Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2016-09-07

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

All times shown according to UTC.

Time Nick Message
04:08 anon joined #metacpan
06:23 oiami joined #metacpan
06:49 [Tux] joined #metacpan
06:49 [Tux] Date: Wed,  7 Sep 2016 07:44:06 +0200 (CEST)
06:49 [Tux] [Request] ** [https://fastapi.metacpan.org:443]-[599] Could not read from SSL socket: 'SSL read error'
06:49 [Tux] at /pro/lib/perl5/site_perl/5.22.0/HTTP/Tiny.pm line 1217., called from sub Search::Elasticsearch::Scroll::next at /work/www/make-perl.pl line 265. With vars: {'request' => {'path' => '/_search/scroll','method' => 'POST','ignore' => [],'qs' => {'scroll' => '5m'},'mime_type' => 'application/json','serialize' => 'std','body' => 'c2NhbjszOzE0NTk0OTg6clZfQVAyNHVSOC05QjViRzFMNGN2dzsxNDYyMTI3Ok1zMVJwdFU1UlgyRGwyOUZUaUdPMHc7MTQ2MDExODpiUzUzdk1wdVRqNkt3d0RYR2t
06:49 [Tux] XVDRBOzE7dG90YWxfaGl0czoyOw=='},'status_code' => 599}
07:24 neilb joined #metacpan
07:47 osfabibisi joined #metacpan
08:04 Relequestual joined #metacpan
08:11 neilb joined #metacpan
08:12 edward joined #metacpan
09:13 ranguard [Tux]: cheers
09:23 mickey Tux, ranguard: maybe for sports we should have your stuff tested with a given alternative ua (the client takes 'ua => <obj>' as a parameter
09:24 mickey just to see it's not an HTTP::Tiny thing
09:25 mickey (saying that, because ranguard pointed out some http-tiny dependency restrictions earlier in relation to the ssl issues)
09:32 ranguard mickey: it works 99% of the time (I'm running it every 2 mins and get one or 2 failures a day)
09:32 ranguard I couldn't match it to a request on our server, so need to get fastly logging running to see what's happening
09:33 ranguard pingdom monitoring of nginx on lw-mc-03 show's 100% uptime
09:34 mickey cool, just wanted to eliminate a ua issue on the client side... but i'm not sure i know what i'm talking about :)
09:34 ranguard it's a good idea - just one step later in the debug process :)
11:18 osfabibisi is Penfold still active?  Longstanding install bug in https://rt.cpan.org/Public/Bug/Display.html?id=104224 that bites us, wondering whether it's worth knocking up a patch and asking for maint
11:24 Tempesta joined #metacpan
11:26 * ranguard shrugs
11:26 ranguard have you emailed mike@altrion.org
11:31 osfabibisi good point ta
11:39 damog joined #metacpan
14:00 Relequestual joined #metacpan
14:25 Lee joined #metacpan
17:03 neilb joined #metacpan
17:56 ranguard evening, oalders so I've got some stuff for mc-web and mc-api - should we just start MetaCPANX:: name space and put stuff there?
17:56 oalders ranguard: we also do have the MetaCPAN namespace already
17:57 oalders ie MetaCPAN::Moose
17:57 * ranguard ideally wondered if Project::MetaCPAN:: would be better
17:57 ranguard oalders: oh, cool, I'll go with that
17:57 oalders sounds good
17:58 ranguard should we put Type into MC::Moose then ?MetaCPAN::Moose
18:01 ranguard e.g. MetaCPAN::Moose::Types I mean - merge https://github.com/metacpan/metacpan-api/blob/master/lib/MetaCPAN/Types.pm and https://github.com/metacpan/metacpan-web/blob/master/lib/MetaCPAN/Web/Types.pm and https://github.com/metacpan/metacpan-api/blob/master/lib/MetaCPAN/Types/Internal.pm
18:01 ranguard hmm, maybe not the last one
18:03 ranguard anyway, I'll release MetaCPAN::Role which will have MC::Role::Fastly (for scripts) and MetaCPAN::Role::Fastly::Catalyst (for -web -api)
18:06 ranguard oalders: please move p5-metacpan-moose repo into 'metacpan' gh org, rather than your own account
18:06 * ranguard didn't even know it existed :)
18:07 oalders weird. i thought it was already there
18:07 oalders ranguard: yeah, merging the types would be good
18:07 oalders i'll move that repo
18:08 ranguard oh, it is there, but the official repo (from the link on mc is yours)
18:09 ranguard that's even worse :)
18:09 ranguard as yours is v00001 and mc/p5- is v00002
19:46 oalders i'll release a new version tonight with a fixed repo url
19:58 ranguard oalders++
19:59 ranguard ah,ha! - can't have "has 'config'" need  "sub config" for these roles consumming roled and requires()
20:09 oalders already sounds complicated ;)
20:14 ranguard https://gist.github.com/ranguard/3f370c8b62cec19047f88b090d8f35bf BOOM - https://gist.github.com/ranguard/a1fac0fb6a5211623d7ae872f9ab13b6
20:20 gordonfish joined #metacpan
20:21 oalders yeah, i generally avoid requires() because i can't be bothered trying to work around that kind of thing
20:26 ranguard heh, I was just trying to be nice as it was in a CPAN module
20:38 mst ranguard: err, not strictly the case
20:38 mst ranguard: did you forget the explanation I already gave you? :(
20:39 mst in Moose, attribute methods don't actually exist until the role is composed onto the class
20:39 mst because moose made the design choice/mistake of using the class' meta-attribute class when that happens
20:39 mst so e.g. a role with "has foo => (is => 'rw');"
20:39 mst being consumed by a class using SemiAffordanceAccessor
20:39 mst would get 'sub foo' and 'sub get_foo' generated
20:40 mst so the list of methods isn't technically known in advance
20:40 mst of course this means if your role did $self->foo($value) instead of $self->meta->get_attribute-by_name('foo')->set_value($self, $value) your role is now broken
20:40 mst so it's a silly feature and I've never seen code that actually relied upon it
20:40 mst so Moo doesn't
20:41 mst ranguard: I mean, the simple solution here is "convert both roles to Moo" but that's sort of cheating ;)
20:43 ranguard mst: I didn't quite get the explination, hence experimenting - thanks for the extra clarification - make sense
20:43 mst ranguard: yeah, well, if you don't, do ask rather than flailing :)
20:43 ranguard thanks :)
20:44 ranguard I was actually flailing with something else mostly, but then came back to get my head around this :)
21:36 ranguard mickey: just thought - indexing issue could be because now via the queue, the rsync might be a min or so diff in the actual worker box
21:36 ranguard so the dist might not be there to index
21:36 ranguard can discuss tomorro - nn :)

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