IRC logs for #baserock for Monday, 2016-07-18

*** gtristan has quit IRC06:38
*** CTtpollard has joined #baserock07:22
*** rdale has joined #baserock07:44
*** toscalix has joined #baserock07:55
*** jonathanmaw has joined #baserock08:16
*** gtristan has joined #baserock08:22
*** gtristan has joined #baserock08:24
*** tiagogomes has joined #baserock08:26
*** rdale has quit IRC08:52
*** rdale has joined #baserock08:54
*** rdale has quit IRC09:16
*** rdale has joined #baserock09:23
*** faybrocklebank has quit IRC09:35
*** faybrocklebank has joined #baserock09:35
pedroalvarezbunch of updates for next genivi release:
pedroalvarezreviews will be appreciated :)12:23
*** CTtpollard has quit IRC12:41
*** CTtpollard has joined #baserock12:42
*** faybrocklebank has quit IRC13:15
*** fay has joined #baserock13:17
*** fay is now known as Guest5230113:18
*** Guest52301 is now known as faybrocklebank13:22
paulsherwoodpedroalvarez: it all makes sense. how do i provide +1 for all 11 patches?14:18
pedroalvarezsadly one by one, although you could just vote one of them saying that the vote is valid for all of them :)14:20
pedroalvarezany other reviewers? :) :)14:30
paulsherwoodpedroalvarez: have you built and tested anything?14:41
pedroalvarezyes, built and deployed a genivi baseline14:42
pedroalvarezbooted also, and run the mock-navigation-demo14:42
paulsherwood+2 then14:45
paulsherwoodi can add that to the gerrit if you like14:45
pedroalvarezno need, I'll make a note when merging14:46
pedroalvarezthanks paulsherwood :)14:46
paulsherwoodthis is the second use-case i've seen recently for tooling to confirm that unpetrify-ref contains ref15:04
persiaWhy do we have both again?15:08
persiaI remember something about human-readable vs. machine-readable, but surely there's a more sensible way to handle that wihich is not so prone to mistakes.15:09
pedroalvarezi seem to remember that having it there has been useful for some tooling that cannot handle sha's directly15:09
paulsherwoodpersia: ref: is supposed to be a SHA. a SHA could get lost if it's not in a branch or tag... so unpetrify-ref is supposed to represent the branch or tag15:10
paulsherwoods/lost/lost by git gc/15:10
persiaYes, so why use a SHA for ref:?  Could just be a lightweight (non-editabe) tag.15:11
persiaTooling only needs a git ref, which either a sha or tag can be.15:11
persiaA branch is also a ref, although branch refs are usually expected not to be stable trees.15:11
persiaThis was the old "petrify"/"unpetrify" behaviour, where one used branches, and then ran "petrify" to switch to SHAs, and freeze everything, except freezing didn't work, and without CI (which was absent back then), unpetrified trees were usually broken.15:12
persiaNote that I think the answer to the above question is complex, but I think it's worth considering in the context of thinking about tooling to have unpetrify-ref be anything other than a comment.15:14
SotKistr that the freezing did work, its just that most of the refs in definitions were branches which led to many rebuilds15:20
persiaOh, the freezing and unfreezing did work.15:22
persiaThe issue was that if one wasn't frozen, things kept changing in frustrating ways, so we moved to an always-petrified model, hence the leftover "unpetrify-ref", which used to be strictly optional (as it was just "ref" when unpetrified)15:23
persiaI suppose if one only used stable branches for well-behaved upstreams, it would be fine, but I'm not sure that was what was being done by those who complained about the unpetrified state.15:24
*** gtristan has quit IRC15:39
*** jonathanmaw has quit IRC16:11
*** lachlanmackenzie has quit IRC17:06
*** lachlanmackenzie has joined #baserock17:10
*** gtristan has joined #baserock17:44
*** toscalix has quit IRC17:48
*** gtristan_ has joined #baserock20:08
*** gtristan_ has quit IRC20:12
*** gtristan_ has joined #baserock20:12
*** gtristan_ has quit IRC20:15
*** cosm has quit IRC23:43

Generated by 2.15.3 by Marius Gedminas - find it at!