IRC logs for #baserock for Tuesday, 2016-04-05

*** zoli_ has quit IRC01:27
*** zoli_ has joined #baserock01:39
*** cosm has quit IRC02:41
*** cosm has joined #baserock02:45
*** locallycompact has joined #baserock05:57
*** locallycompact has quit IRC07:21
*** locallycompact has joined #baserock07:22
*** bruce_ has joined #baserock07:28
*** edcragg has joined #baserock07:30
*** locallycompact has quit IRC07:41
*** gtristan has joined #baserock07:43
*** bashrc_ has joined #baserock08:01
*** rdale has joined #baserock08:23
*** anahuelamo has joined #baserock08:23
*** franred has quit IRC08:33
*** tiagogomes has joined #baserock08:34
*** bashrc_ has quit IRC08:38
*** jonathanmaw has joined #baserock08:39
*** bashrc_ has joined #baserock08:39
*** franred has joined #baserock08:47
*** gtristan has quit IRC08:53
*** edcragg has quit IRC09:15
*** ssam2 has joined #baserock09:15
*** ChanServ sets mode: +v ssam209:15
*** locallycompact has joined #baserock09:22
*** franred has quit IRC09:24
*** edcragg has joined #baserock09:29
*** franred has joined #baserock09:34
*** gtristan has joined #baserock09:43
franredpedroalvarez, reviewed - I think some of them need some changes because they apply valid patches instead of changing the submodule URLs and I think there are a sha1 which misses the first character - feel free to merge all the series if you agree with the observations and you modify them10:24
pedroalvarezfranred: I was looking at the reviews, thanks for taking the time!10:25
franredno probs :)10:25
pedroalvarezfranred: I hope you understand now why I didn't want to address all of these changes when doing the migration10:40
franredyeah, it was a long series10:41
franredpedroalvarez, I've merged the last 2 - Thank you for the explanation about James patches10:42
pedroalvarezfranred: ta10:42
franrednow waiting for mason :)10:42
pedroalvareznote, the series isn't finished10:42
pedroalvarezI'll try to continue soon10:42
*** gtristan has quit IRC10:43
franredpedroalvarez, ok, I will review anything is in gerrit whenever I have some spare time10:44
* pedroalvarez restarts mason so that it tests useful stuff10:44
pedroalvarezfranred: tvm10:46
*** gtristan has joined #baserock10:46
*** gary_perkins has quit IRC11:24
*** CTtpollard has quit IRC11:39
*** CTtpollard has joined #baserock11:40
*** paulwaters_ has joined #baserock11:50
* pedroalvarez broke findutils:
*** CTtpollard has quit IRC11:57
*** CTtpollard has joined #baserock11:57
*** bruce_ has quit IRC12:00
*** bruce_ has joined #baserock12:00
*** faybrocklebank has quit IRC12:01
*** bruce_ has quit IRC12:02
*** fay_ has joined #baserock12:02
*** CTtpollard has quit IRC12:05
*** CTtpollard has joined #baserock12:05
*** gtristan has quit IRC12:05
*** paulwaters_ has joined #baserock12:10
*** CTtpollard has quit IRC12:10
*** CTtpollard has joined #baserock12:11
*** gary_perkins_ has joined #baserock12:11
*** paulwaters_ has joined #baserock12:15
*** bruce_ has joined #baserock12:16
*** paulwaters_ has joined #baserock12:26
*** anahuelamo has quit IRC12:43
*** toscalix has joined #baserock13:03
*** anahuelamo has joined #baserock13:20
*** fay_ has quit IRC13:22
*** vgrade_ has quit IRC13:22
*** wdutch has quit IRC13:22
*** thrace has quit IRC13:22
*** fay_ has joined #baserock13:24
*** vgrade_ has joined #baserock13:24
*** wdutch has joined #baserock13:24
*** thrace has joined #baserock13:24
edcragghi, i'm looking at making the script in definitions work with ybd and morph, i was wondering if it would be ok to change the licence check script to use the original git caches for ybd and morph, rather than copying them to a new repo folder? it would be quicker...13:30
pedroalvarezbut the cached repos are bare, right? you can't run against them13:31
pedroalvarezmaybe they are not for ybd?13:31
edcraggah, good point, i hadn't got as far as testing that yet13:32
pedroalvarezimo, using morph tooling to access morph cache is better than accessing the cache directly. Internal implementation of morph might change in the future13:33
pedroalvarezsame for ybd13:33
