Camelia, the Perl 6 bug

IRC log for #mojo, 2013-08-25

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

All times shown according to UTC.

Time Nick Message
00:38 btyler joined #mojo
00:44 good_news_everyone joined #mojo
00:44 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/ecoW4Q
00:44 good_news_everyone mojo/master f233adc Sebastian Riedel: mention default ports
00:44 good_news_everyone left #mojo
01:09 basic6_ joined #mojo
01:20 hotsnow joined #mojo
01:36 hotsnow joined #mojo
01:39 jnbek joined #mojo
01:41 KindTwo joined #mojo
01:54 asarch joined #mojo
02:04 hotsnow joined #mojo
02:40 phillipadsmith I'm getting an error when using Mojolicious::Plugin::JSONP, 'Can't locate object method "render_json"', any idea what I'm doing wrong here? https://gist.github.com/philli​padsmith/edd40a94a98c0cd0fc69
03:18 preflex_ joined #mojo
03:20 salparadise phillipadsmith: maybe your mojo version?
03:21 phillipadsmith salparadise: mojo version is newer than the plugin expects?
03:21 phillipadsmith Mojolicious is up to date. (4.26)
03:24 salparadise oh plugin, nm im sleepy
03:42 arpadszasz joined #mojo
04:37 arpadszasz joined #mojo
05:36 yakudza joined #mojo
06:09 stephenenelson joined #mojo
06:10 stephenenelson Can you put subdirectories in the public/ folder? I'm getting weird 403s for files I know are there.
06:10 stephenenelson 404s even.
06:17 moritz works fine here
06:19 stephenenelson Strange. I have files in a js subdirectory. One file will get served just fine. Another one will give me a 404. I am unquestionably doing something stupid.
06:32 stephenenelson …and I was. Good night.
06:32 stephenenelson left #mojo
06:41 Vandal joined #mojo
06:52 ilbot2 joined #mojo
06:52 Topic for #mojo is now 🎩, indubitably | http://mojolicio.us | http://irclog.perlgeek.de/mojo/today
07:26 marcus I just read scrollback, and I have one thing to add:
07:26 marcus NERDS!
07:28 marcus Also, i don't quite understand why we want SO_REUSEPORT - Tried to read about it on google, and I understand you can reuse the port, but I still don't get the use case quite.
07:30 dqw joined #mojo
07:45 dotan joined #mojo
08:11 d4rkie joined #mojo
08:12 d4rkie joined #mojo
09:09 arpadszasz joined #mojo
09:18 basiliscos joined #mojo
09:19 sh4 joined #mojo
09:20 asdf joined #mojo
09:21 asdf What does perl mojoliteapp daemon -l localhost:80  do? cause its not the same as -l http://localhost:80/
09:28 dqw joined #mojo
10:05 denisboyun joined #mojo
10:22 dqw joined #mojo
10:24 hummeleBop joined #mojo
11:12 bowtie_ joined #mojo
11:22 basiliscos joined #mojo
11:26 basiliscos1 joined #mojo
11:46 sri marcus: cheap zero downtime upgrades of course
11:46 sri you can run ./myapp.pl daemon -l http://*:3000?reuse=1 while another daemon is still active, then stop the old one and the new one will take over without any noticeable downtime
11:47 sri under linux both can actually serve requests concurrently
11:47 sri the kernel does load balancing for you
11:53 batman that's just crazy.
11:53 batman :)
11:59 espent joined #mojo
12:05 sri i suppose if nobody actually understands SO_REUSEPORT, maybe we shouldn't support it
12:08 sri http://freeprogrammersblog.vhex.net/post/linux-3​9-introdued-new-way-of-writing-socket-servers/2 # article with examples again
12:10 poopy joined #mojo
12:10 poopy how do you stop/kill morbo, ctrl+c doesn't seem to kill it, while ctrl+c does kill daemon
12:11 batman poopy: i think you got some evil code blocking the signal
12:12 batman ctrl+c works like a charm here, unless when i have a nasty bug that keeps the process from receiving the signal
12:12 batman i usually do ctrl+d, kill -9 $SOME_PID and then i try to debug my way to finding out why the signal did not get through
12:13 batman ooops! s/ctrl-d/ctrl-z/ :S
12:13 poopy no evil code, just the most basic lite app, one template, i'm debugging some jquery stuff, i'm on win32, and like I said ctrl+c kills the daemon
12:16 sri win32 perl is evil code
12:16 poopy and they call me poopy
12:17 sri some versions of it work, others don't, but all have weird bugs
12:18 dotan joined #mojo
12:21 poopy oh, its probably the waitpid, it waits forever
12:22 poopy or something like that (looking for perlmonks thread now)
12:24 sri morbo works on windows, just not yours
12:25 sri so, something that morbo depends on is broken in your version of perl
12:25 poopy sri: sure, nothing wrong with morbo , its just me thats special and my version of perl
12:26 marcus sri: I thought hypnotoad already handled that for you :)
12:26 marcus sri: would make more sense if hypnotoad just used it if available.
