Fire is step THREE! | github.com/perl6/toolchain-bikeshed | Channel logs: irclog.perlgeek.de/perl6-toolchain/today | useful prior art: metacpan.org/pod/CPAN::Meta::Spec
Set by moderator on 25 January 2016.
06:59 domidumont joined 07:04 domidumont joined 07:28 FROGGS joined 09:11 leont joined 10:02 hankache joined
ranguard jdv79: github.com/perlorg/ - github.com/perlorg/cpanorg - or github.com/andk/pause depending what you are after :) 10:16
jdv79 ranguard: i dont see it. the code that gen'd the dists for psixdists. 12:32
ranguard you're an owner! - so it's on your drop down in github.com/ 13:23
github.com/perl6modules/perl6-module-uploader 13:24
there's a bug I need to fix (where someone added meta6.json to their .gitignore) so I need to fix that and run again soon
jdv79 oh, yeah. was looking at the wrong repo. thanks. 15:06
ugexe jdv79: let me know when there is a dist with 2 different versions up on your metacpan 16:24
16:42 domidumont joined
jdv79 ugexe: :) 17:47
i will 17:48
18:17 Kassandry joined 18:55 FROGGS joined
tony-o_ curiously strong 19:27
jdv79 altoids? 19:28
19:29 perlpilot joined
tony-o_ the mints' mint. 19:31
jdv79 ugexe: paste.scsys.co.uk/504795 20:08
any ideas?
ugexe jdv79: if you run it enough it eventually doesn't segfault. i think its fixed in the `dev` branch. fwiw i think it was/is caused by the run($un-tar command) and reading/closing its .out handle 20:37
jdv79 so its a rakudo bug?
ugexe im not sure 20:38
jdv79 i seem to remember those type of bugs being discussed a long time ago
ugexe it seemed to be caused by using the output from one `run` as part of the input to another in combination with binding/lazyness 20:42
jdv79 well, dev hits p6c 20:43
so that's not what i'm looking for 20:44
ugexe you can change it in the config. under p6c add "disabled" : 1 (same for ::LocalCache), and "disabled" : 0 for cpan
jdv79 paste.scsys.co.uk/504798 20:45
ok. curious how it gets it from p6c but then tries cpan. or it looks like that from the log.
ugexe the mention of cpan is from the Plugin debug messages. its just saying its going to check if it can use that plugin, and the mentions its disabled 20:46
jdv79 ok. 20:47
ugexe if i can ever figure out a nice way to pass command line options plugins you'll be able to disable/enable them that way instead of having to edit the config file 20:48
jdv79 disabled has a negative connotation
;) 20:49
ah, it seems to work with those config changes. cool.
ugexe im open to any such suggestions
jdv79 enabled 20:50
ugexe default to enabled:1, and require enabled:0 to disable? 20:51
jdv79 idk how metacpan is going to handle the auth/ver/api stuff yet
hmm...
ugexe yeah, i think im going to have to do a final search of the metacpan results to make sure everything matches (or pick items if a range is used) 20:52
but it seemed to work for exact matching values, and with * wildcard 20:53
i.e `install github:tony-o:CSV-Parser:0.1.*` might already just work 20:54
jdv79 disabled:0 reminds me of stuff like $no-whatever = 0 which can unnecessarily confuse people because of the double sense.
ok 20:55
it is a bit nitpicky though. whatever:)
FROGGS I agree fwiw
ugexe yeah i agree, i hate `no-` prefixes. someone asked yesterday why i use --/test instead of --no-test
flussence has seen at least one program with a --no-foo and a --no-no-foo somewhere 20:58
ugexe jdv79: do you know if the metacpan api has a way to get a requested distribution's entire dependency chain? that'd be nicer than making multiple queries to build it manually... 21:16
jdv79 ugexe: not sure about that 21:26
that could be neat
i'm guessing no cause not all deps are static, right?
was there any movement on that actually? dyn deps i mean. 21:27
ugexe right, but i thought maybe for static deps
jdv79 hmm...
ugexe they have dependency graphs and reverse dependencies on a modules page, so the functionality seems like its there (although maybe not part of the api) 21:29
jdv79 maybe we'll have to add it. i'll make a note to look into it.