Camelia, the Perl 6 bug

IRC log for #metacpan, 2013-08-06

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

All times shown according to UTC.

Time Nick Message
01:03 shmuel joined #metacpan
01:37 cooper joined #metacpan
01:55 klapperl_ joined #metacpan
02:10 oalders omega: https://github.com/CPAN-AP​I/metacpan-web/issues/897
02:10 dipsy [ If Changes does not follow the CPAN spec the displayed bit ought to be <pre> formatted · Issue #897 · CPAN-API/metacpan-web · GitHub ]
02:11 ribasushi oalders: since you are awake
02:11 oalders i am!
02:12 ribasushi oalders: are there any plans to recognize/linkup qr/RT[\#\-\:]?(\d+)/ to the correct RT ticket?
02:12 ribasushi would be awesome
02:12 oalders ribasushi: in the Changes file?
02:12 ribasushi no, in the bit that gets displayed
02:12 ribasushi not in the entire changes file view
02:13 ribasushi oalders: like read through the bit I linked: https://metacpan.org/release/DBIx-Class
02:13 dipsy [ DBIx-Class-0.08250 - Extensible and flexible object <-> relational mapper. - metacpan.org ]
02:13 ribasushi oalders: I always include what RT's I am closing
02:13 ribasushi would be handy for a casual observer
02:13 ribasushi same can be done for GH#..., or somesuch
02:13 oalders well, there aren't any plans as this is the first i've heard of it.  but it's a very good idea
02:14 ribasushi oalders: I open an issue with this irclog then?
02:14 oalders ribasushi: please :)
02:14 oalders regarding the original issue, we're using CPAN::Changes to parse the log
02:15 omega oalders: uh oh
02:16 omega ahh, it is because CPAN::Changes is able to parse it
02:16 oalders yeah :)
02:16 omega but the DBIx::Class changelog uses unstandard groups
02:16 ribasushi omega: my dates are fine ;)
02:16 ribasushi precisely
02:16 ribasushi hence why I said "unparseable"
02:17 omega ok, I'll try to make it nicer
02:18 omega ribasushi: thanks for the bug report :)
02:18 ribasushi omega++
02:19 ribasushi omega: now, if you come up with a standard to how to link to tickets (i.e. what the regex above will be exactly) - this spec I will follow religiously ;)
02:19 oalders :D
02:19 * omega is so tempted to ask "why not follow the group spec as well", but I'll refrain, since the ticket forbids it :)
02:20 * ribasushi covers his bases well ;)
02:20 ribasushi hmmm
02:20 omega CPAN::Changes strips whitespace btw, so it isn't simply a <pre>
02:20 ribasushi ugh
02:20 ribasushi ffs
02:21 ribasushi omega: if it can't work easily - don't put too much effort into it, I will poke CPAN::Changes to stop being so nazi-like
02:21 ribasushi mo: why do you specifically want to avoid monospace?
02:21 ribasushi mo: I often do align things with a specific purpose
02:22 * ribasushi is confused where to talk here or on the ticket ;)
02:22 oalders ribasushi: omega is the one who implemented the Changes on the release page
02:23 oalders no need to browbeat mo ;)
02:23 ribasushi oalders: mo made a comment on the ticket
02:23 oalders ah, then fire away :)
02:23 haarg the spec doesn't seem to match the code very well
02:24 haarg the spec is 'changes is a number of lines that are indented'
02:25 haarg while the code is 'changes is a set of lines where deeper indentation means a line continuation'
02:30 haarg of course, the spec doesn't say much about how to parse files.  DBIx::Class arguably complies with the spec based on the limited amount it says about parsing rules.
02:31 oalders bricas: ^^^
02:37 ribasushi ok last issue for today, promise ;)
02:38 oalders :) issues are cheap
02:38 ribasushi making you guys sift through them is not
02:40 oalders well, it's all part of the process.  we just need an intern to do all the sifting...