12:27 sri marcus: in a very different much more complicated way
12:27 sri marcus: it doesn't work that way
12:27 marcus bummer
12:27 * marcus is coding ruby today.
12:27 sri this is simple zero downtime for *all* our servers
12:27 marcus Their documentation culture sure do suck compared to our
12:27 marcus <3 perlfo Foo
12:27 marcus perldoc Foo even
12:28 poopy ha, I just proved something, it takes 2-5 requests after morbo says it reloaded my mojolite app, before this mojolite app starts serving requests
12:28 marcus wanted to find out what methods Rack::MockResponse provided, and it was such a hassle.
12:31 * poopy back to away thanks for noting :)
12:35 human39 joined #mojo
12:40 sh4|2 joined #mojo
12:41 espent joined #mojo
12:46 peaches joined #mojo
12:56 sri marcus: you get the SO_REUSEPORT use case now?
12:56 marcus sri: yes.
12:56 marcus sri: I'd probably not use it tho, since hypnotoad does this for in a cross-platform way.
12:57 sri some people don't like putting processes in the background though, reason we have the prefork command now
12:57 marcus sri: why would I *not* use hypntoad in production? :)
12:57 marcus people are weird
12:58 marcus hypnotoad even
12:58 sri you could use ./myapp.pl prefork -l http://*:3000?reuse=1 now and get zero downtime
12:58 sri marcus: reason mostly stated is automation with chef/puppet
12:58 marcus sri: and what would it gain me over hypnotoad-f  myapp.pl  ?
12:58 sri configure the prefork daemon from the outside
12:59 sri marcus: you can't have zero downtime with -f
12:59 marcus sri: good point.
12:59 marcus sri: I'm sure it'll be useful to someone.
12:59 marcus hmm, maybe I'll even use it with docker
13:00 marcus Although I don't really need zero downtime inside the containers as long as I have a load balancer in front.
13:00 sri zero downtime in the cloud could be another use case
13:01 sri whenever your daemon needs to stay in the foreground
13:01 marcus sri: running apps with supervisord in the foreground is a pretty common scenario
13:06 sri so, we are in favor of keeping it now?
13:06 sh4 joined #mojo
13:09 marcus sri: +1 from me
13:13 basiliscos joined #mojo
13:16 * sri wonders how well linux can balance between two listen sockets
13:17 sri perl -Mojo -E 'a({text => "FIRST!"})->start' daemon -l http://*:3000?reuse=1
13:17 sri perl -Mojo -E 'a({text => "SECOND!"})->start' daemon -l http://*:3000?reuse=1
13:17 sri if anyone has a brand new linux kernel and mojolicious from github, try that :)
13:21 hummeleBop it should work ?
13:23 dqw joined #mojo
13:32 gryphon joined #mojo
13:34 basiliscos joined #mojo
13:41 dqw joined #mojo
13:47 basiliscos joined #mojo
14:01 batman Your vendor has not defined Socket macro SO_REUSEPORT
14:01 batman trying to upgrade IO-Socket-IP
14:01 batman same error message :/
14:02 batman got linux 3.9.4-030904-generic
14:03 sri at least there's a good error message
14:03 batman :)
14:04 batman Your vendor has not defined Socket macro SO_REUSEPORT, used at /usr/local/share/perl/5.14.2/IO/Socket/IP.pm line 472
14:04 batman ^ all the details i got
14:04 sri i've only tested it on os x so far, where it works
14:04 sri you may have to upgrade Socket too
14:05 batman Successfully installed Socket-2.011 (upgraded from 2.009)
14:05 batman same error
14:06 asarch joined #mojo
14:07 batman perl -MSocket -e'SO_REUSEPORT' # Your vendor has not defined Socket macro SO_REUSEPORT, used at -e line 1
14:08 batman super weird
14:08 sri perl -MSocket=SO_REUSEPORT -E 'say SO_REUSEPORT'
14:08 sri 512
14:09 batman SO_REUSEADDR exists
14:09 marcus Your vendor has not defined Socket macro SO_REUSEPORT, used at -e line 1
14:09 sri hmm
14:09 batman which version of linux are you running marcus ?
