Camelia, the Perl 6 bug

IRC log for #mojo, 2011-01-13

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

All times shown according to UTC.

Time Nick Message
00:01 perlrocks Twitter: "@kraih Hello! I look forward to start with Mojolicious! Coming from the oldschool CGI with scalability issues, it might feel fresh :)" --f055 http://twitter.com/f055/status/25341523067404291
00:12 GitHub114 joined #mojo
00:12 GitHub114 mojo: master Sebastian Riedel * ec199b0 (1 files in 1 dirs): made attribute generator a little more versatile - http://bit.ly/freyoP
00:12 GitHub114 left #mojo
00:15 willem left #mojo
00:19 ashleydev left #mojo
00:22 megatron left #mojo
00:22 alnewkirk left #mojo
00:28 xantus[] joined #mojo
00:28 DaTa just played a bit with perlrocks.pl, it can receive github's post service hook (basically) :) http://data.rbfh.de/p/2011-01-13TwZedE7J
00:29 xantus well, time to shut down this server
00:29 xantus left #mojo
00:31 xantus[] now that I have a stable 50mb connection at home...
00:32 ispy_ joined #mojo
00:32 sri DaTa: heh, that looks like fun
00:33 * sri is jealous of xantus
00:33 * sri can only get 3mb atm :(
00:34 DaTa sri: yes it is :)
00:35 DaTa but it spam's a lot when there are more commits on a push, there is a risk of an excess flood
00:36 sri not unlike twitter :)
00:37 baton8 joined #mojo
00:37 * sri needs sleep
00:37 sri nn
00:38 DaTa dito
00:39 ispy_ left #mojo
00:48 ashleydev joined #mojo
00:52 daviddelikat1 joined #mojo
00:54 marty left #mojo
00:57 daviddelikat left #mojo
01:00 ispy_ joined #mojo
01:08 avkhozov left #mojo
01:08 avkhozov joined #mojo
01:10 ispy_ left #mojo
01:15 ispy_ joined #mojo
01:21 perlrocks Twitter: "@dbbolton What framework (if any) are you using in Perl? Mojolicious looked nice, the last time I looked. http://www.mojolicious.org/" --peteraronoff http://twitter.com/peteraron​off/status/25361777671798784
01:23 megatron joined #mojo
01:23 alnewkirk joined #mojo
01:44 verrens left #mojo
01:58 xantus[] ;)
02:36 ispy_ left #mojo
02:41 ryan left #mojo
02:53 perlrocks Twitter: "@ Cnhacktnt @ fenng @ virushuo @ kdolphin framework can be considered Dancer, Mojolicious or simply get Plack + Moose, Catalyst that it is too heavy" (zh-CN) --nightsailer http://twitter.com/nightsail​er/status/25385021229498368
03:42 avkhozov left #mojo
04:08 ashleydev left #mojo
04:17 ispy_ joined #mojo
04:26 ashleydev joined #mojo
04:36 megatron left #mojo
04:36 alnewkirk left #mojo
05:29 su-bzero-0 is now known as su-bzero[A]
05:48 ashleydev left #mojo
05:53 su-bzero_ left #mojo
05:56 kaare joined #mojo
05:58 su-bzero joined #mojo
06:00 su-bzero[A] is now known as su-bzero`
06:01 su-bzero` is now known as su-bz3ro
06:01 su-bzero_ joined #mojo
06:09 ashleydev joined #mojo
06:21 koban joined #mojo
06:26 ltriant left #mojo
06:32 su-bzero1 joined #mojo
06:32 su-bzero2 joined #mojo
06:34 ashleydev left #mojo
06:36 su-bzero_ left #mojo
06:36 su-bzero left #mojo
06:40 su-bzero1 left #mojo
06:40 su-bzero2 left #mojo
06:50 su-bz3ro is now known as su-bzero[A]
07:34 Christian joined #mojo
07:34 Christian morning
07:58 ysyrota joined #mojo
08:06 su-bzero[A] is now known as su-bzero
08:07 breaker313 joined #mojo
08:10 zipkid Hoi
08:17 marcus http://www.dagolden.com/ go chansen
08:18 su-bzero is now known as su-bzero2
08:18 su-bzero2 is now known as su-bzero
08:25 fhelmberger joined #mojo
08:28 marcus RAWR
08:43 chansen =)
09:12 spleenjack joined #mojo
09:20 marcus sri: I see you've reinvented Mouse :)
09:22 su-bzero_ joined #mojo
09:23 su-bzero_ __PACKAGE__ will work?