03:14 grantm left #metacpan
03:14 grantm joined #metacpan
03:15 Mike-PerlRecruiter_ joined #metacpan
03:20 preflex_ joined #metacpan
03:30 metacpan joined #metacpan
03:30 metacpan [cpan-api01] oalders pushed 2 new commits to master: http://git.io/MTnIqw
03:30 metacpan cpan-api/master dab6cc9 Olaf Alders: Adds tidyall config and pre-commit setup.
03:30 metacpan cpan-api/master e004306 Olaf Alders: Renames perltidyrc.
03:30 metacpan left #metacpan
03:30 dipsy [ Comparing bc111fce9895...e004306c15d0 · CPAN-API/cpan-api · GitHub ]
03:31 omega Ok, pushed a PR to CPAN::Changes, and a branch to my metacpan-web fork
03:53 metacpan joined #metacpan
03:53 metacpan [cpan-api01] oalders force-pushed oalders/commonsense from 10ae232 to 531ece2: http://git.io/mvJiXQ
03:53 metacpan cpan-api/oalders/commonsense 2fbb4fa Olaf Alders: Allow for package name for pm.PL files to be in __DATA__.
03:53 metacpan cpan-api/oalders/commonsense 81bd5ff Olaf Alders: Adds a .pm.PL test.
03:53 metacpan cpan-api/oalders/commonsense 2f19129 Randy Stauner: Fix case on new module name in Makefile prereqs
03:53 metacpan left #metacpan
03:53 dipsy [ Commits · CPAN-API/cpan-api · GitHub ]
03:55 metacpan joined #metacpan
03:55 metacpan [cpan-api01] oalders merged oalders/commonsense into master: http://git.io/ghTOxQ
03:55 metacpan left #metacpan
03:55 dipsy [ Comparing e004306c15d0...531ece2ca22f · CPAN-API/cpan-api · GitHub ]
03:56 metacpan joined #metacpan
03:56 metacpan [cpan-api01] oalders deleted oalders/commonsense at 531ece2: http://git.io/PjoY8g
03:56 metacpan left #metacpan
03:56 dipsy [ Fixes undef warning in t/document/file.t. · 531ece2 · CPAN-API/cpan-api · GitHub ]
03:57 oalders omega++ # working on CPAN::Changes
03:58 grantm left #metacpan
03:58 grantm joined #metacpan
04:00 oalders https://metacpan.org/module/common::sense
04:00 dipsy [ common::sense - metacpan.org ]
04:01 omega oalders++ #fixing common::sense
04:01 oalders rwstauner++ # fixing oalders' tests
04:01 omega which judging by the number of issues opened about it, is much more important :)
04:01 oalders heh. well, i'm glad to have it sorted. it just wouldn't go away
04:01 rwstauner :-)
04:01 rwstauner awesome
04:02 rwstauner did you index all the backpan releases?
04:02 oalders i'm just indexing the releases manually
04:02 rwstauner ok
04:06 oalders the amazing thing is that nobody could find it, but it already had 4 ++
04:08 oalders the go to version links are broken, but that's an issue for another day
04:09 metacpan joined #metacpan
04:09 metacpan [metacpan-web01] oalders pushed 2 new commits to master: http://git.io/CiAuiA
04:09 metacpan metacpan-web/master 6f1f1df Andreas Marienborg: Add a vmethod that links RT(#:-)(\d+) to RT issues...
04:09 metacpan metacpan-web/master 72b6e8d Olaf Alders: Merge pull request #900 from omega/omega/link-issues-vmethod...
04:09 metacpan left #metacpan
04:09 dipsy [ Comparing 445a7460bcb5...72b6e8d157bb · CPAN-API/metacpan-web · GitHub ]
04:09 grantm Is that /module/common::sense link meant to work? I get a 404
04:10 oalders grantm: thanks -- fixed!
04:11 grantm oh that's better
04:12 oalders :)
04:12 oalders ribasushi: omega has your RT links working
04:12 oalders just deployed
04:13 omega I also sent a PR for teh #whatsnew bug :)
04:15 metacpan joined #metacpan
04:15 metacpan [metacpan-web01] oalders pushed 2 new commits to master: http://git.io/ruREEg
04:15 metacpan metacpan-web/master 1458025 Andreas Marienborg: Add section anchors to the release page...
04:15 metacpan metacpan-web/master c5625b7 Olaf Alders: Merge pull request #901 from omega/omega/section-anchors...