14:09 marcus Linux beast 3.2.0-4-amd64 #1 SMP Debian 3.2.46-1 x86_64 GNU/Linux
14:10 sri haha
14:10 sri dinosaur!
14:10 marcus That's stable debian, no? :)
14:10 marcus Linux stagolee 3.8.0-26-generic #38~precise2-Ubuntu SMP Thu Jun 20 18:29:36 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
14:10 marcus better?
14:10 batman not really
14:10 sri 3.9+
14:10 marcus any actual linux distros ship with 3.9+ ?
14:10 batman wget http://kernel.ubuntu.com/~kernel-ppa/mainli​ne/v3.9.4-saucy/linux-headers-3.9.4-030904-​generic_3.9.4-030904.201305241545_amd64.deb http://kernel.ubuntu.com/~kernel-ppa/mai​nline/v3.9.4-saucy/linux-headers-3.9.4-0​30904_3.9.4-030904.201305241545_all.deb http://kernel.ubuntu.com/~kernel-ppa/mainli​ne/v3.9.4-saucy/linux-image-3.9.4-030904-ge​neric_3.9.4-030904.201305241545_amd64.deb
14:11 batman who cares? ;)
14:11 marcus I don't think I've compiled linux since the 90s...
14:11 marcus :p
14:11 batman ^ it's deb packages
14:11 batman and another wirc bug....
14:12 marcus for saucy
14:12 marcus hmm
14:12 marcus Anyways, like sri said, works on my mac ;)
14:12 batman http://screencloud.net/img/screenshots​/114b3fef8495c7ad2390332ff0555bf7.png
14:13 marcus mm, only handles one url per string
14:14 marcus /g ? :)
14:14 batman i don't think you're right
14:14 marcus sri: works on freebsd too btw.
14:14 sri \o/
14:14 sri that's apparently where os x stole it
14:17 * marcus & # driving lillehamer -> oslo
14:17 sri batman: perl -Mojo -E 'sub IO::Socket::IP::SO_REUSEPORT { 15 } a({text => "FIRST!"})->start' daemon -l http://*:3000?reuse=1
14:17 sri that might work :)
14:19 sri no actually not... can't redefine a constant
14:19 batman seems to work really nice
14:19 batman THIRD!SECOND!FIRST!SECOND!SECOND!THIRD!FIRS​T!SECOND!SECOND!SECOND!THIRD!SECOND!FIRST!
14:19 sri :O
14:20 sri oh right, it's not a constant in your case
14:20 sri \o/
14:20 batman THIRD SECOND FIRST SECOND SECOND THIRD FIRST SECOND SECOND SECOND THIRD SECOND FIRST SECOND THIRD SECOND FIRST THIRD FIRST THIRD THIRD SECOND THIRD SECOND FIRST THIRD SECOND FIRST FIRST FIRST FIRST SECOND THIRD SECOND FIRST THIRD SECOND SECOND THIRD THIRD SECOND THIRD FIRST FIRST THIRD FIRST THIRD THIR D
14:20 batman ^ stopping now :)
14:21 sri very cool that it load balances
14:21 batman it seems a bit random..?
14:21 sri batman: if you do a wrk load test, is the cpu load even?
14:22 batman how do i do that?
14:22 sri wrk -c 100 -d 10s http://127.0.0.1:3000/
14:24 batman No command 'wrk' found, did you mean: ...
14:24 batman do i get it from github or something?
14:24 batman right.
14:24 sri well, i suppose you can also use ab
14:24 sri wrk is just better
14:26 batman using ab for now...
14:27 denis_boyun joined #mojo
14:28 batman looks like every process is about 100%
14:28 batman ab -c 100 -t 10 http://localhost:3000/
14:30 batman http://screencloud.net/img/screenshots​/fafdef610c1b7d8e79cc9c2e9ce43ad3.png
14:56 sri \o/
15:04 sri would be nice if someone who actually tried perl -MSocket=SO_REUSEPORT -E 'say SO_REUSEPORT' on linux 3.9 could open a ticket
15:04 sri against Socket.pm
16:11 d4rkie joined #mojo
16:15 sri hmm, i think i know how to make SO_REUSEPORT testable
16:16 sri and perhaps a lot more
16:16 sri but it requires adding Mojo::IOLoop::Server::handle and Mojo::Server::Daemon::acceptors
16:21 Britzel joined #mojo
16:22 dod joined #mojo
16:24 * jberger yawns (and moves slowly)
16:24 * jberger was at a fun wedding last night
16:25 Britzel Someone has married just for fun? Wow.
