Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-01-24

| Channels | #metacpan index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:32 klapperl joined #metacpan
04:14 chansen joined #metacpan
06:29 [Sno] joined #metacpan
09:14 dpetrov_ joined #metacpan
10:07 dpetrov_ joined #metacpan
11:17 ribasushi folks did we ever figure out a standard for github-linkable stuff in the changelog?
11:17 ribasushi akin to RT#xxx
11:28 ranguard Hmm, I thought we were going to use CPAN::Changes only, but looks like there are some modifications: https://github.com/CPAN-API/metacpan-web/blob/master/lib/MetaCPAN/Web/Model/API/Changes/Parser.pm - I'm guessing that's the place
11:28 Mithaldu i don't see anything specific about that in https://metacpan.org/pod/release/BRICAS/CPAN-Changes-0.27/lib/CPAN/Changes/Spec.pod
11:28 dipsy [ metacpan-web/lib/MetaCPAN/Web/Model/API/Changes/Parser.pm at master · CPAN-API/metacpan-web · GitHub ]
11:28 dipsy [ CPAN::Changes::Spec - Specification for CPAN Changes files - metacpan.org - Perl programming language ]
11:29 ribasushi Mithaldu: fuck Cpan::Changes, seriously
11:29 * ribasushi grumps on mention of that
11:29 Mithaldu man, what's your beef
11:29 ribasushi ranguard: omega was doing metacpan-only stuff couple months ago, which is what my question is based upon
11:30 ranguard ribasushi: then you know more than me
11:31 Mithaldu ribasushi: considering that i quite often see people do `git log >> Changes`, i'm happy about ANYTHING that yells at those
11:31 ribasushi Mithaldu: because it is a badly designed spec that keeps being rammed down CPAN by folks who both know and don't know better
11:31 Mithaldu what's bad about it?
11:31 ribasushi Mithaldu: and when a change goes from x => lesser x - it really drives me up the wall
11:31 Mithaldu uuuh
11:31 Mithaldu wat
11:32 bowtie_ joined #metacpan
11:32 ribasushi Mithaldu: the spec is incomplete, and focuses too much on machine readability without acknowledging that a changelog is first and foremost for humans to read
11:32 Mithaldu you dislike reverse chronological odering?
11:33 ribasushi Mithaldu: there is plenty said about this topic already, starting with the spec having no mention of how to do multiple level nesting, and how continuation lines work
11:33 ribasushi Mithaldu: so at this point I just go "shitty spec, please can we stop talking about it"
11:34 Mithaldu sounds like your main complaint is that it doesn't specify enough
11:34 Mithaldu rather than being actively bad in what it does specify
11:35 Mithaldu also, as mentioned, as long as people think git logs are fine change files it's the only thing that even remotely helps me
11:35 ribasushi it is bad in what it specifies as well - it is overly strict on the date line, with no reason, and turns the changelog into an aesthetic nightmare with the mandated []s
11:35 ribasushi but I could live with that if it was an actual spec on the stuff that mattered
11:36 ribasushi Mithaldu: I don't think git logs are fine ;)
11:36 Mithaldu will you go and yell at people if i show them to you? :P
11:37 ribasushi Mithaldu: depends on the size of the list
11:37 Mithaldu usually singular people
11:38 ribasushi fwiw this is what a changelog must look like (and it is obviously not C::C compliant): https://github.com/dbsrgits/dbix-class/compare/v0.08250...59d624cf78#diff-4
11:38 dipsy [ Comparing v0.08250...59d624cf78 · dbsrgits/dbix-class · GitHub ]
11:38 ribasushi s/must/should/
11:38 ribasushi meh fuck it s/should/must/ :D
11:38 Mithaldu err
11:38 Mithaldu i see a git log
11:38 Mithaldu wat
11:39 Mithaldu anyhow, here, enjoy this particularly useless ay of doing it: https://metacpan.org/changes/distribution/PDL
11:39 dipsy [ Changes - metacpan.org - Perl programming language ]
11:40 Mithaldu made all the more strange by the existence of this: https://metacpan.org/source/CHM/PDL-2.007/Release_Notes
11:40 dipsy [ Release_Notes - metacpan.org - Perl programming language ]
11:41 ribasushi that I will yell about sure
11:42 * ribasushi puts on top of "yell list"
11:42 Mithaldu haha
11:42 Mithaldu i think the last time i mentioned that to them the response was something along the lines of it being GNU-compliant
11:43 ribasushi rms is also GNU-compliant, doesn't make him a good idea
11:43 Mithaldu mhm
12:15 brilo joined #metacpan