09:29 Christian afaIk
09:31 fhelmber_ joined #mojo
09:33 verrens joined #mojo
09:37 sri moin
09:38 sri marcus: in about 7 lines of code!
10:08 breaker313_ joined #mojo
10:12 breaker313 left #mojo
10:19 GitHub45 joined #mojo
10:19 GitHub45 mojo: master Sebastian Riedel * 43de411 (2 files in 2 dirs): simplified Mojo::Base a bit - http://bit.ly/dEm75n
10:19 GitHub45 left #mojo
10:19 yko sri: if 'has' show itself good, what your plans about attr? deprecate and remove at 2.0 or support for a longer time?
10:20 sri don't see a reason to deprecate attr
10:20 sri not like __PACKAGE__->attr(...) has become useless
10:20 sri or $instance->attr(...)
10:21 yko uhmm. sounds good. my first thought was that youll remove sooner or later
10:31 sri i'm quite sure "has" will stay btw. :)
10:32 sri maybe i'll rename it to "haz" before release, but thats it :D
10:33 yko me too, but people are a bit scared and wondering if you are going to remove attr in nearest future
10:33 yko lol. than it will goes to ojo-like stuff ;D
10:35 perlrocks Twitter: "For those wondering, the attr method will not be removed from Mojo::Base, it is still useful. ;) #mojolicious #perl" --kraih http://twitter.com/kraih/status/25501080779497472
10:35 yko thanks
10:42 * sri wonders why Moose requires is => 'rw'
10:46 moritz sri: about which part do you wonder? about defaulting to ro? or about the syntax?
10:50 sri oh, actually i just got confused by Moo
10:50 sri Moo requires "is" to be present
10:53 janus left #mojo
10:53 janus joined #mojo
11:27 forwardever joined #mojo
11:39 forwardever left #mojo
11:41 forwardever joined #mojo
11:46 alnewkirk joined #mojo
11:46 megatron joined #mojo
11:55 breaker313_ left #mojo
11:55 breaker313__ joined #mojo
11:57 forwardever left #mojo
11:58 forwardever joined #mojo
12:04 kaare left #mojo
12:04 kaare joined #mojo
12:06 * sri wonders if :base would be better than -base
12:07 perlrocks Twitter: "Mojolicious - Web framework for Perl http://www.findbestopensour​ce.com/product/mojolicious  #web-framework" --bestopensource http://twitter.com/bestopenso​urce/status/25524349209411584
12:11 GitHub32 joined #mojo
12:11 GitHub32 mojo: master Sebastian Riedel * ab06f03 (1 files in 1 dirs): cleanup - http://bit.ly/fgcEr7
12:11 GitHub32 left #mojo
12:14 marcus all your :base
12:15 yko ...now belong to us
12:16 marcus what is Moo ?
12:16 purl Moo!
12:16 marcus MOO!
12:20 sri MOO!
12:35 marty joined #mojo
12:40 esskar joined #mojo
12:40 esskar hello
12:41 * sri waves
12:41 esskar long time no see
12:41 purl No...I've been watching you. That thing you do with the candle is most impressive.
12:41 sri :)
12:41 * esskar pets purl
12:41 purl o/` purrrrrr o/`
12:42 esskar good job on mojolicous btw
12:43 esskar it's really getting better and better
12:43 esskar i'm back in grmany btw
12:43 esskar s/grmany/germany/
12:48 sri welcome home, the cold must have been a shock :)
12:51 esskar yes
12:51 esskar i think the swine flu caught me before christmas
12:52 esskar i was "ausser gefecht" for almost 3 weeks
12:53 Christian is there a method to encode utf8 to latin1? waht i found is that: http://search.cpan.org/dist/Mojolicious/lib​/Mojolicious/Guides/Rendering.pod#Encoding
12:54 sri Christian: what for?
