IRC logs for #baserock for Friday, 2015-12-11

*** gtristan has quit IRC03:42
*** gtristan has joined #baserock04:13
*** locallycompact has joined #baserock06:49
*** locallycompact has quit IRC07:16
*** locallycompact has joined #baserock07:17
*** locallycompact has quit IRC07:35
*** locallycompact has joined #baserock07:37
*** locallycompact has quit IRC07:43
*** locallycompact has joined #baserock07:44
*** fay_ has joined #baserock08:24
*** ctbruce has joined #baserock08:38
*** toscalix has joined #baserock08:45
*** rdale has joined #baserock08:53
*** fay_ has quit IRC09:06
*** bashrc_ has joined #baserock09:07
*** locallycompact has quit IRC09:27
*** locallycompact has joined #baserock09:29
*** jonathanmaw has joined #baserock09:52
* tiagogomes_ spots a baserock logo on gerrit09:55
Zaraooh, looks good09:58
*** locallycompact has quit IRC10:03
*** locallycompact has joined #baserock10:05
*** ssam2 has joined #baserock10:05
*** ChanServ sets mode: +v ssam210:05
*** locallycompact has quit IRC10:08
*** locallycompact has joined #baserock10:09
*** locallycompact has quit IRC10:13
*** fay_ has joined #baserock10:18
*** Lachlan1975 has joined #baserock10:25
*** edcragg has joined #baserock10:48
*** locallycompact has joined #baserock10:49
*** locallycompact has quit IRC11:25
*** locallycompact has joined #baserock11:25
*** fay_ is now known as faybrocklebank11:57
rdaleare morph releases no longer tagged? the most recent looks like baserock-14.46, but that is from a year ago13:22
persiaIt may be that morph no longer releases.13:24
persiaI don't think morph has enough CI for CD, but it may be that we've fallen back to our old habit of just picking some recent SHA to put in a deifnitions release, rather than waiting for morph to release.13:25
* SotK suspects that is the case13:25
rdaleit makes it tell to tell whether morph version x, supports definition format y, if morph doesn't have any sort of version13:28
*** gtristan has quit IRC13:34
*** gtristan has joined #baserock14:35
*** toscalix has quit IRC14:51
*** toscalix has joined #baserock14:54
*** edcragg has quit IRC15:04
*** edcragg has joined #baserock15:08
*** VLetrmx has joined #baserock15:47
VLetrmxnice work on the logo whoever did that15:48
tiagogomes_that was Pedro15:49
rjekbtw, can somebody who has control over's DNS please change the MX records to be "MX 10", that'd be dandy15:50
rjek(rather than MX 20 and MX 10
ssam2rjek: that would be the Codethink sysadmins16:06
ssam2so gary_perkins16:06
* rjek looks at gary_perkins16:07
*** tom__ has joined #baserock16:10
*** CTtpollard has quit IRC16:10
*** tom__ is now known as CTtpollard16:10
persiaHrm?  I thought the baserock DNS keys were known to some Baserock folk that weren't Codethink sysadmins, and with separate credentials to Codethink credentials.16:13
persiaIf this isn't true, then we should ask Codethink for independent DNS.16:14
ssam2please do16:15
ssam2i agree16:15
*** franred has quit IRC16:15
rjekI am happy to host DNS as well as mail, btw16:15
rjekSo you can administer it through the same interface16:16
rjekBut that will require some degree of careful management of moving it16:16
persiaSponsors of infratructure are always welcome :)  I think we need a clear understanding of the current state before taking action.16:16
* persia only has memories of rumours, which is unlikely to be sufficiently authoritative16:17
gary_perkinsrjek: done. There is now just the one MX record, priority 10, pointing to
rjekgary_perkins: Ta16:20
* rjek also sets the "customer is competent" flag on the Baserock ppf account so you can add and delete mail and dns domains yourselves16:22
rjekDo not click the delete buttons unless you mean it16:22
ssam2seems like a good solution, thanks for the offer16:24
ssam2but needs some time to actually orchestrate the move16:26
*** franred has joined #baserock16:27
persiaWe should probably have a formal discussion on the mailing list and get +1s from some of the mergers before taking action.16:28
*** franred has quit IRC16:28
*** franred has joined #baserock16:31
*** franred has quit IRC16:54
*** CTtpollard has quit IRC17:16
*** ctbruce has quit IRC17:17
VLetrmxrichard_maw, can you clarify the point about blocking on ? i *think* i know what you mean, but i could be wrong17:47
*** Lachlan1975 has quit IRC17:50
VLetrmxit would be nice if gerrit added cover letters17:51
richard_mawVLetrmx: if tar fails before opening the fifo, we block when we open() the FIFO17:51
richard_mawfifos are special17:51
richard_mawthe problem with a regular pipe, is that we can either use a fixed string input and safely get the output, or we can throw away stdout/stderr and feed it input with morphlib.util.copyfileobj17:53
richard_maweither that, or we need an event loop for providing input and reading output17:54
*** bashrc_ has quit IRC17:54
richard_mawbut GKJFDing Popen.communicate only takes a fixed string, rather than any more useful way of providing input17:54
richard_mawas for chances of the subprocess dying before we open the FIFO, likely if it's busybox tar instead of gnu tar, since it will complain about the command-line arguments17:56
*** locallycompact has quit IRC17:56
*** toscalix has quit IRC17:59
*** Lachlan1975 has joined #baserock17:59
* richard_maw kicks APIs that don't take iterables18:02
*** jonathanmaw has quit IRC18:03
persiaI *like* that gerrit disallows cover letters: in my opinon most of that data either belongs in the commit message or in the change discussion.18:03
VLetrmxi don't know, sometimes you might want to describe what a series is trying to describe overall18:04
VLetrmx*achieve overall18:04
persiaIf you haven't managed to do that through the narrative of your commit messages, attempting to explain it away in a cover letter doens't help the person who needs to understand why you were doing things a decade later.18:05
VLetrmxi guess18:05
persiaNot every cover letter does this, but many seem to describe the point of a series of things, most of which have one-line commit messages that only make sense to people who read the cover letter.18:05
persiaSince the cover letter doesn't go into the repo, that means the commit messages don't make sense when using git blame.18:06
persiaWhich makes it hard to understand why some code construction that is confusing was put there in the first place.18:06
ratmice______I think they are useful especially in large multi-maintainer projects where maintainers review some subset of a patch series which overlaps their domain18:16
*** ssam2 has quit IRC18:22
*** edcragg has quit IRC18:31
*** tiagogomes_ has quit IRC18:33
*** Lachlan1975 has quit IRC19:07
*** rdale has quit IRC19:43
*** locallycompact has joined #baserock20:26

Generated by 2.15.3 by Marius Gedminas - find it at!