Camelia, the Perl 6 bug

IRC log for #metacpan, 2013-05-20

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

All times shown according to UTC.

Time Nick Message
00:06 mo SineSwiper: https://github.com/CPAN-API/cpan-api/blo​b/master/lib/MetaCPAN/Script/Pagerank.pm
00:06 dipsy [ cpan-api/lib/MetaCPAN/Script/Pagerank.pm at master · CPAN-API/cpan-api · GitHub ]
00:06 mo I wrote this a while ago and it calculates the pagerank for modules
00:06 mo probably told you about that already
00:06 mo https://gist.github.com/monken/1142048 # top 100
00:06 dipsy [ PageRank top 100 results ]
00:06 mo from 2 year ago :)
03:59 preflex_ joined #metacpan
06:17 ether_ joined #metacpan
06:33 bouncy joined #metacpan
06:35 Mike-PerlRecruiter_ joined #metacpan
06:36 bbb joined #metacpan
11:43 danaj joined #metacpan
12:09 bowtie joined #metacpan
12:28 SineSwiper mo: okay, so if this is in place with MetaCPAN already, we can probably improve upon it with my vitality score
12:29 SineSwiper I wanted to at least get the score on the left sidebar first, and then we can integrate it into the page rankings
12:39 SineSwiper mo: what controls this?  http://api.metacpan.org/v0/distribution/Moose
12:39 SineSwiper I'd like to get some more information into that bugs hash
12:50 SineSwiper mostly around oldest/newest dates, etc.
13:38 mo trs: ^^
13:38 mo SineSwiper: I would love to index all bugs in their entirety and not just the compressed view that rt provides
16:47 trs mo: define "entirety" ?
16:48 trs mo: there is a nightly trimmed down sqlite export of tickets from rt.cpan.org
16:50 trs it doesn't include transactions, but it does include created and updated: https://github.com/bestpractical/rt-extension-rt_c​pan_org/blob/master/bin/rt-cpan-export-db#L50-L58
16:50 dipsy [ rt-extension-rt_cpan_org/bin/rt-cpan-export-db at master · bestpractical/rt-extension-rt_cpan_org · GitHub ]
16:50 trs SineSwiper: that info comes from https://rt.cpan.org/Public/bugs-per-dist.tsv via MetaCPAN::Script::Tickets
16:53 trs the sqlite export I'm talking about is at https://rt.cpan.org/NoAuth/cpan/rtcpan.sqlite
16:53 trs the tsv is updated every hour, the sqlite export every night.
16:54 trs SineSwiper: not sure where you're located, but are you going to YAPC::NA by chance?  If so, I'm happy to talk rt.cpan there too. :)
18:35 Mike-PerlRecruiter_ joined #metacpan
20:00 thaljef joined #metacpan
20:23 przemek joined #metacpan
21:48 SineSwiper trs: I wish, but too new at my new job for vacation... definitely plan on going next year, though
21:48 trs nod
21:48 trs well, I'm happy to talk rt.cpan here as well :)
21:50 SineSwiper trs: so, do you know which part of MC grabs the existing RT data?
21:50 SineSwiper I like the idea of using that SQLite file
21:50 SineSwiper though, I would have to figure out a good analog for Github as well
21:50 SineSwiper (I'm sure they have some sort of API...)
21:51 trs SineSwiper: from 5 hours ago :) 09:50 < trs> SineSwiper: that info comes from  https://rt.cpan.org/Public/bugs-per-dist.tsv via  MetaCPAN::Script::Tickets
21:52 trs in the cpan-api repo
21:52 SineSwiper oh, sorry, didn't see the module name
21:53 trs the tsv is a more frequently updated data source, so metacpan probably wants to use the SQLite export in addition to  the tsv.
21:53 trs SQLite will get most recently created/updated ticket times for each queue.
21:54 SineSwiper trs: MC:S:T is getting this info real-time, right?
21:55 trs SineSwiper: real-time? no. it's a cronjob, maybe every 2 hours iirc.
21:56 trs SineSwiper: the tsv data feed from RT is generated every hour.
21:56 trs so metacpan could check it twice as often as it currently does if it was desired.
21:56 trs The query you'd want for created/updated against the sqlite: select distribution, max(created), max(updated) from ticket group by distribution order by distribution;
21:57 SineSwiper where's the cronjob for it?  I didn't see mention of it in /etc/cron.d/metacpan
21:57 trs SineSwiper: that file is out of date (last updated in 2011)
21:57 SineSwiper ahh, okay
21:57 trs I tripped over the same thing when I first started poking metacpan too :)
21:58 trs I believe you want to look at the metacpan-puppet repo for the cron config that most matches production.
21:58 trs modules/metacpan/manifests/cron.pp
21:59 SineSwiper cool, gotta pick up my son from daycare, bbl
21:59 trs bah, and I may be wrong then about frequency: that file claims `metacpan tickets` runs at 03:12 every day.
22:21 SineSwiper trs: heh, https://github.com/CPAN-API/cpan-api/blob/​master/lib/MetaCPAN/Script/Tickets.pm#L95
22:21 dipsy [ cpan-api/lib/MetaCPAN/Script/Tickets.pm at master · CPAN-API/cpan-api · GitHub ]
22:21 SineSwiper so, GH data is already in its raw form, we just don't use it
