Camelia, the Perl 6 bug

IRC log for #mojo, 2010-07-29

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

All times shown according to UTC.

Time Nick Message
00:02 GitHub111 joined #mojo
00:02 GitHub111 mojo: master Sebastian Riedel * bd61ed0 (3 files in 3 dirs): cleanup - http://bit.ly/d66lu4
00:02 GitHub111 left #mojo
00:03 xantus_ http://steve-yegge.blogspot.com/2010/07/wi​kileaks-to-leak-5000-open-source-java.html
00:03 garfield [ Stevey's Blog Rants: Wikileaks To Leak 5000 Open Source Java Projects With All That Private/Final Bullshit Removed ]
00:03 garfield http://xrl.us/bhunwr
00:05 sri heh
00:06 sri Zzz
00:10 Dennis okay, i got file upload working. thanks guys!
00:37 xantus_ \o/
01:21 MojoGuest248 joined #mojo
02:19 MojoGuest675 joined #mojo
02:20 MojoGuest488 joined #mojo
02:47 GitHub100 joined #mojo
02:47 GitHub100 mojo: master Sebastian Riedel * 3a12919 (3 files in 3 dirs): made inline base64 files look prettier - http://bit.ly/cq3FSz
02:47 GitHub100 left #mojo
02:48 sri couldn't sleep without making it prettier first :)
02:48 sri Zzz (rly)
02:56 MojoGuest680 joined #mojo
02:56 MojoGuest680 left #mojo
03:13 janus joined #mojo
03:27 GitHub80 joined #mojo
03:27 GitHub80 mojo: master Sebastian Riedel * 0d81f76 (3 files in 3 dirs): made inline base64 files a little prettier - http://bit.ly/aY0M1p
03:27 GitHub80 left #mojo
03:34 Alias_ joined #mojo
04:04 tl joined #mojo
04:35 MojoGuest773 joined #mojo
04:43 Foxcool joined #mojo
04:48 ashleydev joined #mojo
05:37 MojoGuest354 joined #mojo
06:09 alex joined #mojo
06:21 MojoGuest421 joined #mojo
06:22 MojoGuest421 left #mojo
06:54 yko joined #mojo
07:05 vti_mob joined #mojo
07:17 MojoGuest782 joined #mojo
07:24 MojoGuest90 joined #mojo
07:26 MojoGuest91 joined #mojo
07:39 MojoGuest757 joined #mojo
07:39 MojoGuest757 ccc
07:48 jest joined #mojo
07:54 yko MojoGuest757: it works.
08:28 marcus IT works
08:59 MojoGuest48 joined #mojo
08:59 MojoGuest48 hellow
09:01 yko MojoGuest48: hello. it works.
09:02 MojoGuest48 :)
09:02 MojoGuest48 cool
09:02 yko yep, rly
09:05 achromic joined #mojo
09:09 achromic left #mojo
09:10 marcus orly
09:20 MojoGuest765 joined #mojo
09:44 achromic joined #mojo
09:54 MojoGuest945 joined #mojo
10:00 MojoGuest945 test
10:00 MojoGuest945 /mp MojoGuest945
10:11 dotan joined #mojo
10:29 vti sri: i am trying to implement layouts for tt, but i when rendering i get 'header footer + header content footer', it's like the first time layout is rendering without content and is prepender to the final result
10:56 diegok joined #mojo
11:12 Akron joined #mojo
12:06 Schmidt Hmm, I am trying to make a Mojolicious::Lite application automaticly reload the server upon saving changes to the file. I've tried perl $filename daemon {reload,-reload,--reload} and none of them seemed to have worked. What am I doing wrong ?
12:30 MojoGuest857 joined #mojo
12:30 MojoGuest611 joined #mojo
12:38 y0j joined #mojo
12:41 sri Schmidt: --reload
12:47 sharifulin joined #mojo
12:48 Schmidt sri: alrighty, thanks :)
12:48 sharifulin sri: hi! why do you replace name of dispatching controller to 'Dispatching controller.' http://github.com/kraih/mojo/blob/mast​er/lib/MojoX/Dispatcher/Routes.pm#L86
12:48 garfield [ lib/MojoX/Dispatcher/Routes.pm at master from kraih's mojo - GitHub ]
12:48 garfield http://xrl.us/bhuo79
12:49 sharifulin sri: I can't understant what kind of controller works
12:49 sri sharifulin: dunno, must have been a side effect of the new app embedding code
12:50 sharifulin sri: please, fix it. thanks :)
12:50 sri patches welcome
13:21 MojoGuest910 joined #mojo
13:23 MojoGuest910 test
13:23 sri failed
13:25 MojoGuest910 =)
13:27 MojoGuest910 left #mojo
13:36 marcus I see rakudo star is out
13:36 marcus sockets or no sockets?