ssam2rather than copying the whole repo, you could just check out one commit13:33
ssam2into a directory outside the "bare" clone13:33
ssam2i think both Morph and YBD have code to do this. In Morph it's called "extract_commit"13:33
Zarahm, krotscheck mentioned yesterday he was writing a tool to check licenses for dependencies, is it worth me pointing him to any of this or likely to be totally non-generalisable?13:34
pedroalvarezZara: our script only clones repositories that we are using in the definitions, and then run, which I believe is borrowed from...13:36
pedroalvarezI don't know :/13:36
edcraggyes this just parses definitions, clones a git repo, and runs licensecheck.pl13:37
Zaraokay. wondered as I think in general what he's doing has some overlap; he mentioned it offhand but he's recently also been setting up mirrors for npm stuff, so i think that's like the openstack-infra equivalent of a trove13:37
Zarait's nice seeing similar stuff in the wild used to solve problems.13:38
pedroalvarezquicker solution: fall back to git-clone13:39
pedroalvarezbetter solution: ybd implementing something like get-repo13:39
pedroalvarezpro and complex soultion: Use morphlib and ybd modules to achieve the same as previous solution13:39
pedroalvarezany other opinions?13:39
edcraggis it not a bit silly to have to reimplement everything to use multiple apis?13:40
edcraggseems unsustainable to me, definitions should surely not depend on features of ybd or morph?13:41
edcraggi don't know...13:42
edcraggi think the most sustainable is the first option, pedroalvarez, but it's also going to be the slowest13:44
ssam2does the script need to be in definitions.git? surely it's genivi-specific, and was put in definitions just because that was where things go13:45
ssam2baserock-license-check could be its own project and could depend on either morphlib or ybd13:45
*** gtristan has joined #baserock13:46
pedroalvarezit is and it isn't genivi specific. Only  used for genivi releases, but might be useful for other things13:47
ssam2really i figure we should make a sort of 'basreock-tools' repo that would contain stuff like this13:49
ssam2neither morphlib or ybd provide any API compatibility guarantees, so it should just embed one or other of those as a submodule13:50
ssam2and then reuse the functions from within13:50
Zarahm, I feel like some combination of the npm import tool + would do what krotscheck was after, except 1) it wouldn't account for transitive dependencies (because I didn't really get the significance of semver when I worked on it), 2) import tool probably broken these days 3) he's probably not going to want to set up the environment in which it'd run.13:51
ssam2import tool could run in any OS probably, now that morphlib embeds a copy of cliapp13:54
ssam2it probably has bitrotted though13:54
Zarayeah, I suspect your and ripsum's work holds up better than mine, but afaik nobody has been resourced to maintain it, and I certainly haven't looked at it for ages.13:55
Zarait may just be a hobby project of his, anyway13:56
ZaraI'll mention it, jic, thanks.13:57
pedroalvarezyay, new contributor! thanks for the patch anahuelamo!14:08
pedroalvarezedcragg: btw, I still think that falling back to git-clone is the easiest, quicker, and less bikeshedding solution atm14:10
edcragg:) yes i agree14:11
edcraggssam2: i have just thought of one problem with a separate baserock-tools thing... surely in this case it means the caches have to be available, and obtained by morph. so if you're starting from the point of having built a system with another tool, the cache won't necessarily be available anwyay.14:12
pedroalvarezas I understand it, the main goal is to make the script work in an environment where you don't have `morph` available14:12
*** gtristan has quit IRC14:13
edcraggit is simpler, will always work, but has no cacheing14:14
edcraggor, it will generate its own cache essentially14:15
pedroalvarezultimate solution, make ybd and morph use the same library for repocaching :)14:15
edcragginteresting idea :)14:15
edcraggthey do appear to use the same format as it is14:16