22:22 trs huh, I'd have thought that summary data was provided by GH's api
22:24 SineSwiper not quite sure how to expand it without providing some sort of new API destination, though
22:25 SineSwiper distribution->{bugs}{open_tickets} seems like it would be huge for certain distros
22:27 SineSwiper so, what's the best way of get some sort of dev env set up?
22:27 trs there's a wiki page on one of the cpan-api repos
22:27 trs detailing that
22:27 trs frankly, importing all bug info into metacpan seems not terribly useful and an unlimited scope.
22:28 SineSwiper yeah, my needs only extend to more summary information
22:28 SineSwiper specifically "Oldest open RT/GH ticket date (non-wistlist)"
22:29 trs nod. it'd be much simpler to just add the simple stats you need.
22:29 SineSwiper though, I will probably expand that to cover some other basic info
22:30 SineSwiper well, I was planning on transplanting my personal VPS, anyway
22:31 SineSwiper might as well get that up and running and then install a dev MetaCPAN
22:47 oalders SineSwiper: https://github.com/cpan-api/metacpan-developer
22:47 dipsy [ CPAN-API/metacpan-developer · GitHub ]
22:48 oalders that will get you a VM with everything installed
22:48 oalders you can then work off your local checkout and restart the services in the VM to test them
22:48 genehack i'm happy to try to patch https://github.com/CPAN-API/cpan-api/issues/268 if somebody can point me in the right direction
22:48 dipsy [ /pod endpoint incorrectly fails to return =encoding · Issue #268 · CPAN-API/cpan-api · GitHub ]
22:49 oalders rwstauner: ^^
22:51 genehack "incorrectly fails", ugh. somebody smack that bug reporter.
22:52 trs genehack: you want the extract_pod method in MetaCPAN::Server::View::Pod
22:54 trs which I think means you get to chase the bug over to Pod::POM
22:54 trs :)
22:54 genehack \o/
22:54 genehack i started with a bug against utf8::all, managed to avoid filing one against cpandoc ...
22:54 * genehack continues on down the rabbit hole...
22:55 trs (... aaaand the way Pod::POM::View::Pod works looks fragile)
22:55 trs genehack: have fun
22:55 genehack and i don't see a github repo. 8^p
22:56 trs genehack: https://rt.cpan.org/Dist/Display.html?Name=Pod-POM but your bug is reported at least twice already
22:56 dipsy [ Active bugs for Pod-POM ]
22:56 trs perhaps a third one would help :P
22:57 genehack *sigh* there's an open ticket with a patch from 2009.
22:57 * genehack rages against ALL the things
23:02 * genehack has emailed the current maintainer about the best way to send patches and whether he'd be open to giving me a maint bit.
23:02 genehack we'll see how that goes.
23:04 genehack trs: thanks for the help
23:05 trs genehack: no problem. maybe you just want to write Pod::Simple::Pod
23:06 genehack wow the first metacpan hit on that is bonkers
23:08 rwstauner genehack: indeed it is
23:08 * rwstauner wonders why we're using Pod::POM at all
23:08 rwstauner genehack: i forget why that is (about Pod::Simple search result) but i'm pretty sure there are tickets for it
23:08 genehack well, i just looked at Pod:Simple::Text and my head just exploded
23:09 rwstauner rats, there goes another one
23:09 rwstauner i admit, i've submitted some patches for Pod::Simple, and i'm not entirely sure the name fits
23:10 trs genehack: https://metacpan.org/module/Pod​::Simple::Subclassing#SYNOPSIS
23:10 dipsy [ Pod::Simple::Subclassing - write a formatter as a Pod::Simple subclass - metacpan.org ]
23:10 trs it's easier if you don't use the ::Methody subclass.
23:10 rwstauner but i comaint a module based on Pod::Parser and that has it's own issues
23:10 genehack shit, this is starting to sound like work. 8^p
23:10 trs in my POD munging tasks, Pod::Simple's been far nicer than alternatives, even when I have to write patches for it.
23:10 trs genehack: ayup
23:11 trs you can probably get away with just three subclassed methods per that synopsis
23:11 trs for a pod -> pod converter
23:15 genehack that would certainly be a better way to go, given that there's already a Pod::Simple dependency
23:16 genehack (as MetaCPAN::Pod::XHTML subclasses it)
23:16 trs sounds like your work's cut out for you :)
23:17 trs hope you weren't planning on doing anything else today
23:19 genehack heh
23:19 genehack well, i at least made an issue for it
23:19 genehack but no, i'm _supposed_ to be making slides for the training i'm doing at YAPC::NA
23:19 genehack so i should return to that. 8^)
23:44 mo rwstauner: because Pod::POM is  much faster iirc
23:46 mo doesn't matter for the endpoint, but it does when indexing, and we need it to build the full text index
23:53 trs mo: nope, Pod::Text is used, which isa Pod::Simple
23:53 mo ah cool

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