Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2016-06-20

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

All times shown according to UTC.

Time Nick Message
01:47 ilbot2 joined #metacpan
01:47 Topic for #metacpan is now V1 READY FOR TESTING https://github.com/CPAN-API/cpan-api/wiki/MetaCPAN-V1-announcement-and-migration-plan | MetaCPAN Developer VM https://github.com/CPAN-API/metacpan-developer | Chat logs available at http://irclog.perlgeek.de/metacpan/ | Can't find your module on MetaCPAN? https://metacpan.org/about/missing_modules
02:09 vroom joined #metacpan
02:11 Khisanth joined #metacpan
02:16 oalders ranguard: you can also just add real dists to the darkpan that the api tests use
02:16 vroom joined #metacpan
02:21 oalders ranguard: also, scripts/endpoints/release in metacpan-examples has some good starting points for queries
02:26 vroom joined #metacpan
02:28 vroom joined #metacpan
02:30 vroom joined #metacpan
02:34 vroom joined #metacpan
02:39 vroom joined #metacpan
02:56 vroom joined #metacpan
03:10 vroom joined #metacpan
03:52 vroom joined #metacpan
06:23 oiami joined #metacpan
07:54 neilb joined #metacpan
08:25 Relequestual joined #metacpan
08:29 neilb joined #metacpan
08:45 Tux joined #metacpan
09:00 vroom joined #metacpan
09:12 vroom joined #metacpan
09:30 ranguard oalders: cheers, but I want to import several hundred modules so though full dists might be too much
09:31 ranguard I eventually worked out that 'release' is what I wanted
09:31 ranguard I've created another example in a branch, but haven't tested it yet
09:54 vroom joined #metacpan
10:47 vroom joined #metacpan
11:21 oiami joined #metacpan
11:32 vroom joined #metacpan
11:45 vroom joined #metacpan
11:53 vroom joined #metacpan
12:06 vroom joined #metacpan
12:32 Tempesta joined #metacpan
13:02 vroom joined #metacpan
13:06 oalders ranguard: up to you how you want to handle it, but we cache the downloaded files
13:06 oalders the big hit is just indexing everything before a test run
13:07 oalders but my thinking was that it's easier to just add URLs for dists that we want to test than to have to fake them
13:07 oalders also seems less error prone
13:12 ranguard oalders: bah, where were you all those hours ago when I started this :)
13:12 oalders at the beach ;)
13:12 ranguard though actually just getting the URL's to download was the harder bit
13:12 ranguard oalders: !@£()Q&*SP@£QSW :)
13:12 ranguard nice
13:13 oalders you mean writing a script to extract the download urls?
13:13 ranguard I've written a script you pass it something like 'Data-Page' and it'll download every version
13:13 ranguard this is with a view to being able to create test cases for the download_url end point
13:14 oalders nice.  i saw some activity there.  that'll be an important endpoint to test very well
13:14 ranguard so if someone reports an issue we can just update our test suite to include ALL versions of that
13:15 ranguard module into our test index
13:15 oalders good idea
13:46 vroom joined #metacpan
13:46 vroom left #metacpan
13:57 ranguard boo, I wasn't in the MC table photo for YAPC::NA :(
13:57 jberger oh but I was!
13:58 jberger hehehe
14:00 oalders ranguard: i guess that pic must have been from after you left
14:00 oalders mickey will be speaking shortly https://www.youtube.com/watch?v=73e6quzJU9U
14:01 jberger yay, mickey!
14:01 oalders now i know how to pronounce his last name
14:01 ranguard there he is... setting up... and the crowd gets ready to go wild!
14:01 ranguard doh, more screen problems!
14:02 jberger happens to the best of them
14:02 alh Technology is awful
14:02 ranguard it all sucks :)
14:02 * jberger goes back to chiseling tablets
14:02 oalders :)
14:02 jberger not sure my wife is going to like all these grooves on her tablet though
14:10 oalders for real this time
14:12 rblackwe joined #metacpan
14:14 oalders mickey++
14:16 * ranguard waits for the hundreds of vounteers to flood in :)
14:17 kivanc joined #metacpan
14:17 mickey heh, no speaker notes on Sawyer's computer... damn :)
14:18 jberger mickey++
14:18 * mickey hopes it wasn't that bad
14:18 genehack mickey++ # dealing with that type of situation always sucks
14:18 genehack ramps the anxiety up something fierce
14:19 genehack you did great!
14:19 ranguard mickey++
14:19 mickey thanks
14:21 kyzn joined #metacpan
14:25 oalders mickey: your talk was much better than what i would have come up with. you touched on all of the right points.
14:59 dylan joined #metacpan
16:12 kivanc joined #metacpan
16:14 kivanc joined #metacpan
16:58 kyzn joined #metacpan
16:59 kyzn joined #metacpan
17:02 kyzn joined #metacpan
17:08 kyzn joined #metacpan
17:16 neilb joined #metacpan
17:23 kivanc joined #metacpan
17:23 kyzn joined #metacpan
17:33 kyzn joined #metacpan
18:24 vanstyn_ joined #metacpan
18:28 ilmari the "show changes for all dev releases since the last stable release" only seems to work for dists using underscores in the version number, not for ones using -TRIAL
