Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-07-08

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

All times shown according to UTC.

Time Nick Message
01:45 FROGGS_ joined #metacpan
01:47 klapperl_ joined #metacpan
02:18 talina_ joined #metacpan
03:04 castaway joined #metacpan
05:08 talina_ joined #metacpan
06:24 chankey joined #metacpan
06:36 neilb joined #metacpan
07:29 FROGGS_ joined #metacpan
07:58 chankeypathak joined #metacpan
07:59 oiami joined #metacpan
08:08 neilb joined #metacpan
09:06 oiami joined #metacpan
10:20 neilb If I search for "Pete Armstrong" on MetaCPAN, I get taken to a page for a release of Acme-MetaSyntactic-Themes-Abigail. Huh?!
10:21 * neilb suspects he's missing something...
10:21 omega https://metacpan.org/pod/Acme::MetaSyntactic::moonwalkers here?
10:22 dipsy [ Acme::MetaSyntactic::moonwalkers - People who have walked on the moon. - metacpan.org ]
10:22 * ranguard gets moon walkers as well
10:22 ranguard which has a 'pete' and an 'armstrong' on the page
10:22 omega is it something about if there is only one result, redirect to that page?
10:23 * ranguard guesses so
10:34 rashi joined #metacpan
10:38 neilb Aha
10:38 omega maybe it should have a notification or something saying "we found only this one result"
10:39 * ranguard thinks it just redirects so that'd be a bit tricky
10:40 omega it has a session cookie, so could perhaps set something there maybe
10:42 ranguard ick, but yea
10:42 * ranguard wants to really limit the cookies at some point, it's what's stopping use caching... well almost anything
10:42 neilb maybe that "one result let's redirect" should only fire in "obvious cases"? so if I search for Acme::Pink, I'd get redirected to the relevant pod, but not if it's just a list of words?
10:42 ranguard although maybe that should be via ESI or something
10:43 ranguard neilb: nice idea
10:43 neilb shall I add an issue on gh?
10:44 neilb or will you have cut the code by the time I submit it? ;-)
10:44 ranguard heh, issue
10:45 * ranguard hasn't actually done much on MC or the API, just the sysadmin/puppet stuff
10:45 ranguard someone will have to work out what 'obvious case' actually means in code
10:57 neilb BOGOFF - two issues for the price of one!
11:01 ranguard neilb: what ever happened to perlhub and your reviewing all templating systems? - still on the list ?
11:02 neilb gosh, is it that time already? time for lunch!
11:03 ranguard :)
11:03 neilb it is still on the list, but the list has grown spectacularly. I blame Yanick for my Yak-shaving tendencies.
11:04 ranguard you seem to have a whole yak hairdressers when you get going ;)
11:04 neilb cpanwhack: find amusing search terms that go to surprising pages on MetaCPAN
11:04 * ranguard thinks he's even seen pink ribbons on some of your yaks :)
11:04 neilb :-)
12:58 mst \o/
12:58 mst p3rl.org search box now using metacpan.org
12:58 mst Juerd sends his apologies for missing it when he converted everything else
13:15 ranguard woot
13:16 * ranguard wonders what else is left and how we convince google the point to MC
13:17 ranguard s/the/to/
13:17 ranguard the new servers should at least get us faster load times
14:27 ether should metacpan also search the 'keywords' metadata field?  is it possible to search metadata quickly? (does elasticsearch index it?)
14:28 ether https://metacpan.org/pod/CPAN::Meta::Spec#keywords
14:28 dipsy [ CPAN::Meta::Spec - specification for CPAN distribution metadata - metacpan.org ]
14:40 FROGGS joined #metacpan
14:45 oalders that info should be in release.metadata
14:46 neilb joined #metacpan
14:47 oalders but i don't see metadata in the release mapping
16:49 rwstauner oalders: metadata isn't in the mapping which means it's not searchable :'(
16:49 rwstauner ether: ^^
16:50 oalders so, it's there but it isn't ;)
16:50 oalders something for v1, i guess
16:52 ether want a ticket? :)
17:04 rashi joined #metacpan
17:04 oalders ether: sure :)
17:04 oalders actually, hang on
17:04 oalders there may already be one open
17:06 * ether hangs.
17:08 oalders ether: i think we do need an issue
17:08 oalders i was thinking of this https://github.com/CPAN-API/cpan-api/issues/145
17:08 dipsy [ Add contents of META file to ES · Issue #145 · CPAN-API/cpan-api · GitHub ]
17:08 oalders but that's already covered.  we need a new issue for adding some kind of a mapping for metadata
17:09 oalders i think we may need to figure out which parts of the metadata should be searchable
17:11 ether created: https://github.com/CPAN-API/cpan-api/issues/320
17:11 dipsy [ index distribution metadata in elasticsearch · Issue #320 · CPAN-API/cpan-api · GitHub ]
17:11 oalders ether++
17:12 oalders i'm not sure if we can take advantage of dynamic mapping here or if that just applies to an entire type
17:14 metacpan joined #metacpan
17:14 metacpan [metacpan-web] Talina06 pushed 1 new commit to talina/stargazer: http://git.io/BJjxDQ