12:54 purl for fun.
12:54 * sri bites purl
12:54 * purl bites sri back
12:55 esskar btw. MOJO_RELOAD=1 nolonger effects the templates
12:55 esskar is there a new option?
12:55 esskar could not find it
12:55 Christian here is a database on latin and my data are in utf8 so i have hieroglyphs in my database if there are umlauts
12:56 moritz Christian: what kind of database?
12:56 moritz most DB drivers support an automatic encoding and decoding option
12:56 Christian mysql
12:56 moritz anyway, 'use Encode;' => you should be fine
12:57 sri yea, thats normal perl stuff
12:57 moritz the search the DBD::mysql docs if it can automate it for you
12:57 Christian moritz, ok i will read a little bit ;)
12:57 Christian sri, is there a mojo method for encoding
12:58 sri there is but you wouldn't really benefit from it
12:58 sri Encode or a DBD driver option should be all you need
12:59 sri like i said, this is plain perl stuff
12:59 sri esskar: do you have a test case?
13:00 sri because --reload works just fine here for templates
13:00 esskar external templates in /templates ?
13:00 sri yes
13:00 esskar hmm
13:00 esskar will give you a test case
13:02 breaker313_ joined #mojo
13:02 Christian ok thx moritz and sri
13:03 sri esskar: you mean lite + external template?
13:03 esskar sri, yes
13:03 sri hmm, there could be a problem
13:03 * esskar nodes
13:07 breaker313__ left #mojo
13:08 sri ok, this is weird
13:08 breaker313_ left #mojo
13:09 sri it seems if you change the lite app once all template changes will be picked up
13:09 jdixon sri: didn't I mention something like that a couple weeks ago?
13:10 lee7 joined #mojo
13:10 jdixon hmm, no
13:10 esskar :D
13:10 jdixon that was something else
13:10 jdixon the mysql AutoCommit issue
13:10 esskar i remember that it has been like this a couple of month ago
13:10 sri i'm sure this can be fixed in Mojo::Loader, just not sure how
13:10 esskar then you fixed it
13:10 esskar now it's broke again
13:10 esskar seems to be an unstable feature
13:11 sri of course it is
13:11 sri reloading is just crazy
13:11 esskar tschaka
13:13 sri ok, i'm stupid
13:13 jdixon I beg to differ.  ;)
13:14 sri wait till you see the fix -.-
13:14 moritz reloading a DBIx-Class result class often leads to "inconsistent c3 linearization" or so
13:15 moritz but I guess that's just the way things are
13:15 GitHub64 joined #mojo
13:15 GitHub64 mojo: master Sebastian Riedel * 46ccafe (2 files in 2 dirs): fixed --reload delay problem - http://bit.ly/eR81WB
13:15 GitHub64 left #mojo
13:16 esskar nice
13:16 sri should have been obvious :S
13:17 esskar t/mojo/loader.t ............................... 1/36
13:17 esskar #   Failed test 'reloaded successfully'
13:17 esskar t/mojo/loader.t ............................... 36/36 #   at t/mojo/loader.t line 87.
13:17 esskar #          got: '23'
13:17 esskar #     expected: '26'
13:17 esskar # Looks like you failed 1 test of 36.
13:17 esskar was kicked by sri: esskar
13:17 esskar joined #mojo
13:17 * esskar waves
13:18 sri never paste into an irc channel
13:19 esskar i just slipped and fell
13:19 sri heh
13:20 sri the test passes here though
13:21 esskar well, it fixes the bug here
13:22 sri which operating system?
13:23 sri (the test fail)
13:26 esskar windows 7
13:26 purl i heard windows 7 was out, which has a lighter base install, and more inline stuff so they can limit piracy or same song, different verse
13:26 sri oh ok
13:27 sri not much i can do then i'm afraid
13:28 esskar where is MojoTestReloader defined?
