Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2016-08-30

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

All times shown according to UTC.

Time Nick Message
05:41 [Tux] ranguard, I run it every 15 minutes purely for myself, indeed resulting in the page you mentioned
05:42 [Tux] there you also see a link "other dashboards", where I have run that script manually for a bunch of other id's
05:58 ohoushyar joined #metacpan
06:22 oiami joined #metacpan
07:07 ilmari joined #metacpan
08:17 edward joined #metacpan
08:18 ranguard mickey: can we make MC::Client depend on IO::Socket::SSL 1.42 and/or Net::SSLeay 1.49 - seeing as those are needed for SSL with HTTP::Tiny
08:19 ranguard oh, looks like that should be an AND
08:23 uree joined #metacpan
08:40 ranguard ok, cron setup to run every 2 mins, lets see what happens
09:39 osfabibisi joined #metacpan
09:43 mickey ranguard: sure, did you just volunteered to do it? :)
09:44 mickey no worries, i'll look into it
09:53 mickey ok, we probably should set to a newer HTTP::Tiny version (at least from the one i got installed)
10:01 metacpan joined #metacpan
10:01 metacpan [metacpan-client] mickeyn pushed 1 new commit to master: https://git.io/vifnd
10:01 metacpan metacpan-client/master 75a25bd Mickey Nasriachi: some version restrictions for SSL over HTTP::Tiny
10:01 metacpan left #metacpan
10:03 mickey ranguard: https://v1.metacpan.org/release/MICKEY/MetaCPAN-Client-1.025000-TRIAL
10:03 mickey Tux: ^^
10:03 mickey can you please give it a try and let me know if it helps
10:03 mickey ?
10:04 * Tux pulls
10:05 Tux git related comment: after you build, it modifies cpanfile, which is (also) in git and thus breaks successive pull's
10:05 Tux error: Your local changes to the following files would be overwritten by merge:
10:05 Tux cpanfile
10:05 Tux Please commit your changes or stash them before you merge.
10:05 mickey which local changes did you make ?
10:05 Tux none
10:05 Tux $ dzil test
10:06 Tux $ dzil install
10:06 mickey then just check it out
10:06 Tux It is now installed. The next run will be in 9 minutes I think
10:07 mickey cool, let's hope for the best :)
10:55 Tux No mail so far
10:56 Tux and I've had at least three runs now
10:58 mickey nice, i'll give it a couple more hours before making it into a proper release
11:04 Tux does M::C also return cpantesters results?
11:05 Tux my results are currently not as well show as it could be
11:05 Tux e.g. with Text::CSV_XS it shows 0 / 0 / 0 where it should be 214 / 0 / 0
11:06 Tux and it looks like I get a timeout on my current implementation of finding the number of downriver deps
11:06 Tux would be nice to get that through M::C too
11:10 mickey Tux: we have river data is in the 'distribution' index
11:10 mickey MetaCPAN::Client->new( version => "v1" )->distribution("Text-CSV_XS")->river
11:11 Tux {   bucket           => 3,
11:11 Tux immediate        => 75,
11:11 Tux total            => 119
11:11 Tux }
11:11 mickey cpantesters i don't think we have, but i don't mind having the client fetch that for you... can you please open an issue for it on GH?
11:12 Tux yep, will do :)
11:21 Tux Failed to fetch 'https://fastapi.metacpan.org/v1/distribution/Release-Checklist': Not Found at /pro/lib/perl5/site_perl/5.22.0/MetaCPAN/Client.pm line 206.
11:21 Tux Hmmm, that should probably not fail but return something empty
11:31 Tux mickey++; # river-data
11:31 Tux https://github.com/metacpan/metacpan-client/issues/45
11:37 mickey i'm guessing the Failed is from the ua because the api returned 404
11:38 Tux new dashboards now running for the "other" authors
11:38 Tux looks like not all dists have a ->river
11:39 Tux my $rd = $mcpd ? $mcpd->river->{total} : undef;
11:39 Tux caused a Can't use an undefined value as a HASH reference at ./makewww.pl line 344.
11:39 mickey yeah, not all of them do
11:40 Tux then ->river should return an empty hash imho
11:40 Tux now I have to add checks
11:40 mickey i can fix that
11:41 Tux yes please
11:41 Tux I'll commit the collected changes to Release::Checklist
11:44 mickey hmm... i don't see anything about Release::Checklist on either v0 or v1
11:44 Tux https://github.com/Tux/Release-Checklist/commit/337611243d8f10d21af80e499dc97ccc9ff81d61
11:47 mickey yeah, but you're looking it up in metacpan api
11:47 mickey i don't see it there at all
11:48 Tux then I do not understand your question
11:48 mickey you were looking for a non-existing distribution
11:49 mickey that's why you got a 404 and the ua said Faile...
11:49 Tux correct: that has been fixed in makewww.pl
11:49 Tux my error. not yours
11:49 mickey that's different from an existing distribution with no river data which i can fix to {}
11:49 Tux that yes, please
11:49 mickey ok, just to be clear on the fix :)
11:50 * Tux notices he has been unclear somewhere in the discussion and apologizes
11:50 mickey hehe, no worries Tux... that's what we have discussions for
11:50 mickey and i appologize for grasping slowly
11:51 Tux using ->river is *EXTREMELY* much faster than using the old method
11:51 Tux \o/
11:51 mickey do you have an example of a dist that is missing it?
11:51 mickey just to quickly test
11:51 Tux so even if the values could be different, at least I now have values that are reliable
11:52 Tux mickey, wait a bit: they are still running, so I cannot check which one failed
11:52 mickey ok
11:53 Tux ILMARI - Acme::Holy::Shit::Here::Come::s::An::S
