Camelia, the Perl 6 bug

IRC log for #mojo, 2012-10-09

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

All times shown according to UTC.

Time Nick Message
00:00 motoboi Well seems that calling to_string creates a cached version, which won't be invalidated after appending a new header entry.
00:25 d4rkie joined #mojo
00:39 zivester joined #mojo
00:51 laouji joined #mojo
00:55 motoboi joined #mojo
02:18 tempire joined #mojo
02:20 noganex_ joined #mojo
02:21 tempire If we're critiquing the front page, I think the code needs a higher line-height
02:21 tempire or maybe a shorter block.  it looks complicated on first glance.
02:22 tempire damn.  it doesn't look complicated to me anymore.  I've adapted.  But it used to.
02:24 tempire An amazing real-time web framework allowing simple one-file Mojolicious::Lite apps, large structured web apps for teams, and any size in between.
02:30 tempire 1.7em: https://img.skitch.com/20121009​-xd22nuihypbh28p9h5cnqbisni.jpg
02:39 janliesendahl_de joined #mojo
03:42 memowe_ joined #mojo
03:46 motoboi joined #mojo
05:04 Foxcool joined #mojo
05:04 motoboi joined #mojo
05:07 yakudza joined #mojo
05:37 Britzel joined #mojo
06:17 dpetrov_ joined #mojo
06:41 spleenjack joined #mojo
07:11 sawtooth joined #mojo
07:12 yakudza joined #mojo
07:13 batman joined #mojo
07:13 Vandal joined #mojo
07:17 Vandal joined #mojo
07:21 Mike-PerlRecruiter_ joined #mojo
07:33 ladnaV joined #mojo
07:42 pau4o joined #mojo
08:04 fhelmber_ joined #mojo
08:24 bannaN joined #mojo
08:28 bannaN I get the message "Template templatename.json.ep not found. Nothing has been rendered, expecting delayed response". Is it possible to just render whatever is in the json stash value without a json template? If so, ho is it done?
08:36 crab $self->render... oh, gone.
08:40 yakudza_ joined #mojo
08:43 bannaN joined #mojo
08:44 Vandal joined #mojo
08:47 rem_lex joined #mojo
08:59 lukep joined #mojo
08:59 lukep joined #mojo
09:27 ladnaV joined #mojo
09:28 yakudza__ joined #mojo
09:44 mjy joined #mojo
09:45 mjy hi, is there a way to set production mode in a Mojolicious::Lite script other than by setting the environment var?
09:52 dpetrov_ mjy: probably https://metacpan.org/module/Mojolicious::Lite#Mode ?
10:03 mjy I've read this dpetrov, thanks, but I don't use the daemon mode, I run the script directly under fcgi or mod_perl environments
10:03 dpetrov_ what is wrong with the env variable? :)
10:03 mjy -m production will probably work in the FcgidWrapper string
10:04 mjy well it's not as practical as app->mode('production')->start would be for example ;)
10:09 fhelmber_ joined #mojo
10:10 daxim joined #mojo
10:11 dpetrov_ I think app->mode('production') should work as well
10:12 mjy doesn't from what I've tested :(
10:15 dpetrov_ or I am wrong :)
10:15 Lucas1 joined #mojo
10:27 nic mjy: I found the same thing
10:27 nic If I remember correctly, the docn examples only work in a full app
10:43 ObseLeTe joined #mojo
10:52 yakudza joined #mojo
11:01 rem_lex|pivo joined #mojo
11:05 laouji joined #mojo
11:17 fhelmbe__ joined #mojo
11:28 ovnimancer joined #mojo
11:35 fhelmber_ joined #mojo
11:38 ovnimancer joined #mojo
11:51 holgr joined #mojo
11:52 holgr left #mojo
11:56 Lucas11 joined #mojo
12:05 yakudza hello ppl . Is there any easy good looking way to show select multiple input field with items selected from previous request. I get params from standart param helper.
12:10 yakudza I mean to solve this issue directly in template. Currently I do this half in controller half in template. Controller: map {$params_checked->{networks}->{@_} = 1} @$networks; and then pass to template checked params and in template <%= 'selected' if ($params_checked->{$net->{id}}) %>
12:10 ovnimancer joined #mojo
12:16 Lucas1 joined #mojo