16:25 jberger hehe
16:26 jberger regarding SO_REUSEPORT, it seems useful, are there any potential downsides to conclusions (other than testability, which it looks like sri has an idea about)?
16:27 sri i suppose it could end up not getting used
16:27 jberger that's a matter of adding examples in the docs
16:27 jberger cookbook etc
16:27 sri on the other hand, it's a new way to deploy web apps, so it could also end up being huge
16:28 jberger I will admit, I'm not a sysadmin of any kind, I basically deploy using the examples in the cookbooks verbatim
16:29 sri btw. yes, i can test it now
16:30 jberger the thing I'm working on (once I get the memorize thing out the door) is X-Accel-Redirect for caching/protecting certain static files, and I'm having a hard enough time understanding that
16:30 jberger sri: cool, great work!
16:30 lammel2 joined #mojo
16:31 jberger anyway, all I'm saying is, if you make good cookbook examples, I'm sure it will get used by people like me, who look for to the cookbook for deployment help
16:33 good_news_everyone joined #mojo
16:33 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/Nfje6g
16:33 good_news_everyone mojo/master 64c758c Sebastian Riedel: added tests for SO_REUSEPORT and made servers easier to test
16:33 good_news_everyone left #mojo
16:34 sri besides testing, i don't think Mojo::Server::Daemon::acceptors has many uses
16:34 sri but it makes testing some very tricky stuff possible
16:35 jberger that's enough use-case for me
16:35 jberger :-)
16:35 sri could possibly also add diagnostics events to Mojo::IOLoop::Server
16:36 sri already got Mojo::IOLoop->acceptor($daemo​n->acceptors->[0])->on(accept => sub { ... })
16:36 sri allows you to do "stuff" with the newly accepted socket
16:38 sri $daemon->acceptors is basically the equivalent to $tx->connection
16:38 jberger you're over my head atm, but of course that sounds useful
16:38 jberger oh, that analogy helps
16:39 sri (for listen sockets)
16:39 sri which we call acceptors, because that sounds cool
16:39 jberger sounding cool is the most important thing
16:39 jberger obviously
16:39 sri indubitably
16:40 jberger 🎩
16:40 * jberger is thinks he can keep 'expires' as epoch time, and add 'duration' as delta seconds, both using steady_time under the hood
16:41 * jberger is rewriting the tests now
16:41 * sri would just get rid of epoch times
16:41 jberger I will document that duration is preferred, but having a drop-in replacement would be nice IMO
16:42 sri actually... supporting both is easy enough
16:42 jberger exactly
16:42 jberger its only about two extra lines
16:43 sri oh, we were thinking the same thing i guess
16:44 jberger convert the epoch time to a delta using the current time, then use that delta against steady_time as with duration
16:44 sri my $delta = $args->{expires_after} // ($args->{expires} - time);
16:45 jberger I picked 'duration' rather than expires_after, since the distinction seemed easy to miss, but yeah
16:45 * sri nods
16:45 jberger actually, that's a line shorter
16:45 * jberger steals
16:45 jberger but the logic was the same
16:49 good_news_everyone joined #mojo
16:49 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/FeS1UQ
16:49 good_news_everyone mojo/master ad76974 Sebastian Riedel: more consistent acceptor tests
16:49 good_news_everyone left #mojo
16:53 KindTwo joined #mojo
16:56 good_news_everyone joined #mojo
16:56 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/B5m8wg
16:56 good_news_everyone mojo/master 81c022b Sebastian Riedel: documentation tweaks
16:56 good_news_everyone left #mojo
16:56 denisboyun joined #mojo
17:01 KindOne joined #mojo
17:22 denisboyun joined #mojo
17:42 jberger sri: the short logic made the zero behavior difficult, so I'm using the longer logic
17:42 jberger but its ready now I think!