17:14 metacpan metacpan-web/talina/stargazer 4455ced Talina06: Displaying the starred modules under the account tab.
17:14 metacpan left #metacpan
17:14 dipsy [ Displaying the starred modules under the account tab. · 4455ced · CPAN-API/metacpan-web · GitHub ]
17:18 metacpan joined #metacpan
17:18 metacpan [metacpan-web] Talina06 pushed 1 new commit to talina/stargazer: http://git.io/0ph4OQ
17:18 metacpan metacpan-web/talina/stargazer 60b1626 Talina06: Logic for stargazers display for $user.
17:18 metacpan left #metacpan
17:18 dipsy [ Logic for stargazers display for $user. · 60b1626 · CPAN-API/metacpan-web · GitHub ]
17:40 heytitle hi, i have an question about api.
17:43 heytitle I see POD, description and abstract are mapped to /module
17:43 heytitle But actually, I can't see them from http://api.metacpan.org/v0/module/_search?q=distribution:App-cpanminus&sort=date:desc&size=1
17:43 dipsy urgh. long url. Try http://tinyurl.com/msxpest
17:51 ranguard anyone know what's eating the CPU on mc?
17:52 ranguard http://munin.metacpan.org/metacpan.org/bm-n2.metacpan.org/elasticsearch_cache.html looks dodgy?
17:52 dipsy urgh. long url. Try http://tinyurl.com/pyzsc5p
17:52 dipsy [ Munin :: metacpan.org :: bm-n2.metacpan.org :: elasticsearch cache ]
17:53 ranguard oalders: ^^ ?
17:53 ranguard rwstauner: ^^ ?
17:55 oalders yeah, i guess ES is hogging the CPU
17:55 oalders it's coming down, though
17:59 oalders heytitle: if you just get the module directly, you'll get those fields http://explorer.metacpan.org/#/6c639ac32c96a55f54e0
17:59 dipsy [ MetaCPAN Explorer ]
17:59 oalders not sure why they're not in the search
17:59 oalders rwstauner?
17:59 dipsy rwstauner is too modest
17:59 oalders :)
18:00 heytitle another question, how can I search by ignoring case?
18:00 rwstauner thanks, dipsy
18:01 oalders heytitle: http://api.metacpan.org/v0/module/_mapping
18:01 oalders you'll see that some fields are marked with "include_in_all" : false,
18:01 rwstauner heytitle: your query doesn't have those fields because you've limited size to one and you happen to be getting the MANIFEST
18:02 rwstauner if you get more files, or add filters to actually get a module, you'll see those
18:03 oalders ah, there you go
18:03 rwstauner heytitle: try this, for instance: http://api.metacpan.org/v0/module/_search?q=distribution:App-cpanminus+AND+mime:text/x-script.perl-module&sort=date:desc&size=1
18:03 dipsy urgh. long url. Try http://tinyurl.com/me2ldax
18:04 oalders what's confusing about that search is that it's on /module, but it's returning the MANIFEST, which is not a module
18:04 oalders module is really just file
18:04 oalders in this case
18:05 rwstauner yeah, i always thought the module/file thing was a bit confusing
18:05 heytitle thanks
18:06 oalders i wonder we should just redirect to /file for these types of searches
18:06 oalders to make it obvious what is happening
18:07 rwstauner the "typical" search for a module includes filters like [ {"file.module.indexed": true}, {"file.module.authorized": true} ]
18:07 rwstauner yeah, something better could be done
18:08 rwstauner could just add appropriate filters when using /module
18:08 rwstauner like "exists": "file.modules"
18:08 rwstauner or something like that
18:08 oalders right
18:15 lifeofguenter joined #metacpan
19:07 metacpan joined #metacpan
19:07 metacpan [metacpan-web] Talina06 pushed 1 new commit to talina/stargazer: http://git.io/DN8AVw
19:07 metacpan metacpan-web/talina/stargazer 84f90b6 Talina06: Improve Styling.
19:07 metacpan left #metacpan
19:07 dipsy [ Improve Styling. · 84f90b6 · CPAN-API/metacpan-web · GitHub ]
19:27 neilb joined #metacpan
19:37 alex joined #metacpan
19:39 alex Hi all. I just found the well-hidden pause news from 2014-03-26 stating that the registered modules list is being retired. Does this mean that I don't have to register new namespaces anymore? Because I'm waiting for an approval for some time now.
19:39 alex well-hidden = here: http://pause.perl.org/pause/query?ACTION=pause_05news
19:39 dipsy [ PAUSE: pause_05news ]
19:48 mst alex: you never had to, and nobody's cared about namespace registrations for years
19:50 alex ah ok. I thought that would be something one should do
20:03 metacpan joined #metacpan
20:03 metacpan [metacpan-web] Talina06 pushed 1 new commit to talina/stargazer: http://git.io/2Omg5g
20:03 metacpan metacpan-web/talina/stargazer 1cc2e23 Talina06: Enable star button only for /pod/$module.
20:03 metacpan left #metacpan
20:03 dipsy [ Enable star button only for /pod/$module. · 1cc2e23 · CPAN-API/metacpan-web · GitHub ]
20:41 ether I've been working on a patch to remove module registration from the web ui, but the tuit was dropped and I haven't finished.
22:32 neilb joined #metacpan
22:55 lifeofguenter joined #metacpan

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