:10] <cjwatson> I think others have something a bit more elegant but this is good enough for my purposes :)
[14:12] <mapreri> eheh.  OOI, why do you spawn subshells inside functions?
[14:12] <cjwatson> mapreri: so that I don't get left wherever they've cded to
[14:13] <mapreri> oh, make sense
[14:13] <cjwatson> I could cd back at the end, but then error handling and life's too short, a subshell is simpler
[14:13] <cjwatson> in case it's not obvious, no support implied, if you've found some bug in this stuff I probably already know and don't care. :-)
[14:14] <mapreri> :P
[14:15] <mapreri> let's say, if I'll ever need something similar I think I'll write my own crap :)
[14:15] <mapreri> o.O( at least I've full copyright on it, it must mean something! )
[14:16] <mapreri> (=> it doesn't )
[14:16] <cjwatson> heh
[14:16] <cjwatson> it's not particularly hard
[14:26] <LocutusOfBorg1> actually no-change rebuild is the same as a binNMU :)
[14:30] <cjwatson> LocutusOfBorg1: actually it's not quite
[14:30] <cjwatson> but close enough for at least some purposes
[14:30] <cjwatson> (binNMUs keep the source version unchanged)
[14:32] <LocutusOfBorg1> true, but at the end the result seems the same
[15:04] <cjwatson> LocutusOfBorg1: sometimes, if you don't care about source version != binary version which does in fact sometimes matter :)
[15:22] <LocutusOfBorg1> true :)
[15:29]  * mapreri doesn't care, since already binary versions could be different than source version anyway...
[15:31] <mapreri> another difficult bit of binNMU is that you can have multiple binaries with the same name out of the same (source, version) tuple, I think.  somebody might rely on that?