13:10 brilo1 joined #metacpan
13:41 brilo joined #metacpan
13:59 bowtie_ joined #metacpan
14:01 oalders ribasushi: i don't know that we settled on a particular syntax for github issues in changes, but the regex is here https://github.com/CPAN-API/metacpan-web/blob/master/lib/MetaCPAN/Web/Model/API/Changes.pm#L102
14:01 dipsy [ metacpan-web/lib/MetaCPAN/Web/Model/API/Changes.pm at master · CPAN-API/metacpan-web · GitHub ]
14:06 bobtfish joined #metacpan
14:35 AirDisa joined #metacpan
14:56 brilo1 joined #metacpan
15:41 brilo joined #metacpan
15:46 ribasushi oalders: that's what I wanted to know, cheers
15:49 oalders :)
15:51 mattp_ joined #metacpan
15:54 fernando joined #metacpan
15:55 avenj joined #metacpan
17:15 danaj joined #metacpan
17:45 [Sno] joined #metacpan
18:00 AirDisa joined #metacpan
18:25 trs oalders ribasushi: hmm, I should update the rtcpan regex to match [rt.cpan.org #NNN]
18:47 oalders trs: sounds good
18:53 ranguard trs: that sort of thing I'd just do as a pull request - rather than waiting :)
18:58 maja joined #metacpan
19:00 trs is the reason the Changes aren't shown for this release because the date format isn't good enough? https://metacpan.org/release/App-RecordStream
19:00 dipsy [ App-RecordStream-4.0.2 - System for commandline analysis of data - metacpan.org - Perl programming language ]
19:02 jayallen joined #metacpan
19:13 AirDisa joined #metacpan
19:14 trs answer appears to be "yes"
19:15 trs I thought it was going to be a pain to find examples in the wild to test my regex, but then I remembered cpangrep: cpangrep file:Changes '\[?rt\.cpan\.org\s+\#(\d+)\]?'
19:30 oalders i did not know you had a command line tool for that :)
19:30 trs it's really handy :)
19:30 trs when I remember it
19:30 oalders :)
19:31 oalders output is nicely laid out
19:31 trs thanks.  I spent a while on it.
19:31 oalders thanks for the tip!
19:32 trs there are a few improvements I'd still like to make, but it generally does most of what I use grep.cpan.me for.
19:34 oalders it's quite fast
19:34 trs that's the backend, which I don't run but have submitted patches to :)
19:35 trs thank dg :)
19:40 oalders ah, it's the query caching
19:40 oalders i was trying some stuff that was really slow
19:40 oalders second time around it's quick
19:41 trs yeah, as it reads pages into memory, it keeps them around for further queries. but at some point someone else runs a query that pushes those pages out.
19:46 trs oalders: you may also find App::cpanchanges useful
19:49 oalders trs: another one i didn't know about. thanks :)
19:51 oalders trs: in App::cpanchanges Pod "App::cpangrep is simply a package placeholder for the included cpanchanges script"
19:51 oalders ?
19:58 trs oalders: whomp. :)
19:58 oalders copy/paste?
19:58 dipsy copy/paste is the only bit that's hard to get used to
19:58 trs I used cpangrep as a base for cpanchanges when I wrote it
19:58 trs yeah
19:59 trs they're separate
20:03 trs I'll have to fix that later tonight when I'm at home
20:13 AirDisa joined #metacpan
20:29 trs bah, I can't get WWW::Curl to install on OS X.
20:32 oalders apparently neither can i
20:41 brilo joined #metacpan
20:51 BinGOs Net::Curl can pretend to be WWW::Curl if you can get that installed.
20:57 oalders ah, good tip
21:12 trs for me, it appears to be an issue with the Makefile.PL parsing the cpp output of curl.h for the constants
21:14 AirDisa joined #metacpan
22:15 AirDisa joined #metacpan
22:41 trs oalders: well, after a break to interview someone, I fixed it with a patch to WWW::Curl: https://rt.cpan.org/Ticket/Display.html?id=92448
22:41 dipsy [ Bug #92448 for WWW-Curl: [PATCH] Parse enums generated by ISO-compliant cpp from curl.h ]
22:43 oalders trs++
22:45 daxim joined #metacpan
22:53 trs tangentially, let me state for the record that I hate build systems which modify version controlled files in place.
23:02 trs and finally, the thing that started this rabbit hole: https://github.com/CPAN-API/metacpan-web/pull/1044
23:02 dipsy [ Link more bug refs in changes by tsibley · Pull Request #1044 · CPAN-API/metacpan-web · GitHub ]
23:04 haarg trs: what build system?
23:08 trs haarg: the custom Makefile.PL for WWW-Curl
23:16 AirDisa joined #metacpan
23:26 nbezzala joined #metacpan

| Channels | #metacpan index | Today | | Search | Google Search | Plain-Text | summary