13:29 sri temporary file
13:30 esskar ah
13:30 esskar now i see it
13:38 ashleydev joined #mojo
13:40 esskar sri, the problem is, that the temp MojoTestReloader.pm is not in %INC
13:43 sri ok, but why does it only affect windows?
13:44 daviddelikat1 left #mojo
13:44 sri push @INC, $dir; require MojoTestReloader;
13:44 sri that has to be broken on windows
13:44 esskar i know; looks like it
13:48 esskar funny
13:48 esskar if i dump %INC in loader.t it's in there
13:48 esskar but not when i dump in Mojo::Loader->reload
13:49 sri smells like broken perl
13:51 esskar wait a minute
13:53 esskar :D
13:55 marcus the broken windows theory
13:55 marcus ( http://en.wikipedia.org/wiki/Broken_windows_theory )
13:56 sri lol
13:56 sri you can even apply it to the operating system
13:57 esskar found it
13:57 sri microsoft is obviously responsible for all internet crime
13:57 esskar it's in %INC
13:58 esskar but what's also in %INC is, is  t/mojo/loader.t
13:58 esskar and this seems to be modified (for a reason i do not understand)
13:59 esskar and the require fails
13:59 esskar and reloading is interupted
13:59 sri oh wait
14:00 esskar Require failed You tried to plan twice at t/mojo/loader.t line 6.
14:00 sri https://github.com/kraih/mojo/bl​ob/master/lib/Mojo/Loader.pm#L62
14:00 sri this is why it thinks loader.t got changed i guess
14:01 sri but i don't understand why it's only a problem on windows
14:01 esskar yes
14:01 sri it does not plan twice on os x
14:01 sri even if run directly
14:02 esskar setting it to 0 and checking if ($STATS->{$file} && $mtime > $STATS->{$file}) {
14:02 esskar would that make sense?
14:02 sri not at all
14:02 sri would break everything again
14:02 esskar hmm
14:03 sri realoading $0 has to happen every single time
14:03 sri otherwise we can't recover from syntax errors in a lite app
14:03 esskar i see
14:04 sri i guess you understand now what i meant when i said reloading is crazy :D
14:05 sri can someone please test this on linux and possibly other unix operating systems?
14:06 sri latest github master, just run the tests
14:07 sri fine on ubuntu at least
14:07 esskar whats yout Test::More version?
14:07 sri i ran it manually under 5.8.9 5.10.0 and 5.12.0
14:07 yko ./t/mojo/loader.t fails
14:08 yko ubuntu, v5.10.1
14:08 sri are you sure?
14:08 purl You still have ALL THREE lifelines left!
14:08 sri it passed on my ubuntu box with 5.10.1
14:08 yko sri moment
14:08 esskar This is perl 5, version 12, subversion 2 (v5.12.2) built for MSWin32-x86-multi-thread
14:08 jdixon t/mojo/loader.t works on openbsd with v5.10.1
14:09 yko no, that was fake, sorry
14:09 sri ;p
14:10 sri ok, then it's really only windows
14:10 yko pass all tests on ubuntu,  v5.10.1
14:11 jdixon is IO::Socket::SSL 1.3.7 required for just the test, or all SSL features?
14:11 jdixon (separate question)
14:11 sri all ssl features
14:11 jdixon damn
14:11 sri you can't do ssl without openssl
14:12 jdixon no, it has it.  but only 1.3.3
14:12 yko hmm. fails on WinXP v5.12.2
14:12 jdixon er, 1.33
14:12 sri ah
14:12 jdixon (openbsd)
14:12 sri ye, IO::Socket::SSL had many bugs
14:15 sri esskar: what does $0 say right before the ->reload call?
14:15 sri i suspect there could be a \ bug
14:16 sri for us it is t/mojo/loader.t
14:17 esskar t/mojo/loader.t
14:17 yko also t/mojo/loader.t on xp
14:17 sri ok, out of ideas then
14:17 yko just at line 86 of .t
14:18 esskar yes
14:18 esskar strange
14:18 esskar well
14:18 esskar everything else is fine
14:19 gabiruh left #mojo
14:19 sri do we have to deactivate the loader test on windows too?
14:20 ispy_ left #mojo
14:25 sri yko: broken on xp too?