17:44 lammel2 joined #mojo
17:48 lammel2 left #mojo
17:49 jberger https://metacpan.org/release/JBERGE​R/Mojolicious-Plugin-Memorize-0.01
17:49 jberger batman ^^
17:52 jberger (http://irclog.perlgeek.de/​mojo/2013-08-14#i_7450610)
17:54 jberger \o/
17:54 lammel2 joined #mojo
18:04 sri jberger++
18:06 sri jberger: i would have allowed $self->memorize(foo => 'bar') too
18:07 sri $self->memorize(foo => sub {'bar'}) is a bit verbose
18:07 jberger I guess it seems nice, but it makes the argument handling inconsistent
18:08 sri imo not, last argument is required and can be a coderef or string
18:08 jberger perhaps ou are right
18:09 sri oh wait
18:09 sri no you are right
18:09 sri memorize begin ... end is valid too
18:10 sri memorize(sub {...}) vs memorize('foo')
18:10 sri or actually
18:10 sri i'm right!
18:10 sri ;p
18:11 jberger I think you are
18:11 sri somehow i thought memorize('foo') was a get operation
18:13 jberger is there some reason your old helper didn't allow this? remember I basically just kept your old logic, though now updated for steady_time and OO
18:13 sri it's old
18:13 sri some newer helpers do that kinda stuff
18:14 jberger then again, why memorize a string?
18:14 jberger I guess if it is interpolated
18:14 sri %= memorize some_other_helper('lalalala')
18:15 jberger that would still call some_other_helper, just ignore the return value
18:15 sri true
18:15 sri i've not thought this through
18:16 sri sri: USE CASES!!!1
18:16 jberger thats ok, spitballing is good
18:16 jberger if anyone has a good use case, I will add it, but for now, I think I will leave it as is
18:17 jberger for example if some_other_helper has side effects, I think that people might not expect that it gets called and be wrong
18:18 jberger in the meantime, time to port Galileo to this
18:18 jberger and other cleanups, like using some of the newer Mojo constructs
18:19 jberger that said, I think Galileo might be passe to some extent, I'm really liking NoSQL for CMS now
18:20 jberger sri: that video you shared a while back really sold me on that
18:20 jberger Mango + something like my MangoModel is more like what I want, plus, nonblocking is cool!
18:24 denisboyun joined #mojo
18:25 sri cms is pretty much the ideal use case for a document database imo
18:27 jberger agreed
18:28 jberger btw, preaction gave a good talk on Mojo+Angular for this month's Chicago.pm
18:29 * sri is still not sure if he likes angular or not
18:30 jberger he made a pretty compelling case
18:30 jberger for specific purposes, and when combined with an exposed REST api, it seems dead useful
18:32 jberger he made an admin page for permissioning site users, he binds the site + data + api using angular and its so easy
18:32 jberger preaction: have you uploaded it yet?
18:32 preaction one sec
18:32 jberger preaction: o/
18:35 preaction http://preaction.github.io/Perl/Mojo-Angular.html
18:36 jberger preaction++
18:38 jberger oh that reminds me
18:38 lammel2 looks nice
18:38 lammel2 (the mojo-angular thing)
18:38 sri the BAMM stack? :)
18:38 lammel2 buzzwords!
18:38 jberger sri: is there any way we could petition for upgrading _id to a bson_oid in cases where its the only reasonable usage?
18:38 preaction oh right
18:39 jberger sri: nice!
18:39 sri jberger: don't think that would be mongodb compliant
18:39 preaction it took me quite a bit to figure out that the reason why my Mango searches weren't working was because I sent ID as a string, not as an oid
18:39 sri anything can be a valid _id
18:39 preaction damn. that is also true.
18:40 sri sounds like something an ODM would manage, since it knows the schema
18:40 preaction yeah, that'd have to be the layer for that
18:40 jberger indeed
18:41 jberger actually, mine essentially does that anyway, but its an interesting idea
18:41 jberger I guess if you want something to be a unique key, you could essentially use that as the _id
18:42 jberger though just making that a unique index is probably the right way
18:42 sri haha, i was about to say slide 7 was risky... but remembered that mango now has fork support... so carry on :) http://preaction.github.io​/Perl/Mojo-Angular.html#7
18:42 preaction eh, i'm becoming more and more a fan of never having the ID of a document be useful information
18:43 preaction sri: yeah, jberger mentioned that too. i should change that to a helper.
18:43 sri not anymore ;p
18:43 jberger its still a good habit to get into, for two reasons
18:44 jberger 1) get used to it for non fork-safe handles
18:44 jberger 2) for use in ./myapp.pl eval '...;
18:44 jberger s/;/'/
18:45 jberger I use eval to check on the state of database entries all the time now
18:45 jberger so much easier
18:46 sri true
18:46 sri hmmmm
18:46 preaction i'm going to turn those app-##.pl footnotes into links so people can see the whole code
18:46 sri i wonder if the return value for eval -v should run through the dumper() helper
18:47 jberger sri: what if you explicitly print?