edcraggfor gits14:16
pedroalvarezI think it would be good to separate the functionality from the build tools14:16
pedroalvarezand sharing code will help both build tools at the same time14:17
locallycompactI had thought about doing this as a WrapFS once14:18
ssam2edcragg: both morphlib  and ybd's library are available to clone repos into the local repo cache14:19
ssam2edcragg: but I guess it'd be annoying if their cache formats differed, and you used one tool but needed to use the other14:19
ssam2i mean, you used YBD, and the baserock-tools stuff expected a Morph case14:19
ssam2ideal solution for me would be to have a standard library that both could use to manage repo caches in a compatible way14:20
*** gtristan has joined #baserock14:20
ssam2which is actually super easy because they are compatible, in practice14:20
ssam2i think "shared local cache of Git repos" is a useful tool outside Baserock. the wrapdb idea is interesting14:21
locallycompactparameterised wrapfs14:21
ssam2but this is massive scope-creep. Maybe we could just embed YBD in Morph for now and use YBD's repo caching code in morphlib ;-)14:22
*** richard_maw has left #baserock14:23
*** toscalix has quit IRC14:30
*** richard_maw has joined #baserock14:30
bashrc_ssam2: yes, I've tried implement something like a baserock trove in other projects14:33
ssam2did you get to the point of publishing any such thing?14:39
ssam2i know richard maw began such a library14:40
*** gtristan has quit IRC14:43
bashrc_this doesn't do any of the fancy conversions to git
ssam2cool, thanks15:09
*** anahuelamo has quit IRC15:10
*** anahuelamo has joined #baserock15:11
*** toscalix has joined #baserock15:12
*** toscalix has quit IRC15:13
*** toscalix has joined #baserock15:13
* paulsherwood is only skimming this conversation...15:15
*** richard_maw has left #baserock15:15
bashrc_what that script does is scan for repos and then mirror htem under a special "mirrors" user. Other users can then clone from that one via ssh. In this way one box can support another, even if the original upstream repos aren't available15:17
*** toscalix has quit IRC15:20
*** toscalix has joined #baserock15:22
*** bruce_ has quit IRC15:23
*** richard_maw has joined #baserock15:27
*** toscalix has quit IRC15:32
*** toscalix has joined #baserock15:34
*** toscalix has quit IRC15:39
*** toscalix has joined #baserock15:48
*** toscalix has quit IRC15:48
*** toscalix has joined #baserock15:52
*** fay_ has quit IRC15:56
*** cosm has quit IRC16:04
*** toscalix has quit IRC16:04
*** gtristan has joined #baserock16:06
*** ssam2 has quit IRC16:22
*** toscalix has joined #baserock16:29
*** jonathanmaw has quit IRC17:01
*** bashrc_ has quit IRC17:02
*** toscalix has quit IRC17:03
*** rdale has quit IRC17:32
*** locallycompact has quit IRC17:34
*** anahuelamo has quit IRC17:41
*** edcragg has quit IRC17:46
*** locallycompact has joined #baserock18:15
*** cosm has joined #baserock18:17
*** gtristan has quit IRC18:26
*** toscalix has joined #baserock18:53
*** toscalix_ has joined #baserock18:54
*** toscalix has quit IRC18:58
*** toscalix_ has quit IRC19:01
*** toscalix_ has joined #baserock19:07
*** toscalix has joined #baserock19:08
*** toscalix_ has quit IRC19:11
*** toscalix has quit IRC19:18
*** toscalix has joined #baserock19:19
*** toscalix has quit IRC19:26
*** toscalix has joined #baserock19:27
*** toscalix has quit IRC19:39
*** toscalix has joined #baserock19:54
*** toscalix has quit IRC20:00
*** toscalix has joined #baserock20:03
*** toscalix has quit IRC20:03
*** toscalix has joined #baserock20:04
*** toscalix has quit IRC20:05
*** toscalix has joined #baserock20:09
*** edcragg has joined #baserock20:09
*** toscalix has quit IRC20:10
*** toscalix has joined #baserock21:01
*** toscalix has quit IRC21:10
*** locallycompact has quit IRC21:39
*** locallycompact has joined #baserock21:48
*** locallycompact has quit IRC21:54
*** radiofree has quit IRC22:34
*** radiofree has joined #baserock22:34
*** edcragg has quit IRC23:43
*** gtristan has joined #baserock23:43

Generated by 2.15.3 by Marius Gedminas - find it at!