The web in a box - a next generation web framework for the Perl programming language

IRC log for #mojo, 2014-12-08

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

All times shown according to UTC.

Time Nick Message
00:01 fdgdgsdfg joined #mojo
00:05 asarch joined #mojo
00:23 noganex joined #mojo
00:27 KCL joined #mojo
00:37 franzkafka I have a callback in a UserAgent get that does an emit... I am wondering, is there any way to ensure the entire DOM is loaded before doing the emit?
00:50 franzkafka ok, figured that out
00:51 franzkafka Should this not work: $dom->at('h1.title')->text; if <h1 class="title"> BLAH </h1>
00:51 franzkafka I.e., the spaces
00:58 franzkafka It does work in a test. There must be something else here messing things up.
01:07 jberger sri: re https://github.com/kraih/mojo/pull/713
01:07 jberger http://pastie.org/9767028
01:07 jberger just a thought
01:07 jberger its pay for what you use as well
01:08 punter joined #mojo
01:15 asarch joined #mojo
01:20 sri jberger: Mojo::Base does not require Mojolicious
01:20 sri oops
01:20 sri missed that require
01:21 sri if you load Mojolicius anyway, why not just use that for the version check?
01:22 sri anyway, i'm -1 on the whole concept
01:23 jberger mostly I was curious if I could make it work
01:24 sri when can we use "package Mojolicious 5.69;" anyway?
01:25 jberger hmmm, good question
01:25 purl Yeah, it is. I'm stumped.
01:27 jberger package block (and package version block) are supported in 5.14
01:27 * jberger checks earlier for package version
01:28 jberger yeah, 5.12
01:28 jberger https://metacpan.org/pod/distribution/perl/pod/perl5120delta.pod#New-package-NAME-VERSION-syntax
01:29 woz joined #mojo
01:29 sri hmm
01:29 * sri always hated use statements with versions... and the whole concept of "lets install shit from cpan until our script stops to die"
01:30 sri you should just have a Makefile.PL or cpanfile
01:30 sri they have their use for optional deps (as we still use them)... but not for hard deps
01:31 jberger I get that
01:31 jberger in fact, when I bump a version number in a Build.PL (or whatever) I comment why I need that version
01:32 sri i wish more perl people learned that stuff, and how you can just freeze your deps to specific versions you've actually tested
01:32 jberger https://github.com/jberger/Galileo/blob/master/Build.PL#L14
01:33 sri i think i vaguely remember that marcus promised a cpanfile cookbook recipe
01:33 jberger hmmm, it seems I'm not as good at that I thought I was
01:34 sri and we thought about a cpanfile generator
01:34 jberger I do use cpanfile for the FWBM stuff
01:34 jberger its handy there
01:34 jberger and carton
01:35 sri of course the fact that miyagawa said carton was dead and there would be something new makes that seems like such a great idea now...
01:35 jberger how would a cpanfile generator work? scan source for deps?
01:35 jberger there must be some PPI-ish tool out there for that
01:35 sri just like the Makefile.PL generator
01:36 * jberger has never used that
01:36 sri just something to encourage people to use it
01:36 * jberger uses Build.PL
01:36 sri it does pretty much nothing
01:36 jberger I'm trying to switch to Module::Build::Tiny actually
01:37 sri Module::Build does not look liek a healthy project
01:38 jberger I used it quite a bit for Alien::Base, and while I understand it has issues, I was able to make a really powerful build tool with it
01:38 sri and i think it didn't copy non .pm/pl files in the lib dir, which really annoyed me
01:38 jberger sri: yes, that is a major difference
01:39 jberger actually, hmmmmm
01:39 jberger so leont made App::ModuleBuildTiny which is meant as the companion authoring tool for Module::Build::Tiny
01:39 jberger its less advanced than the authoring portions of EUMM or MB
01:39 jberger so I've been tinkering with some extensions (like meta merge)
01:40 jberger and now you have me wondering, I bet it would just bundle the whole lib directory
01:40 * jberger checks
01:40 * sri tries to stay far far far away from build tools
01:41 sri back in the catalyst days i wrote Module::Build and Module::Install extensions for catalyst... and it left scars...
01:42 jberger I think App::MBT would work, it just bundles everything in the manifest