18:48 sri then you don't need -v
18:48 jberger oh
18:48 jberger I missed the -v
18:48 jberger oh that would be cool!
18:49 jberger or perhaps another switch? -d?
18:50 jberger_ joined #mojo
18:50 * jberger has to start getting ready for yet another event
18:50 jberger I will try to keep an eye on the channel from my phone
18:50 sri you need google glass :)
18:50 jberger hehe
18:51 sri -d doesn't look very nice next to -v
18:51 sri --verbose --dump
18:52 jberger_ Hahahaha
18:53 sri preaction: nice examples
18:53 jberger_ I had meant that -d wouldn't need -v
18:54 sri sure, i meant in the option overview
18:54 sri -d would be a variation of -v, which doesn't seem very logical
18:55 sri this is just a command though, we can change argument names if necessary
18:55 jberger_ Well however you would implement it, I know I would use it all the time
18:59 sri perhaps -V for very verbose :)
19:00 lammel2 joined #mojo
19:04 ruz hi, is there "what's new in mojo" slides or blog posts that cover past year?
19:04 sri http://blog.kraih.com/post/50517069291/mojolic​ious-4-0-released-perl-real-time-web-framework # for the highlights
19:04 sri and Changes for the details
19:05 sri everything after 4.0 is only covered by Changes so far
19:06 ruz changes don't describe reasons for deprecation and new additions
19:06 ruz thanks for the link
19:06 sri reasons are mostly just found in the irc logs
19:06 ruz heh
19:10 sri jberger_: http://paste.ubuntu.com/6026139/ # eval -V
19:12 jberger_ sri: Looks great
19:15 jberger__ joined #mojo
19:17 good_news_everyone joined #mojo
19:17 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/faiGhw
19:17 good_news_everyone mojo/master 3468c4b Sebastian Riedel: added -V option to eval command
19:17 good_news_everyone left #mojo
19:20 good_news_everyone joined #mojo
19:20 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/Ojjjng
19:20 good_news_everyone mojo/master 4a648b8 Sebastian Riedel: fixed typo in Changes
19:20 good_news_everyone left #mojo
19:23 good_news_everyone joined #mojo
19:23 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/eattOA
19:23 good_news_everyone mojo/master 4795bfc Sebastian Riedel: mention eval -V in cookbook
19:23 good_news_everyone left #mojo
19:30 lammel2 joined #mojo
19:37 lammel2 left #mojo
19:58 lammel2 joined #mojo
20:01 lammel2 joined #mojo
20:26 lammel2 joined #mojo
20:32 russum joined #mojo
20:33 russum left #mojo
20:59 lammel2 left #mojo
21:28 asarch joined #mojo
21:58 phillipadsmith Anyone using Mojolicious::Plugin::JSONP? Thoughts on this: http://irclog.perlgeek.de/​mojo/2013-08-25#i_7496407
22:02 jberger__ render_json was removed in mojolicious 4.0
22:04 dr joined #mojo
22:06 jberger__ File a bug, or better, submit a patch
22:31 phillipadsmith jberger__: I wondered about that; couldn't find reference to it...
22:34 jberger__ https://github.com/kraih/mo​jo/blob/master/Changes#L146
22:35 phillipadsmith I'll see if I can sort that out and send tempire a pull request...
22:37 phillipadsmith Seems odd that render_json is being used at all, on first glance: https://github.com/tempire/Mojolicious-Plugin-JSON​P/blob/master/lib/Mojolicious/Plugin/JSONP.pm#L23
22:39 sri the use of render(text => is also broken
22:40 sri can get you double encoding issues
22:40 * sri bonks tempire on the noggin
22:41 phillipadsmith sri: what's the preferred way of replacing render( text => … ) now?
22:41 sri always was render(data =>
22:42 * phillipadsmith makes that change too
22:46 phillipadsmith https://github.com/tempire/Mo​jolicious-Plugin-JSONP/pull/1
22:46 phillipadsmith tempire ^^^
22:47 phillipadsmith github++
22:58 russum joined #mojo
22:59 russum left #mojo
22:59 mire joined #mojo
23:09 human39 joined #mojo
23:17 quibbit joined #mojo
23:21 jberger_ joined #mojo
23:33 jberger__ joined #mojo
23:40 tott joined #mojo
23:57 denisboyun joined #mojo

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