13:38 vti it has lwp::simple
13:38 marcus and no mojo? :(
13:39 sri no non blocking sockets
13:39 sri i won't bother without
13:41 sri rakudo is still years away from production quality
13:43 sri i've been asking many times about non blocking sockets, but there is no interest
13:43 marcus sri: write a patch? :)
13:43 Akron Hi!
13:43 Akron I'm wondering if there is any step-rendering function in Mojolicious' rendering engine. And - if this would be useful. ;)
13:43 Akron (Sorry - not Rakudo* related.)
13:43 marcus Akron: what do you mean by step-rendering?
13:43 sri marcus: do you really think i have that much spare time?
13:44 marcus sri: less tv!
13:44 Akron Can I post example code for that idea?
13:44 MojoGuest725 joined #mojo
13:44 marcus Akron: nopaste it?
13:45 Akron marcus: I mean - it's 25 lines Mojolicious::Lite code.
13:45 sri marcus: and burn out in a month?
13:47 marcus Akron: stuff it in http://gist.github.com/ ?
13:47 garfield [ Gist - GitHub ]
13:48 Akron http://gist.github.com/498162
13:48 garfield [ gist: 498162 - GitHub ]
13:48 marcus sri: is it really that much work?
13:49 marcus (I have no idea)
13:49 sri marcus: there is no support in parrot
13:49 sri Akron: that can't be done portably
13:50 sri and even if it was done non portably it couldn't look that simple
13:51 sri Akron: have you thought about the side effects? content length calculation, cookies...all the header stuff that needs to be finished in advance?
13:51 Akron sri: Ah ... okay. I thought of just rendering it partally, setting content-length to 0 and returning an iterator to the positions.
13:51 sri invalid http
13:51 marcus ield> [ gist: 498162 - GitHub ]
13:51 marcus chunked mode
13:51 sri the only way to do it is chunked
13:51 sri returning an iterator
13:52 Akron But that's not really portable, isn't it?
13:52 sri isn't
13:52 sri nothing is
13:52 Akron :)
13:52 sri fastcgi for example will often just buffer the whole response until it send something
13:52 sri same for cgi
13:52 sri depends entirely on the server
13:52 marcus asp and php have directives to control buffering :)
13:53 Akron Ah - Okay.
13:53 sri marcus: which rarely works
13:53 sri marcus: RTFM!
13:53 marcus sri: it works well with asp and IIS
13:53 marcus I've used it
13:53 sri ...
13:53 sri because they control the whole stack
13:53 marcus and probably with mod_php as well
13:53 marcus sure
13:53 sri we do not control the whole stack
13:53 vti http://use.perl.org/~Alias/journal/40470
13:53 marcus time to write mod_mojo :p
13:53 garfield [ Journal of Alias (5735) ]
13:53 sri mod_php can do it because it controls the whole apache process
13:54 sri marcus: feel free
13:54 sri marcus: why are you flooding me with impossible features today?
13:55 marcus sri: bored?
13:55 sri -.-
13:55 marcus anyways, I've been nagging about streaming support for ages.
13:55 marcus it's one of my pet peevees.
13:56 sri and we do support it, don't think it can be made any less messy
13:56 marcus vti: I totally agree, moose stack traces are horrible to read.
13:56 marcus I also think that moose in general has grown out of proportion. In particular, the accessors support way too many options.
13:57 sri the solution is what Akron suspected, returning an iterator to generate a chunked response, rendering partials
13:58 marcus mm
13:58 sri if you think it can be done more elegant on a higher level be my guest :)
13:59 sri i've thought this through and actually checked my facts
14:01 sri adding some features to the renderer to allow this to become a somewhat cleaner plugin (like the rails one) is possible
14:01 marcus neat
14:01 sri but making it a core feature is out of the question, due to portability issues
14:01 marcus I guess the json generator could be made to output json in chunks as well.
14:02 sri why?
14:02 marcus stream parsing the output?
14:02 marcus http://stackoverflow.com/questions/444​380/is-there-a-streaming-api-for-json
14:02 sri ok, but why?
14:02 garfield [ Is there a streaming API for JSON? - Stack Overflow ]
14:02 garfield http://xrl.us/bhupcf
14:03 sri json is not a good streaming format
14:03 marcus sri: I agree, xml is better
14:03 sri i disagree ;p
14:03 sri websocket streams containing json/xml messages
14:03 sri thats a real streaming protocol
14:03 marcus sri: one use case for me recently was to start displaying data in my app before the 3 mb xml stream was done downloading
14:04 sri websockets
14:04 Akron Plugin possibility would be great.