01:42 sri and i'm sure it scarred everyone trying to maintain it afterwards
01:42 jberger and the manifest generation is pretty lax
01:42 jberger problem is, build tools are important
01:42 jberger I really dislike make and EUMM (at least for perl, C maybe)
01:43 jberger for now that leaves MB proper
01:43 jberger but MBT is MUCH more sane
01:45 jberger here is where I am on Module::MetaMerger btw: https://github.com/jberger/Module-MetaMerger
01:45 jberger in case anyone is interested
01:51 good_news_everyon joined #mojo
01:51 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/p6XwyA
01:51 good_news_everyon mojo/master 6b24270 Sebastian Riedel: removed deprecated has_conditions method from Mojolicious::Routes::Route
01:51 good_news_everyon left #mojo
01:58 sri really wish we had somebody hacking on http/2 preparations
02:06 sri i would even be happy with a prototype patch based on https://metacpan.org/pod/Protocol::HTTP2
02:11 akhasanov joined #mojo
02:45 human39 joined #mojo
02:55 klapperl joined #mojo
03:15 noganex_ joined #mojo
03:17 tencendur joined #mojo
03:34 hexreel joined #mojo
03:45 disco joined #mojo
04:27 franzkafka joined #mojo
04:42 mattastrophe joined #mojo
04:49 Eke- joined #mojo
04:52 franzkafka sri, you seem to be in-the-know when it comes to things... do you use a framework client side, and if so, which one?
04:53 preaction i recommend angular, i've been using it on all my recent apps to great effect. combine with bootstrap for UI
05:08 * sri is trying to outline the steps for adding HTTP/2 support to mojolicious https://github.com/kraih/mojo/issues/423#issuecomment-65975427
05:12 franzkafka preaction, awesome, that's what I am using. I love when I make the right choices.
05:12 preaction i wouldn't call them the right choices :p
05:12 preaction there's wrong choices, absolutely, but there's no right choice
05:13 preaction react and its ilk have been getting popularity recently. but extjs still exists (we don't talk about yui anymore)
05:17 sri until angular 2.0 breaks everything ;p
05:19 rem_lex|pivo joined #mojo
05:21 preaction yeah, that oughta be fun
05:38 basiliscos joined #mojo
05:44 sunil joined #mojo
05:45 franzkafka Pulseaudio is a steaming pile of shit
05:51 basiliscos joined #mojo
06:01 basiliscos joined #mojo
06:53 neyasov__ joined #mojo
06:56 Adura Try the ALSA... Ice Bucket, think it's called, instead.
06:58 dod joined #mojo
06:59 Dandre left #mojo
06:59 Dandre joined #mojo
07:08 Eke- joined #mojo
07:27 jamesaxl joined #mojo
08:01 McA3 joined #mojo
08:07 davido_ joined #mojo
08:15 eseyman joined #mojo
08:27 vytas joined #mojo
08:36 absolut_todd joined #mojo
08:38 Vandal joined #mojo
08:44 stryx` joined #mojo
08:46 ashimema joined #mojo
09:13 sugar joined #mojo
09:31 sugar_ joined #mojo
09:32 neyasov__ joined #mojo
09:40 woz joined #mojo
09:41 woz_ joined #mojo
09:47 denis_boyun joined #mojo
10:00 punter joined #mojo
10:13 vytas joined #mojo
10:44 neyasov___ joined #mojo
10:50 meshl joined #mojo
10:51 meshl Is there rapid app framework, "a la http://www.rapidapp.info/" available on Mojolicious?
11:04 denis_boyun joined #mojo
11:07 ashimema joined #mojo
11:17 woz joined #mojo
11:40 alnewkirk joined #mojo
11:47 amon joined #mojo
11:53 mattastrophe joined #mojo
12:10 basiliscos joined #mojo
12:16 asarch joined #mojo
12:18 woz joined #mojo
12:25 Lee_ joined #mojo
13:16 sugar__ joined #mojo
13:32 mattastrophe joined #mojo
13:45 woz_ joined #mojo
13:47 akhasanov joined #mojo
13:49 neilhwatson joined #mojo
14:29 Ptolemarch joined #mojo
14:41 sugar joined #mojo
14:51 sugar_ joined #mojo
15:02 zivester joined #mojo
15:05 rsperl joined #mojo
15:07 mattastrophe joined #mojo
15:11 sh4 joined #mojo
15:11 rsperl joined #mojo
15:32 muraiki joined #mojo
15:36 Ptolemarch joined #mojo
15:38 apl i'm trying to use minion with a rather high load: 30jobs are queued every second
15:38 apl after about 60k jobs the performance starts to degrade
15:39 apl could this be because of the postgres-trigger?
15:41 apl it looks like it sends a notification for every row on every insert
15:43 apl i'm running 10 workers, the postgres box has 32 cores and like infinity gb of ram
15:48 akhasanov joined #mojo
15:54 sri explain analyze the queries and you'll see
