| onefang | I'm off now. | 00:01 |
|---|---|---|
| Centurion_Dan | KatolaZ: can you remember how to fix the issue wit arch="all" packages failing on not finding the amd64.changes ? | 00:02 |
| Centurion_Dan | in builds? | 00:02 |
| Centurion_Dan | see http://ci.devuan.org/job/cloud-init-binaries/architecture=all,label=all/15/ | 00:02 |
| KatolaZ | it's a problem in the buildhost | 00:08 |
| KatolaZ | it's actually a problem in pbuilder | 00:09 |
| KatolaZ | that we need to fix | 00:09 |
| KatolaZ | but the "all" should be built bu a buildhost that had the "correct" pbuilder version | 00:09 |
| KatolaZ | IIRC, build15 does | 00:09 |
| KatolaZ | Centurion_Dan: ^^ | 00:09 |
| Centurion_Dan | KatolaZ: Thanks... I'll compare and fix it :-) | 00:41 |
| Centurion_Dan | I' | 00:41 |
| Centurion_Dan | I'd like to move alot of that config out of pbuilder and into the job scripts and add parameters to provide distrobution, and flavours - we'll need to do this to support kfreebsd and hurd and alternatives to libc etc. | 00:44 |
| KatolaZ | Centurion_Dan: the problem is that the pbuilder interface has changed | 00:50 |
| KatolaZ | the parameters are already in the scripts | 00:50 |
| KatolaZ | but they don't get passed properly to the new version of pbuilder | 00:50 |
| Centurion_Dan | ah, right. | 00:50 |
| KatolaZ | will get back to that soon I guess | 00:51 |
| Centurion_Dan | KatolaZ: I'm working on a new release of jenkins-debian-glue that should fix the issue. | 07:50 |
| KatolaZ | golinux: the "mirror broken" error was just a user issue, as expected | 10:58 |
| onefang | Yay! B-) | 11:22 |
| gnu_srs1 | Centurion_Dan: KatolaZ: Are you working on eudev and policykit-1 for ceres/beowulf? Why not contact me? | 23:05 |
| fsmithred | gnu_srs1, Centurion_Dan said he was just using policykit to test jenkins | 23:22 |
| fsmithred | getting the build system ready. | 23:22 |
| KatolaZ | gnu_srs1: wazzup mate? | 23:32 |
| Centurion_Dan | gnu_srs1: not yet, although I'm keen to get on with beowulf... just beating the ci back into shape after jenkins updates broke it... | 23:47 |
| KatolaZ | Centurion_Dan: you probably mean ceres :P | 23:51 |
| Centurion_Dan | both actually... | 23:52 |
| KatolaZ | nope | 23:52 |
| KatolaZ | only ceres | 23:52 |
| Centurion_Dan | ?? | 23:52 |
| KatolaZ | we won't drop any package in beowulf, right? | 23:52 |
| KatolaZ | just in ceres | 23:52 |
| KatolaZ | unstable/testing/stable, right? | 23:52 |
| KatolaZ | that's what we have been talking so far | 23:52 |
| Centurion_Dan | we can't ignore beowulf ... so both will get attention - but major changes- ie dropped or newly forked pacakges will of course need to start in ceres | 23:54 |
| KatolaZ | Centurion_Dan: nope | 23:54 |
| KatolaZ | we will build for ceres | 23:54 |
| KatolaZ | and then move the same packages to beowulf | 23:54 |
| KatolaZ | which is effectively ceres | 23:54 |
| KatolaZ | otherwise we won't ever stop building ad-hoc stuff | 23:55 |
| KatolaZ | beowulf is testing, so it is basically ceres | 23:55 |
| KatolaZ | minus the really broken stuff | 23:55 |
| KatolaZ | so anything that works in ceres will work in beowulf | 23:56 |
| KatolaZ | (minus the broken stuff, but we have so few packages that we won't notice that perhaps) | 23:56 |
| KatolaZ | we should just build for ceres from now on | 23:56 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!