04:15 metacpan left #metacpan
04:15 dipsy [ Comparing 72b6e8d157bb...c5625b794005 · CPAN-API/metacpan-web · GitHub ]
04:15 omega oalders: I don't know if it is possible, but you could consider "hacking" in the pached CPAN::Changes as well perhaps?
04:16 ribasushi omega++ # fucking awesome
04:16 oalders omega: we aren't doing that for any other modules that i know of. we *could* do it, but we could just see about bugging bricas first :)
04:16 oalders omega++ # keeping ribasushi happy
04:17 omega oalders: ok :)
04:17 ribasushi omega, oalders: perhaps it would be a good idea to blog about this particular bit, ideally with links to the logic etc
04:17 ribasushi since there are many many ways to say RT X
04:17 oalders yep -- that would be helpful
04:18 ribasushi also I personally do not have ideas wrt github issues, as I don't use them much
04:18 ribasushi omega: on your other comment - what happens when the tracker changes
04:18 ribasushi it doesn't matter
04:19 ribasushi since you are rendering just the "latest changes" part
04:19 omega ahh, true :P
04:19 omega hehe
04:19 ribasushi it will naturally refer to... right ;)
04:19 oalders but you know the next request will be to parse the entire Changes file
04:19 oalders that is inevitable
04:20 ribasushi well... not from me ;)
04:20 oalders :)
04:20 omega Well, it wouldn't be _too_ hard to change the changes view from source to rendered with the same html / logic
04:20 omega or so I assume :P
04:20 oalders btw, omega, if you put something like "fixes #899" in your commit message, that will close the relevant issue when the code is merged in
04:20 omega ahh, ok :)
04:20 omega I'll try to remember :P
04:21 ribasushi but yeah...
04:21 oalders i say that having just forgotten to do it myself
04:21 ribasushi I'd say in this case requiring RT<sep><num> or GH<sep><num> is the cleanest approach
04:21 ribasushi since historic changelogs will be linked up properly
04:21 ribasushi and no heuristic will be involved
04:22 ribasushi also one could search for 'closes', 'resolves', etc in all indexed Changelogs
04:22 ribasushi to get an idea "what's out there"
04:23 ribasushi (followed by some digits naturally, to narrow stuff down)
04:23 oalders with github there is the possibility that the actual repo url changes
04:23 oalders but that's not something i'd be concerned about
04:23 ribasushi oalders: true, but there isn't much you can do about that
04:23 ribasushi well... you could
04:24 oalders but that's jumping through hoops
04:24 ribasushi after parsing the changelog you process each part separately ;)
04:24 ribasushi right ;)
04:24 oalders and it can be left as an exercise to the reader :)
04:24 ribasushi but yeah I'd start with just a dumb \d{5} search through all changelogs
04:24 ribasushi just to see what baseline to start with
04:24 oalders right
04:25 ribasushi but now I fall away &
04:25 omega hmm, wonder how to best pass the github repo url to the filter..
04:25 omega might not be possible?
04:31 oalders i'm not sure if it's possible
04:31 oalders but i need to shut it down for the evening.
04:31 oalders omega: thanks for all of your help
04:31 oalders &
06:15 dpetrov_ joined #metacpan
06:26 Mike-PerlRecruiter_ joined #metacpan
07:26 Mike-PerlRecruiter_ joined #metacpan
07:37 avar joined #metacpan
07:55 dolmen joined #metacpan
09:05 daxim joined #metacpan
14:51 mo joined #metacpan
15:58 bowtie_ joined #metacpan
18:30 preflex joined #metacpan
19:27 Mike-PerlRecruiter_ joined #metacpan
19:50 ether_ joined #metacpan
21:06 bowtie_ joined #metacpan

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