12:36 arthas_ joined #mojo
12:45 Lucas1 joined #mojo
13:00 bluescreen joined #mojo
13:01 xxtjaxx_ sri: Is there a replacement mojo plugin/command/tool for require.js r.js?
13:13 pau4o joined #mojo
13:15 jnap joined #mojo
13:46 motoboi joined #mojo
13:52 batman left #mojo
13:57 gryphon joined #mojo
13:58 batman joined #mojo
13:59 d4rkie joined #mojo
14:24 sh4 joined #mojo
14:30 knshaum joined #mojo
14:40 mfontani joined #mojo
14:48 sri tempire: i like that you are mentioning teams, but the sentence as a whole doesn't sound so nice
14:52 Jason__ joined #mojo
14:59 Leandr joined #mojo
15:06 bluescreen joined #mojo
15:15 vel joined #mojo
15:17 inokenty joined #mojo
15:48 batman sri: i agree
15:48 batman it's too complicated
15:50 batman when it comes to the font, i would boost it to 0.9em and line-height 1.4em
15:50 batman i also find the scrolling absolutely super irritating
15:50 batman i'm using metacpan instead just because of the scrolling
15:55 sri i don't care about anything else than that one first sentence
15:55 batman "i agree" ?
15:55 purl somebody said "i agree" was good.
15:56 sri the one we are discussing since yesterday
15:57 batman ok
15:58 * sri just thought about supporting PLACK_ENV as an alternative to MOJO_MODE... but it would cause backwards compatibility problems
16:01 sri re your line height, scrolling... whatever changes... just do the right thing and make a real proposal we can actually vote on
16:03 xaka joined #mojo
16:05 batman ok
16:05 batman will do
16:19 dpetrov_ joined #mojo
16:40 dvinciguerra joined #mojo
16:47 ObseLeTe joined #mojo
17:04 fhelmber_ joined #mojo
17:13 Lucas1 joined #mojo
17:13 fhelmber_ joined #mojo
17:20 sri does someone here have access to a really fast linux box with at least 4 cores and could run two benchmarks for me?
17:21 sri i wonder if 10 is too low for multi-accept on modern servers
17:21 sri http://hastebin.com/raw/nayahavixo
17:22 fhelmber_ joined #mojo
17:22 sri i'd like to see the difference before and after applying that patch for "./script/hypnotoad -f examples/fast.pl" and "ab -c 100 -n 100000 http://127.0.0.1:8080/"
17:23 sri fresh mojolicious master clone will do
17:24 fhelmber_ joined #mojo
17:24 sri difference could be substantial if the server is fast enough
17:24 sri as in thousands of rps
17:26 Lucas1 joined #mojo
17:26 sri (ideal benchmark would be with EV installed on the box)
17:34 crab which patch?
17:34 crab http://hastebin.com/raw/nayahavixo?
17:38 sri ye
17:38 crab Requests per second:    1407.74 [#/sec] (mean)
17:38 crab Requests per second:    2376.97 [#/sec] (mean)
17:39 sri very interesting
17:39 crab before/after on a slow-ass dual-core machine with only 6gb of ram
17:39 sri and if you change 50 to 100? :)
17:39 yakudza joined #mojo
17:39 sri uh... that is weird
17:40 sri i have dual core boxes here too and see no difference
17:41 crab well, i guess this is quad-core if you count the fake cores (hyperthreading)
17:41 crab Requests per second:    2166.98 [#/sec] (mean)
17:41 crab (with 100)
17:41 sri but i guess we should increase the value and make it configurable
17:41 sri fake cores do count
17:42 crab Requests per second:    2259.37 [#/sec] (mean)
17:42 crab with 70
17:43 crab yeah, sounds good
17:43 sri 50 looks like a good number
17:44 sri but i would like to see the results for a really fast server
17:44 jzawodn hrm... we have a lot of 24 core boxes with 72GB RAM.  I could test on one of them a bit later today
17:44 crab octa-core i7 920 @2.67ghz counts?
17:44 sri \o/
17:45 sri 24 core might be overkill though :)
17:46 * sri wonders if i7 920 is the latest generation
17:46 jzawodn heh... ok.  we also have a lot of 16 core boxes that are slightly older... holler at me if you want, I can test later today with whatever versions of Perl and Mojo you'd like
