IRC logs for #baserock for Thursday, 2016-06-30

*** locallycompact has joined #baserock01:21
*** locallycompact has quit IRC01:26
*** gtristan has joined #baserock04:49
*** gtristan has quit IRC06:21
*** edcragg has joined #baserock06:35
*** gtristan has joined #baserock06:48
*** edcragg has quit IRC07:21
*** ctbruce has joined #baserock07:24
*** fay_ has joined #baserock07:29
*** rdale has quit IRC07:40
*** CTtpollard has joined #baserock07:52
*** toscalix has joined #baserock07:57
*** rdale has joined #baserock08:10
*** tiagogomes has joined #baserock08:10
*** jonathanmaw has joined #baserock08:19
*** franred has quit IRC08:32
*** franred has joined #baserock08:36
*** janderJLR_ has joined #baserock09:01
*** paulsherwood has joined #baserock09:01
*** tiagogomes_ has joined #baserock09:04
*** perryl_ has joined #baserock09:04
*** tiagogomes has quit IRC09:08
*** janderJLR has quit IRC09:08
*** leeming has quit IRC09:08
*** paulsher1ood has quit IRC09:08
*** perryl has quit IRC09:08
*** leeming has joined #baserock09:09
*** perryl_ is now known as perryl09:09
*** edcragg has joined #baserock09:10
*** locallycompact has joined #baserock09:27
*** jonathanmaw has quit IRC09:52
*** jonathanmaw has joined #baserock10:01
*** mario-goulart has quit IRC10:43
*** mario-goulart has joined #baserock10:44
leemingmaybe out of scope(?) but for anyone who has used the yaml python libs.... when doing a yaml.dump() to create a yaml/morph file.... how do you get around the key ordering?10:44
leemingive tried what is suggested here >   , but doesn't seem to work all the time :S10:44
edcraggdon't people usually use safe_dump or whatever it is10:45
leemingor am i being too anal wanting 'name' and 'kind' at the top10:45
* leeming goes to look10:45
SotKleeming: morph subclasses yaml.SafeDumper to dump in a defined order:
pedroalvarezalmost snap ...10:47
pedroalvarezwhat SotK said :)10:48
leemingoooo already python world morph code... doh should have looked10:48
pedroalvarezand no, having them in order makes sense. Please, do that.10:48
pedroalvarezlots of things are already implemented in morph, yes10:49
leemingthanks SotK10:49
pedroalvarezpity that nobody wants to use that library anymore :P10:49
* leeming does10:49
* leeming maybe10:49
pedroalvarezI guess you will end up just copying the code10:49
leemingor importing the lib....?10:50
leemingor is it currently in a weird quantum state of working and not working?10:50
pedroalvarezno no, it works10:51
leemingwell, as long as it is portable enough, hopefully shouldnt have to _copy_ the code10:52
leemingwhile im on the topic... any other libs/repos that might be of use to me while developing this spec2definitions tool?10:52
*** cyndis_ is now known as cyndis10:53
pedroalvarezhm.. maybe the import tool10:54
franredit would be nice if the spec2definitions will be a plugin in the import tool, yep10:55
pedroalvarezthe import tool takes [npm, python,cpan....] packages and create definitions for them10:55
pedroalvareznot saying this is the way, if you only want to make something quick that works, but just pointing out that we have something like that already:
*** gtristan has quit IRC11:01
leemingfranred, after you asked about %patch, i was asked to look into one on my project that has them in the spec... so might be on the cards11:09
leemingpedroalvarez, cool thanks.. yes, I am aware of the effort to make a "tool to fix THIS problem" vs generalising/proper(ish) software engineering the tool11:11
leemingjust like to know the options before making some arbitrary design decision that ends up not being compatible with merging with general tools11:11
leemingJust wondering, in the import tools... how does it figure out the git information? if no one knows off the top of their heads it is fine, just starting to scan through the code now.... is it a lorry thing?11:33
pedroalvarezit tries to generate lorry files with the information it can get from npm,pip,etc package managers11:35
leemingis git information contained within them package managers?11:37
leemingor is it created from11:37
pedroalvarezI have no idea, tbh11:37
leeming[software bundle] --> Create git repo--> git, instead of, [software bundle] --> Magically git searching --> git-mirror on trove/whereever11:38
pedroalvarez[software bundle] --> magically *source* checking --> generation lorry file that you can push to a trove11:40
* leeming nods11:43
leemingfeel free to laugh but, are there any api docs for this stuff?11:51
pedroalvarezwhat do you mean by api docs11:51
pedroalvarezthere are some readmes in the repo, not sure what's on them :P11:53