14:25 sri activeperl and strawberry?
14:26 gabiruh joined #mojo
14:26 yko activeperl
14:26 purl On the day Perl becomes "Active", Satan will be skating to work. or ActiveState's port of perl plus other so-called "goodies" or it looks like Satan is skating to work. or available from activestate or at http://www.activestate.com or not Free Software or Free Software or (qv 'lwall on activeperl') or http://activestate.com/Products/Do​wnload/Register.plex?id=ActivePerl
14:27 yko Binary build 1202 [293621] provided by ActiveState
14:27 sri broken?
14:27 purl broken is always relative
14:27 * sri kisses purl
14:27 * purl swoons!
14:28 yko not ok 36 - reloaded successfully got: '23' expected: '26'
14:28 yko sorry for bold
14:28 sri ok
14:28 sri think i'll disable the test :S
14:28 yko until somebody fix it :)
14:29 sri yea
14:33 jdixon sri: I know you don't need more headaches, but fyi --reload still isn't working here. And I'm not using mysql/autocommit at all.  :(
14:33 daviddelikat joined #mojo
14:33 jdixon I'm sure it's something I'm doing.  grrr.
14:34 sri nothing i can do without a test case
14:34 jdixon yeah I know
14:43 GitHub136 joined #mojo
14:43 GitHub136 mojo: master Sebastian Riedel * 32ef1bb (1 files in 1 dirs): disable loader tests on windows - http://bit.ly/hbITno
14:43 GitHub136 left #mojo
14:44 sri we could use a windows expert to review all tests we disable on windows
14:44 * esskar hides
14:45 chansen windows experts using Perl?
14:45 chansen haha
14:45 * chansen hides
14:46 esskar sri, isn't t/mojo/loader.t reloaded on your systems?
14:46 sri esskar: nope
14:47 esskar but you set $STATS{$0} = 1
14:47 jfuller sri: The problems with windows is that it isn't consistent across machines. i.e. I get different results on the Plack tests from machine to machine
14:47 chansen sri: why don't you fire up your vmware? ;P
14:48 esskar so $mtime > $STATS->{$file} is true
14:48 * sri does not own a single windows license anymore
14:50 jfuller sri: Not a single windows license? Where's the challenge in that?
14:51 chansen sri: do you want one? ;P
14:51 sri no thanks ;p
14:51 chansen lol
14:52 marcus :D
14:52 marcus wasnt microsoft supposed to start providing cloud servers for perl developers?
14:52 chansen sri: I have 10-pack, I could send you one ;P
14:52 sri i prefer to keep my excuse :D
14:53 chansen be nice otherwise I'll send you one and you wont have excuse ;P
14:53 sri :o
14:54 chansen don't worry, I won't push such a burden on you ;P
14:54 sri esskar: "t/mojo/loader.t -> t/mojo/loader.t modified, reloading!"
14:55 sri esskar: you can set MOJO_LOADER_DEBUG=1 to see whats happening
14:55 sri so it is in fact reloading the test script, just nothing goes wrong
14:56 sri must be some safety switch not working on windows or so
14:58 esskar MOJO_LOADER_DEBUG is turned on
14:58 sri http://pastie.org/1456421
14:59 sri thats the full output
14:59 megatron left #mojo
14:59 alnewkirk left #mojo
15:02 sri http://venturebeat.com/2011/01/12/google-search/ # interesting
15:02 sri what if facebook joined the web search game
15:02 sri using social data to rank links
15:05 sri i suspect twitter could be a good source
15:05 sri then again spammers are gaming twitter too
15:06 yko in two years fecebook became a garbage collector
15:06 yko people vs bots will be 1:3 or so
15:07 sri facebook is a huge investment bubble, it sure will explode at some point
15:07 sri but social web search can work with a good reputation system
15:08 yko i'm not telling that using social as data source is bad, but it will bring much more bots there
15:24 ashleydev left #mojo
15:35 barwin joined #mojo
15:43 Christian nice evening to you all
15:43 Christian cu
15:46 Christian left #mojo
15:53 ysyrota left #mojo
16:01 koban left #mojo
16:06 GitHub156 joined #mojo
16:06 GitHub156 mojo: master Sebastian Riedel * af761b2 (16 files in 10 dirs): cleanup - http://bit.ly/hR8oL5
16:06 GitHub156 left #mojo
16:10 GitHub141 joined #mojo
16:10 GitHub141 mojo: master Sebastian Riedel * 0323ba0 (1 files in 1 dirs): fixed typo - http://bit.ly/fnETud
16:10 GitHub141 left #mojo
16:11 esskar sri, can you check the return value auf ->reload in your test
16:11 esskar i notice, that MojoTestReloader.pm runs before t/mojo/loader.t
16:11 esskar in my case, it always runs before it
16:12 esskar and it throws $@
16:12 sri yes, it dies
16:12 sri "You tried to plan twice at t/mojo/loader.t line 9."