11:54 ilmari Tux: I blame arc
12:04 Tux always a safe option to choose :P
12:07 ilmari it even has a 'SCAPECOAT section
12:11 Tux ilmari, thi's i's not 'bout the module it'self, it's about MetaCPAN::Client ruturning undef in'stead of an empty ha'sh for downriver dependency count's
12:12 Tux Thi's module i's awe'some for people that do not know the difference between it's and it's
12:18 metacpan joined #metacpan
12:18 metacpan [metacpan-client] mickeyn pushed 2 new commits to master: https://git.io/vifwn
12:18 metacpan metacpan-client/master 2f7bfa7 Mickey Nasriachi: distribution: add default values for missing keys
12:18 metacpan metacpan-client/master 6e4d005 Mickey Nasriachi: release: 1.025000
12:18 metacpan left #metacpan
12:19 ranguard mickey: I got 'bugs()' returning undef from dist I think
12:19 mickey yeah, that happens as well
12:20 ranguard if you are going for consistency, but didn't seem strange to me if there wern't any
12:20 * ranguard hasn't ready docs :)
12:20 ranguard mickey++ # fixing of MC::Client
12:20 mickey this will also get an empty hash now
12:22 ranguard Tux: so since 9:40 GMT I've not had any issues with calling LLAP
12:22 ranguard switching to using HMBRAND
12:23 Tux http://tux.nl/Files/LLAP.html
12:24 mickey Tux: https://v1.metacpan.org/release/MICKEY/MetaCPAN-Client-1.025000
12:24 ranguard Tux: I'm just trying to replicate your fastapi issues rather than generate the output :)
12:25 Tux I know
12:25 neilb joined #metacpan
12:25 Tux and together we work towards a better CPAN :)
12:25 Tux mickey++ ranguard++
12:27 ranguard please let me know when you get another issue and I'll see if I got one at the same time
13:21 Tux http://tux.nl/Files/ILMARI.html successfully updated now
13:22 Tux Currently the delay for the pages mostly comes out of analyzing RT
13:26 mickey Tux: RT info is in distribution("X")->bugs ('type' eq 'rt')... can't that save you some time?
13:26 mickey MetaCPAN::Client->new->distribution("Text-CSV_XS" )->bugs
13:26 * Tux starts experimenting on that right away …
13:34 Tux does it also have the links to the github issues?
13:37 mickey potentially, not for all distributions
13:37 mickey some do
13:38 Tux I more meant like what you now have in rt: open, closed, resolved, rejected, patched, …
13:39 Tux e.g. Text::CSV_XS might have issues in RT (open tickets of type rt) and issues in github
13:39 Tux these are unrelated
13:39 mickey sure, i'm just not sure why some don't have that info
13:44 mickey that's a Pithub (https://v1.metacpan.org/pod/Pithub) thingy... oalders should know more
13:44 mickey oalders: ^^
13:54 mickey oalders: i see we can only have one source type in the bugs index, so we overwrite the data with the latest we find (rt being the latest). why is that? can't we make bugs a hash with separate keys for rt and gh and have both datas?
13:55 jberger I'd also love a way to distinguish feature requests from bugs
13:55 jberger whether by convention or configuration
13:56 jberger maybe something I could work on at a hackathon someday :)
13:57 mickey jberger: hehe... yeah, we're trying to make that "someday" into a string that also contain numbers
13:57 jberger christmas
13:57 jberger errr, chr1stmas, if you need numbers
13:58 mickey well, i guess christmas is pretty deterministic date
13:58 mickey i don't mind, but most ppl will probably not want to use it for a hackathon
14:11 oalders mickey: i think we just wanted one canonical source of bugs to display on the front end, but i'd be happy to change how that works
14:11 oalders since it obviously isn't a good fit across the board
14:26 * Tux would love to see that too
14:27 Tux enev if that breaks my current implementation!
14:27 Tux even
14:41 oalders Tux: could you open an issue for that on metacpan-api?
14:44 Tux yep, will do so :)
14:51 Tux done
15:29 oalders Tux++
15:38 mickey Tux++
15:40 Khisanth joined #metacpan
16:24 skaji hi
16:24 skaji It seems that v1 download_url endpoint does not support "!=" notation.
16:25 skaji ex: https://fastapi.metacpan.org/v1/download_url/Moose?version=%21%3D%202.1805
16:26 skaji Should I add this issue to https://github.com/metacpan/metacpan-api/issues/483 ?
16:28 oalders skaji: that's a good place to have the conversation. maybe you could just clarify the use case and also what your expected output would be
16:30 skaji oalders: OK, I will add a comment to the issue. thanks!
16:36 oalders skaji++
17:21 mickey skaji: this is a separate issue, please remove the comment and create a new one for it for better tracking
17:23 skaji mickey: OK:-)
17:25 mickey thanks, that way we can hold separate conversations and know which problems are fixed
17:46 oalders that works too :)
17:51 gordonfish joined #metacpan
18:35 oiami joined #metacpan
18:43 uree joined #metacpan
19:20 metacpan joined #metacpan
19:20 metacpan [metacpan-api] mickeyn created mickey/bugs_remap (+1 new commit): https://git.io/viJ93
19:20 metacpan metacpan-api/mickey/bugs_remap c61f45f Mickey Nasriachi: distribution: bugs -> rt + github
19:20 metacpan left #metacpan
19:34 metacpan joined #metacpan
19:34 metacpan [metacpan-api] mickeyn force-pushed mickey/bugs_remap from c61f45f to 8f2d4c5: https://git.io/viJQc
19:34 metacpan metacpan-api/mickey/bugs_remap 8f2d4c5 Mickey Nasriachi: distribution: bugs -> rt + github
19:34 metacpan left #metacpan
20:00 neilb joined #metacpan
21:58 mateu joined #metacpan

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