pedroalvarezand some tutorials:
leeminghow come some things are in docs.baserock, yet others in wiki.baserock.. ahh11:55
leemingbut thanks.. but not quite what i was after11:56
leemingi was seeing if there was docs for a developer, telling me what classes/functions exist , with descriptions/purpose/args11:57
franredleeming, good to know that was helpful :)11:57
pedroalvarezi think focuses on the Definitions format, everything else, like tutorials, and more information is on the wiki11:57
leemingi know a lot of projects dont have them, due to time/effort/fast moving11:57
leemingbut nice to have =)11:58
leemingwiki/docs, is just another example of fragmentation that is currently something that keeps cropping up for me =) dont worry about my outbursts haha11:58
persiaPart of wiki/docs is that the wiki was relatively ungardened, although I don't know that we ever completed enough of docs to retire the wiki.11:59
* leeming just never quite knows where to look12:00
persiaFor developer docs, my default is to add sphinx to things and stick them on readthedocs, as they have good google-fu, so things come up faster than if one just renders development docs locally and adds it to a project site.12:00
leemingurm ok you mention tools i have no idea what they are12:00
persiaI don't believe that is done for any of the baserock projects currently, nor that we have a local renderer, making it tricky.12:00
* leeming stacks them on my todo google searches12:00
leemingim familiar with doxygen, but guess that is more in the java ecosystem12:01
persiaMind you, I tend to prefer Markdown to ReStructured Text, but ...12:02
leemingwell... using documenting macro langs like doxy gets around that right? creates multiple renders in diff formats12:03
SotKsphinx is worth using ReStructured Text for imo12:03
leemingok cool, will look into that12:04
leemingmost of my python experience is very hack/slash, so not fully used to the dev workflow for it12:04
* leeming keeps asking about documentation in CT.. is worried he is getting a name for himself12:05
*** locallycompact has quit IRC12:26
*** locallycompact has joined #baserock12:31
*** CTtpollard has quit IRC12:47
*** CTtpollard has joined #baserock12:47
*** tiagogomes_ has quit IRC12:51
*** tiagogomes_ has joined #baserock13:45
*** franred has quit IRC14:01
*** franred has joined #baserock14:02
* paulsherwood wonders how morph deals with assembling a system containing s-i commands, where the executables for the s-i commands themselves are not present on the assembled system14:38
richard_mawit doesn't, s-i commands were intended to be stuff like ldconfig, which requires you to be chrooted14:39
*** gtristan has joined #baserock14:42
paulsherwoodlocallycompact: there's your answer... your system is impossible :)14:49
paulsherwood(from another channel)14:49
locallycompactIs it14:49
paulsherwoodlocallycompact: can you share the actual system-integration commands from what you're doing? i wonder what components are actually necessary14:51
paulsherwoodmaybe you could define an extra (empty or nearly empty) stratum, with b-d that includes the core+foundation etc14:52
paulsherwoodthat would avoid redoing cache-keys for strata you already have14:52
paulsherwoodlocallycompact: does that make sense? ^^14:54
locallycompactnone of this makes sense honestly14:55
paulsherwoodreally? what's the confusion?14:56
locallycompactSo, the system integration command is being run at system build time but defined in the chunk14:58
paulsherwoodyes, correct14:58
locallycompactSo that one that failed is the linux one14:58
locallycompactWhy does it fail here14:58
* paulsherwood had no hand the semantics of this14:58
paulsherwoodiiuc (but i may be wrong) as gtristan expressed 'Conclusion, we cannot run these commands in the sandbox when the sandbox does not contain a shell.'15:00
persiaIf you need to do that sort of thing, the deployment commands are a sensible place...15:01
paulsherwoodgood point15:02
paulsherwoodybd could do with a better message for this situation15:02
*** Jude has joined #baserock15:03
*** toscalix has quit IRC16:10
*** CTtpollard has quit IRC16:13
*** toscalix has joined #baserock16:16
*** ctbruce has quit IRC16:18
*** jonathanmaw has quit IRC16:29
*** edcragg has quit IRC16:51
*** toscalix has quit IRC17:24
*** locallycompact has quit IRC17:25
*** edcragg has joined #baserock17:36
*** edcragg has quit IRC17:46
*** gtristan has quit IRC21:45
*** paulsherwood has quit IRC21:45
*** rdale has quit IRC21:45
*** Jude has quit IRC21:45
*** vgrade has quit IRC21:45

Generated by 2.15.3 by Marius Gedminas - find it at!