16:12 esskar exactly
16:13 sri "Compilation failed in require at /usr/local/lib/perl5/site_​perl/5.12.0/Mojo/Loader.pm line 87."
16:13 esskar same hre
16:13 esskar so, it's not a windows "bug"
16:13 sri ?
16:13 sri of course it is
16:13 esskar nope
16:14 sri if it doesn't pass it's a windows bug
16:14 sri or did i miss something?
16:15 esskar it fails on your system; it only passes because t/mojo/loader.t is executed after MojoTestReloader.pm
16:15 sri ok, but how come it only affects windows?
16:16 esskar you rely on the order of the iteration over an hash
16:16 esskar i could simple fix the test
16:16 esskar wait
16:22 esskar done
16:25 chansen sri: if you need a smoke on win32, I could fire up my wm
16:26 sri esskar: mind sharing?
16:26 sri chansen: thanks, in a few mins :)
16:26 chansen np =)
16:28 esskar one sec
16:28 sri esskar: you don't need to send a pull request if it's simple
16:30 esskar i know
16:30 * sri expects a "die if $ENV{PLAN...yada yada};" or so :)
16:31 sri or actually a clean exit
16:40 forwardever left #mojo
16:47 sri guess the fix isn't that simple
16:47 esskar yes, it is
16:47 esskar sorry
16:47 esskar had to fix a problem away from perl
16:47 esskar some guy in the office is bugging me
16:50 yko is it possible that mojolicious call controller's actions with more than 1 argument  (controller instance) ?
16:51 sri does it?
16:51 yko i mean, no plan for something like that?
16:51 yko no it doesn't
16:51 esskar sri, http://pastie.org/1456810
16:51 esskar sorry, ugly formating
16:51 sri ok, no plans for that
16:51 yko sri, thx
16:52 sri esskar: what did you change and why?
16:52 esskar well, i changed, that $0 is reloaded at the end
16:52 esskar and does not interupt loaded modules
16:52 sri hmm
16:52 esskar so the order is always, load models
16:52 esskar then the script
16:52 purl then the script is screwed
16:53 esskar so
16:53 esskar it's the same as in your case
16:53 esskar you are just lucky
16:54 esskar have you checked, if $0 is really always the last in the iteration?
16:55 sri your paste is bugged in at least two ways
16:55 esskar ?
16:55 sri i'll investigate
16:55 esskar why bugged?
16:56 esskar fuck
16:56 esskar sorry
16:56 esskar wait
16:56 sri i'm on it now
16:59 esskar the last line in reload should be
16:59 esskar return $self->_reload( $0, $0 );
16:59 esskar of course
17:00 esskar everything else is for sure fine
17:01 sri $INC{$0} = $0; is wrong too
17:02 esskar thats not me
17:02 esskar thats you
17:02 esskar https://github.com/kraih/mojo/​blob/master/lib/Mojo/Loader.pm
17:02 sri ?
17:03 sri it is responsible for the random element in the first place
17:04 esskar but you did add it so, you will get it in your loop
17:04 esskar if you do not add it, $0 will not get reloaded
17:04 ashleydev joined #mojo
17:04 esskar you means YOU
17:05 esskar just checked it
17:05 esskar you can remove it
17:05 esskar if you do it, you may also remove next if $file eq $0;
17:05 esskar but you have to call return $self->_reload( $0, $0 );
17:05 sri i have not the slightest idea what you're talking about :S
17:06 sri anyway, fixing
17:06 esskar why not?