14:04 marcus fetching news articles, I could present the user with the first article as soon as it was done downloading
14:04 sri Akron: i'm willing to apply all patches neccessary
14:04 marcus sri: this was for an obj-c app talking to a http rss feed
14:04 marcus sri: just saying, that's an example of a use case.
14:05 sri marcus: not a very good one though
14:05 Akron sri: Great! I will think about it.
14:07 achromic joined #mojo
14:07 MojoGuest166 joined #mojo
14:07 sri marcus: the context switch is still damn expensive, you'll buffer half an xml document, wait until it is written, and after that keep on rendering
14:08 sri double latency overhead
14:09 marcus I thought computers were very good at context switching
14:09 marcus unlike programmers
14:10 marcus also, you can have a reduced memory footprint given that you don't need to have the whole document in memory.
14:11 sri you don't understand, the data has to travel to the user first, including all latency related delays, before you even get a chance to continue
14:12 sri this is only really worth it if you are waiting for an async backend service anyway
14:12 sri to send the html head in advance so images/js/css can be preloaded
14:13 sri in pretty much all other situations you are off worse
14:13 marcus are you quite sure about that? I don't understand why it'll have to travel to the user?
14:14 sri or a varnish has to cache it
14:14 marcus before you can write the next chunk
14:14 sri the handler is blocking until you context switch to the server again
14:14 sri we have no threads
14:15 sri you either write data or do other stuff, never both
14:16 sri there is no magic buffer you can flush and everything gets automatically written to the user without delay ;p
14:16 marcus and because of tcp, you have to wait for the response before it context switches?
14:16 sri ye
14:16 marcus http over udp? ;)
14:16 sri ;p
14:17 sri http://php.net/manual/en/function.flush.php # worth reading
14:17 garfield [ PHP: flush - Manual ]
14:18 marcus sri: if you use the standalone http daemon and put a ngnix in front, you wouldn't have this problem, would you?
14:18 sri marcus: depends on caching behavior
14:18 marcus proxying behavior even?
14:18 sri ye
14:18 sri there is so much to go wrong
14:19 sri you can implement it fine for one specific setup, but never protably
14:19 sri *portably
14:19 sri i'm all for making it a plugin, but it will be a pita to support
14:20 sri php has full control over the apache process, and things still go wrong as soon as you add say mod_gzip
14:21 marcus mmm
14:25 sri you of course also lose all convenience features such as exception tempates, since you've already written the head section
14:26 marcus yupp
14:26 yko and you can not change layout while rendering templates
14:26 marcus I am not a big fan of layouts
14:26 marcus I prefer explicit header / footers.
14:27 sri http://www.w3.org/TR/html5/offline.html # hidden here might be a better solution for the future
14:27 garfield [ 5.6 Offline Web applications — HTML5 ]
14:28 sri like, what if browsers checked a manifest file before loading the page
14:28 sri (just loud thinking)
14:29 marcus yeah, it solves the parallel downloading of assets
14:29 sri the streaming problem is solved by websockets
14:29 sri (already)
14:30 sri and it solves it so well, i really don't care about other solutions :)
14:33 sri in the end, as long as i can serve 200+ req/s per server i don't think i even need sending the html head in advance
14:34 sri this is really micro optimization unless you have blocking DBI calls bringing your req/s down to 0. numbers
14:37 marcus which I always do
14:37 sri bad marcus
14:39 marcus so, which hyper fast 200 req/s backends are you coding against?
14:39 sri couchdb can provide that
14:40 sri (if you keep the views fresh)
14:41 sri just throw another server at it if it gets slower :)
14:42 marcus I really should code a couchdb app sometime soon
14:43 marcus vti: is your couchdb mojo glue done?
14:43 sri it still has some quirks but is bound to become the one true databse :)
14:43 sri *+a
14:43 marcus ye
14:44 sri http://www.paperplanes.de/2010/7/26/1​0_annoying_things_about_couchdb.html # spot on
14:44 garfield [ paperplanes. 10 Annoying Things About CouchDB ]
14:44 garfield http://xrl.us/bhuii2
14:44 marcus I've started an app, actually
14:44 sri and they've made this list their new roadmap :)
14:48 vti marcus: ye, it has even attachments, but it is async
14:50 marcus vti: is it on cpan?
14:50 ysyrota joined #mojo
14:52 vti marcus: of course not
14:52 sri cpan is soooo last year
14:53 sri people can't follow me on cpan, why would i upload something there?