17:47 sri thanks
17:47 sri i7 920 looks like a bloomfield, which would make it rather old
17:48 crab sri: probably quite old, like ~2y
17:48 sri sandy/ivy with 4-6 cores would be cool
17:49 crab 1380.77 to 2628.93 on the i7 920 (10->50)
17:50 sri and 100?
17:50 purl hmmm... 100 is the reference
17:50 sri 50 is pretty much a done deal, open question is just if 100 makes sense :)
17:51 crab 2444.74
17:51 sri hmmmm
17:52 crab i wouldn't suggest making 100 the default anyway, it's the kind of thing that will cause other problems
17:53 crab e.g. overrunning max_connections in postgres and such.
17:53 sri good point
17:53 purl nice and sharp
17:56 pau4o left #mojo
18:03 gryphon joined #mojo
18:07 ObseLeTe joined #mojo
18:12 batman joined #mojo
18:28 good_news_everyone joined #mojo
18:28 good_news_everyone [mojo] kraih pushed 2 new commits to master: http://git.io/MZo06Q
18:28 good_news_everyone [mojo/master] added multi_accept attributes and improved accept performance of all built-in web servers significantly - Sebastian Riedel
18:28 good_news_everyone [mojo/master] added multi_accept setting to Hypnotoad - Sebastian Riedel
18:28 good_news_everyone left #mojo
18:30 sh4 joined #mojo
18:31 Britzel joined #mojo
18:31 fhelmbe__ joined #mojo
18:31 sri please test!
18:32 sri and feel free to tweak the multi_accept setting to see which one works best on your boxes
18:33 sri lower values like 25 or 30 might also be interesting
18:34 sri nginx doesn't appear to have a limit at all, but since we are serving dynamic content i guess it's better to have one for us
19:01 Jason__ joined #mojo
19:03 kevouze joined #mojo
19:21 _bob joined #mojo
19:22 Mike-PerlRecruiter_ joined #mojo
19:22 _bob_ left #mojo
19:23 ObseLeTe joined #mojo
19:25 kevouze2 joined #mojo
19:28 _bob joined #mojo
19:47 abra joined #mojo
20:16 _nightmare joined #mojo
20:23 good_news_everyone joined #mojo
20:23 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/iz0VfA
20:23 good_news_everyone [mojo/master] improved documentation browser CSS - Sebastian Riedel
20:23 good_news_everyone left #mojo
20:28 sri batman: http://mojolicio.us/perldoc # is tha what you had in mind with your complaint about scrolling?
20:29 kevouze2 left #mojo
20:36 lukep joined #mojo
20:39 ZadYree aha, I like the bot name
20:57 ObseLeTe joined #mojo
21:00 gryphon joined #mojo
22:19 jzawodn joined #mojo
22:23 gryphon joined #mojo
22:27 kevouze joined #mojo
22:50 * tempire likes the smooth scrolling
23:07 tempire neato https://blog.mozilla.org/blog/2012/10/0​9/firefox-debuts-new-developer-toolbar/
23:08 sri smooth scrolling is gone, batman killed it
23:33 gryphon joined #mojo
23:35 batman sri: YAY!
23:37 batman tempire: it just takes too much time. i remember the first time i saw it and i thought it was cool.
23:37 batman a bit like flash i guess :P
23:38 * sri never really cared about smooth scrolling
23:38 batman my mind is free! i can now go to sleep :)
23:39 * batman is making Mojo::SNMP
23:41 batman https://github.com/jhthorsen/mojo-snmp
23:42 batman just because i can...
23:42 batman *ttyl*
23:43 sri it's not a bad example for how to hook something without a handle into Mojo::IOLoop
23:46 batman i going for a bit more than just an example...
23:46 sri i like to call ->run just ->wait these days though
23:46 batman good idea
23:47 batman $snmp->start->wait; looks ok
23:47 sri ->wait implies ->start to me
23:47 batman yeah. sorry. a bit too late :P
23:48 sri or rather ->start->stop
23:48 batman what?
23:48 sri ->wait eq ->start; ->stop;
23:48 batman right
23:52 batman i was trying to hook into SNMP.pm, but that wasn't all that easy
23:52 batman so i change to Net::SNMP

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