17:06 esskar at least, it's not a windows bug *puuuuuh*
17:06 esskar it's just a coincidence
17:07 esskar that it was passing on mac and nix
17:11 marcus Some guy was mailing me about writing a proxy plugin. Told him to come here.
17:21 * chansen backlogs
17:26 chansen esskar: next time, why don't you assemble a test-case instead of involving whole channel in a non-existant bug chase?
17:33 GitHub169 joined #mojo
17:33 GitHub169 mojo: master Sebastian Riedel * 8e84af2 (2 files in 2 dirs): enabled loader tests on windows again - http://bit.ly/fwY4oQ
17:33 GitHub169 left #mojo
17:36 sri ok, that should do the trick
17:40 ashleydev left #mojo
17:47 sri please test! *hint hint*
17:48 * sri goes to get a cold beer
17:50 josh is there some timeout on a chunked response? eg, $c->write_chunk('starting long running task'); do_something(); $c->write_chunk('this never goes to the client')
17:52 yeo joined #mojo
17:53 yko ubuntu pass, xp pass
17:53 sri server has an inactivity timeout
17:53 sri \o/
17:53 josh can i change the timeout at runtime?
17:53 megatron joined #mojo
17:53 alnewkirk joined #mojo
17:54 sri yes there's a hack
17:54 josh or alternately i can just send some sort of ping to the client
17:54 josh but i tried with $c->client->ioloop->timer and didnt seem to have any luck
17:54 sri $self->client->async->ioloop->conne​ction_timeout($self->tx->connection => 300); or so
17:55 josh ooh excellent, thanks!
17:56 sri i guess you could also $self->write(...) some garbage the chunked parser will ignore
17:57 spleenjack left #mojo
17:58 josh hmm ill try it as a back up... id rather keep the timeout normally, and then just disable it for my long-running requests
17:58 sri interesting, all hashes in ruby 1.9 are preserve order :o
17:58 sri *-are
18:06 sri http://rim.jobs # lol
18:07 sri best corporate domain ever
18:12 snearch joined #mojo
18:17 avkhozov joined #mojo
18:19 fhelmber_ left #mojo
18:25 fhelmber_ joined #mojo
18:30 sri damn...firefox 4 will be released without websockets :(
18:33 jdixon wtf
18:37 daviddelikat left #mojo
18:59 marcus left #mojo
19:03 pid joined #mojo
19:12 jdixon sri: http://tools.ietf.org/rfcdiff?url2=dr​aft-ietf-hybi-thewebsocketprotocol-04
19:12 jdixon hopefully that motivates FF/Opera to do something
19:12 sri nope
19:13 sri they won't do anything until the spec is finished
19:13 jdixon sigh
19:13 sri and nobody is in a hurry
19:14 sri now they are turning websockets into a crypto protocol
19:14 sri there is so much bullshit on that list
19:15 sri the reason they are pushing the move to a crypto protocol is to avoid broken proxies interpreting frames as http messages
19:15 jdixon right
19:15 ryan joined #mojo
19:15 jdixon it's stupid
19:16 sri like they have totally lost touch with reality
19:16 jdixon doesn't look like they're changing from Upgrade to Connect as suggested
19:16 jdixon tossing more crypto in the fire
19:17 sri i've only seen like one smart post during the last few months
19:17 jdixon oh well, it's not like anyone uses FF anymore, right?  ;)
19:17 sri chrome is unpopular too now
19:18 sri since they removed h.264
19:18 sri it's all just corporate agenda bullshit
19:18 sri html5 started so well...
19:19 jdixon sri: I like you. You're starting to sound like de Raadt.  :)
19:19 sri lol
19:20 janus must be the jdixon i know :)
19:20 jdixon hehe
19:20 jdixon hi simon
19:20 jdixon didn't recognize your nick :)
19:21 janus hey jason
19:22 jdixon are you going to slovenia?