18:29 ilmari compare https://metacpan.org/release/BINGOS/ExtUtils-MakeMaker-7.18 vs. https://metacpan.org/release/EXODIST/Test-Simple-1.302030
18:31 ilmari might want to add a check for (TRIAL RELEASE) on the version line as well as underscores in the verison number?
18:31 ilmari https://metacpan.org/changes/distribution/Test-Simple
18:36 ilmari e.g. check $release->note =~ / \( \s* trial \s+ release \s* \) /x
18:58 ether I believe that was on haarg's tuit list
19:11 ilmari oh, it does on git master, but not on v1.metacpan.org
19:12 ilmari https://github.com/CPAN-API/metacpan-web/commit/01f92167588a4f6075f936aa895e7a9a7042fb2b
19:12 ilmari haarg++ # beat me to it
19:13 ilmari huh, that uses direct hash access to the release objects...
19:15 ilmari oh, it doesn't actually use a CPAN::Changes object
19:55 metacpan joined #metacpan
19:55 metacpan [metacpan-web] oalders pushed 1 new commit to master: https://git.io/vorsP
19:55 metacpan metacpan-web/master 959fdea Olaf Alders: Merge pull request #1733 from CPAN-API/topic/permalinks...
19:55 metacpan left #metacpan
19:55 metacpan joined #metacpan
19:55 metacpan [metacpan-web] oalders deleted topic/permalinks at e48aa49: https://git.io/vorsX
19:55 metacpan left #metacpan
19:56 oalders haarg++ # deploying that now
19:59 ranguard oalders: where in the api test code does it import these full .tar.gz files :) ?
20:00 oalders ranguard: t/lib/MetaCPAN/DarkPAN.pm
20:02 ranguard oalders: cheers
20:14 ranguard oalders: so I've done https://gist.github.com/ranguard/b877a33dde9c2fed74ae3b51a700d687 - but feel like I'm slightly floundering in how to make it useful
20:15 ranguard mickey: ^^ not sure of any of that is useful
20:16 ranguard there is a bug in it doesn't seem to work for TRIAL releases, but then feel like that might be going beyond what we need
20:17 ranguard might be it should be thrown out, or split into useful bits *shrug*
20:18 oalders i guess you could have used the API to get the file names rather than downloading and unpacking
20:19 oalders ranguard: you're asking how this data can be used to test download_url>
20:19 oalders ?
20:22 ranguard oalders: yea, was trying to make something so adding test data was easy
20:22 ranguard then when we find edge cases we can add them
20:24 oalders ranguard: have you tried running t/00_setup.t after creating the files?
20:24 oalders you should get some verbose output to let you know they're being indexed
20:25 oalders after that file is run, you should be able to run any other test in the suite
20:25 ranguard oalders: yea, I mean, I guess I'm asking is that code worth me spending more time on
20:35 oalders ranguard: good question
20:36 oalders probably just adding modules to the DarkPAN is not going to scale well
20:37 oalders but i like the DarkPAN for testing cases where there are issues with the contents of archive that we want to test
20:38 oalders i haven't added new fake CPAN dists because it wasn't clear to me how to easily go about doing that
20:38 metacpan joined #metacpan
20:38 metacpan [cpan-api] mickeyn pushed 1 new commit to master: https://git.io/vorl0
20:38 metacpan cpan-api/master 8469888 Mickey: Merge pull request #497 from CPAN-API/topic/permalinks...
20:38 metacpan left #metacpan
20:38 ranguard yea, I built this so you got all the files, but didn't have to have the actualc ontent
20:38 oalders but it looks like you're on the way to solving that
20:38 ranguard oalders: ok, that's enough to make me carry on - cheers
20:38 oalders :)
20:40 mickey ranguard: I hope this is just cache: after the merge & deploy of #1729, https://v1.metacpan.org/recent should show relative links for pagination, it does if you select specific host but not from this link
20:41 mickey I tried the purge line you posted the last time but it didn't work for me for this URL
20:47 ranguard mickey: run: curl -svo /dev/null https://v1.metacpan.org/recent
20:47 ranguard look for 'Age' and 'X-Cache' headers
20:51 mickey that's actually showing cache miss
20:52 ranguard so it's not fastly
20:52 mickey k, i'll dig in :)
20:59 * ranguard HATES versions.. and on that note I'm calling it a night
21:17 mickey hmm... we still have too many hardcoded references to https://metacpan.org
21:18 mickey but i'm not sure this is the problem here, the links come from req.uri_with and i can't reproduce this
21:19 mickey it works fine on my vm
21:34 oalders this seems like something mst could help with?
22:02 haarg ilmari: metacpan uses a completely rewritten changelog parser.
22:03 ilmari haarg: yeah, I noticed
22:03 haarg i want to get it integrated to CPAN::Changes, but backcompat makes it tricky
22:03 haarg i have it mostly does but i keep having other things take priority
23:16 metacpan joined #metacpan
23:16 metacpan [metacpan-web] haarg created haarg/changes-padding (+1 new commit): https://git.io/vorMj
23:16 metacpan metacpan-web/haarg/changes-padding 7553376 Graham Knop: improve sizing and padding in changes box
23:16 metacpan left #metacpan

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