16:08 sugar_ joined #mojo
16:09 denis_boyun joined #mojo
16:28 sugar__ joined #mojo
16:28 sri apl: and share the results
16:51 apl sry, cluster went down
16:52 apl i will try to reproduce this tomorrow
17:01 sri that's usually the last thing i hear before people vanish :S
17:07 rsperl joined #mojo
17:10 mst sri: but don't you understand? THE CLUSTER IS DOWN
17:18 sri (for the record, no, a notification trigger should not result in performance degradation, the cost is pretty constant)
17:19 sri the more busy your workers the cheaper it gets actually
17:21 sri which is why i'm asking for explain analyze results
17:21 sri to see what the query planner ended up with
17:24 tencendur joined #mojo
17:30 rsperl joined #mojo
17:34 Ptolemarch joined #mojo
17:38 sri in my artifical tests, after 300k jobs i notice no performance degradation
17:38 sri 1900 enqueues per second (Postgres running with an SSD)
17:40 sri with two workers
17:46 sri anyway
17:46 sri we still need volunteers to work on this https://github.com/kraih/mojo/issues/423#issuecomment-65975427
17:50 akhasanov joined #mojo
17:54 marcusr maybe the performance degraded because the cluster went down?
17:56 denis_boyun joined #mojo
17:57 jberger sri: argh! I just ran into this bug at $work: https://rt.cpan.org/Public/Bug/Display.html?id=78656
17:57 sri haha
17:58 jberger Time::HiRes::clock_gettime not implemented on OSX!
17:58 jberger should I try to bug someone on #p5p to get rcaputo's patch applied finally?
17:59 sri YES
17:59 * sri bugs jberger
18:01 sri crazy how it's already been 2.5 years :o
18:02 jberger with code written too
18:02 jberger I understand how bare bug reports/feature requests can sit idle, but its sad when code is ready and it does so
18:03 jberger rjbs just said he would take a look
18:03 sri rcaputo is dngor btw.
18:04 jberger oh I didn't know
18:13 Eke- joined #mojo
18:33 inokenty joined #mojo
18:35 dod joined #mojo
18:45 asarch joined #mojo
18:57 mattastrophe joined #mojo
19:10 neyasov___ joined #mojo
19:14 akhasanov joined #mojo
19:18 trone joined #mojo
19:29 Ptolemarch joined #mojo
19:44 disputin joined #mojo
19:49 woz joined #mojo
19:50 meshl joined #mojo
19:55 rsperl joined #mojo
20:01 jamesaxl joined #mojo
20:17 disco joined #mojo
20:18 disputin joined #mojo
20:24 neilhwatson joined #mojo
20:39 asarch joined #mojo
20:47 sugar joined #mojo
20:51 sri we really need more folks that know the Mojo::UserAgent and Mojo::Server::Daemon internals
21:00 woz joined #mojo
21:06 jberger sri: preaction just did a bunch of stuff with the Daemon
21:06 jberger I know a fair amount of the UA
21:10 Eke- joined #mojo
21:11 neyasov___ joined #mojo
21:21 davido_ joined #mojo
21:57 denis_boyun joined #mojo
21:59 disputin joined #mojo
22:10 woz joined #mojo
22:13 Ptolemarch joined #mojo
22:21 denis_boyun joined #mojo
22:25 Grinnz_ how do i specify form content with the 'get' command, -c 'key=value' doesn't seem to be working
22:25 Grinnz_ (as a POST)
22:29 tencendur joined #mojo
22:41 sri funny you ask, i wanted to make it easier a few days ago, but my patch didn't get much feedback
22:42 * sri wanted to add "mojo get -M POST -f '{"key":"value"}' ..."
22:43 sri Grinnz_: right now you have to do *everything* manually
22:47 Grinnz_ sri: shouldn't that be how to do it manually though?
22:47 Grinnz_ or do you mean... it needs metadata
22:47 good_news_everyon joined #mojo
22:47 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/Akh6Vw
22:47 good_news_everyon mojo/master d05e330 Sebastian Riedel: more detailed example for get command
22:47 good_news_everyon left #mojo
22:48 Grinnz_ ah right, content-type
22:49 Grinnz_ -H 'Content-Type: application/x-www-form-urlencoded' -c 'key=value' works :)
23:36 Grinnz_ in https://metacpan.org/pod/Mojolicious::Command::eval, synopsis, is "app->home" supposed to refer to something? i can't find it in Mojolicious or DefaultHelpers
23:38 Grinnz_ btw, +1 from me for that -f option for get... i thought i saw something like that mentioned before
23:47 preaction my guess is app->home is a Mojo::Home object. maybe it's in Mojo::Base?
23:48 sri http://mojolicio.us/perldoc/Mojolicious#ATTRIBUTES
23:49 Grinnz_ ah, its in Mojo

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