19:22 janus nope, didn't even know
19:23 janus i need to get active again...
19:23 jdixon me too  :(
19:24 janus to update the mojo port unless someone else did it already
19:25 jdixon nope, you're the last
19:25 jdixon 0.991237
19:26 janus :(
19:48 daviddelikat joined #mojo
19:58 pid left #mojo
19:59 pid joined #mojo
20:11 marcus joined #mojo
20:38 ok22 joined #mojo
20:41 ok22 happpy Old New Year!
20:41 marcus urgh
20:42 avkhozov left #mojo
20:42 avkhozov joined #mojo
20:43 marcus ok: snovim godom!
20:43 marcus ok22 even
20:44 marcus that charlie guy who wants me to "mentor him in trying to understand the transaction code in mojolicious". says he tries to avoid irc or instant messaging.
20:44 marcus I'm tempted to tell him I try to avoid email
20:46 sri marcus: charlie from the list?
20:46 marcus sri: yes
20:47 sri ah, think i told him to stop writing me mails directly too
20:47 marcus sri: he mailed me personally to ask me to help him to write a full proxy
20:47 marcus so I told him to come here and we could try helping him
20:48 sri bit of a help vampire
20:48 marcus "Sorry, I run a busy consulting business and do not have time for one-on-one free support. I suggest continuing to ask on the mailing lists if IRC is out of the question for you."
20:49 marcus hmm, new iTerm is pretty neat
20:50 marcus omega++ # for that recommendation
20:50 marcus I like how you can split terminal windows too :)
20:50 sri hmm
20:50 sri not tried the new one yet
20:50 marcus icon is much prettier too =)
20:51 ok22 left #mojo
20:52 ok22 joined #mojo
20:52 jegade how could i define a global block or include in Mojo::Template
20:56 yko jegade: why don't you just define aprtial template to include? http://mojolicio.us/perldoc?Mojolicious/Gu​ides/Rendering#Including_Partial_Templates
20:58 jegade yko: ah, ok - that looks good, ty
21:01 sri oh, 256 colors, i'm sold :D
21:02 omega marcus: yiha :)
21:04 snearch left #mojo
21:18 marcus sri: "[sri has accused me of "spamming" the mailing list, hence my keeping this
21:18 marcus off-list.]"
21:18 marcus :D
21:21 marcus I'm so happy about maintaining inbox 0 for the last week
21:25 ltriant joined #mojo
21:34 sri :)
21:38 avkhozov left #mojo
21:38 fhelmber_ left #mojo
21:43 sri oh, i like fullscreen mode
21:47 marcus yeah
21:47 marcus and you can save a window layout
21:48 marcus including tabs and split windows and stuff
21:48 marcus so once you've set it up the way you like it, you can save it
21:50 jdixon how *do* you do that?
21:51 jdixon I can't seem to find any docs on saving sessions
21:55 marcus jdixon: in iterm2?
21:55 marcus windows -> save window arrangement
21:57 tempire_ joined #mojo
22:07 * sri wonders why there is no better email client for os x yet
22:09 tempire_ there is
22:09 tempire_ gmail
22:09 tempire_ mailplane, even
22:12 * sri can't stand the google design
22:20 marcus <3 mailplane
22:21 barwin left #mojo
22:21 jdixon marcus: ah, I was looking at iterm
22:47 sri Zzz
22:59 kaare left #mojo
23:01 daviddelikat left #mojo
23:19 tempire_ left #mojo
23:21 jdixon marcus: I'm not crazy about the font rendering in iTerm2. Wonder why it looks so different than iTerm/Terminal.
23:22 jdixon ah
23:22 jdixon antialiasing
23:22 purl hmmm... antialiasing is the electronic simulation of a blurry display
23:22 jdixon weird, didn't have the same effect in iTerm
23:29 daviddelikat joined #mojo
23:31 tempire_ joined #mojo
23:32 tempire_ Somewhere along the line, I got the idea that core perl modules would be uber clean code.
23:33 tempire_ Either I'm getting better or more critical.
23:33 tempire_ maybe both
23:33 purl maybe both are wrong.
23:33 tempire_ indeed

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