14:54 MojoGuest897 joined #mojo
14:54 vti true
14:55 marcus guess it's time for cpan.iusethis.com :p
14:56 sri as long as i can login with my twitter id!
14:56 marcus twat id
14:56 MojoGuest786 joined #mojo
14:57 marcus I guess I *should* add that.
14:57 marcus and facebook connect
14:57 vti oauth!!!!!!!
14:57 marcus vti: is it done yet?
14:57 sri vti did it
14:58 sri it even looks quite sleek now i think
14:58 marcus vti is a doer
14:59 vti nice plugin with a simple store interface and yes
14:59 vti just two classes :D
14:59 marcus mm, looking at your example now
14:59 vti it is really low level
15:00 sri he'll make it a plugin he promised
15:00 * sri cracks the whip
15:00 stephan48 marcus: please checkout janrain for such things like login with other services
15:01 vti i think i can add "providers" too, so you don't have to specify urls, and just $oauth->provider('twitter')->request_token
15:03 MojoGuest786 hey guys
15:05 MojoGuest786 you see me?
15:05 stephan48 no.
15:08 sri nope
15:11 MojoGuest786 :-)
15:14 MojoGuest171 joined #mojo
15:17 Akron Oh - I've got another question: Is there currently work going on, adapting an authorization model like RBAC or something for Mojolicious?
15:18 sri don't think so
15:20 Akron Ah. I think a "standard" authorization plugin would be great for further app development (wikis, blogs, etc.).
15:20 sri all kinds of reusable components should be useful :)
15:21 Akron Sure. :)
15:21 sri the main focus at the moment is more on individual auth methods like oauth i think
15:23 Akron That's - of course - closer to the Mojo idea. An RBAC model would need a complete admin app.
15:23 sri and a persistence layer
15:23 Akron Yes.
15:24 sri which is something we don't have baked in by design
15:25 Akron Which is a great thing - although their may be a conventional layer as soon as more Mojolicious apps arise.
15:25 marcus sri: don't care if it's a plugin, I need to use it in catalyst ;)
15:26 sri Akron: likely, my bets are on on one of the couchdb modules
15:26 sri marcus: ;p
15:28 Akron sri: Please have mercy with all the shared-webhost users! ;)
15:29 sri but there is already couchdb hosting ;)
15:30 sri actually, couchdb hosting is free atm
15:30 sri http://www.couch.io/get
15:30 garfield [ Get — CouchDB: The NoSQL Document Database ]
15:33 Akron sri: Hm. I will give it a try. But there's currently no bundled webhosting package in a garage near me. ;)
15:41 sri i bet they will start offering nosql packages soon too
15:41 sri :)
15:42 GitHub82 joined #mojo
15:42 GitHub82 mojo: master Sebastian Riedel * e0414d1 (1 files in 1 dirs): fixed typo - http://bit.ly/dzMGFk
15:42 GitHub82 left #mojo
15:42 sri awesome...i quoted a typo on the list -.-
15:46 Grauwolf left #mojo
15:48 MojoGuest540 joined #mojo
15:48 sri rails 3 is a really good example for why you should have good unit tests
15:50 sri they can't keep their release schedule because new problems keep popping up
15:52 marcus nice
15:52 marcus guess merging two frameworks is kind of hard
15:52 sri turned out to be less of a merge
15:53 sri they just rewrote core components to be more modular
15:53 sri actually, i've not seen any actual merb code in rails yet
15:53 sri was more of a team merge i think :)
15:54 sri http://www.reddit.com/r/programming/comments/cv​33g/more_drama_from_the_rails_teamthis_is_like/ # lol
15:54 garfield [ More drama from the Rails team...this is like reality TV for me : programming ]
15:54 garfield http://xrl.us/bhupoa
16:09 MojoGuest759 joined #mojo
16:13 MojoGuest640 joined #mojo
16:13 MojoGuest640 pretty cool
16:24 sri i love it when a plan comes together
16:27 Akron left #mojo
16:29 sri wonder if they'll make more a-team movies
16:30 marcus I hated the a-team movie
16:30 MojoGuest291 joined #mojo
16:30 MojoGuest291 fasdfsdfsadf
16:30 sri it was ok
16:46 ysyrota left #mojo
16:51 MojoGuest994 joined #mojo
17:01 ashleydev joined #mojo
17:24 dotan left #mojo
17:47 MojoGuest369 joined #mojo
17:48 sri wow, safari 5.0.1 has actually been updated to draft 76 :O
18:01 GitHub22 joined #mojo
18:01 GitHub22 mojo: master Sebastian Riedel * abf4990 (4 files in 4 dirs): removed websocket draft 75 bandaid fix because Safari 5.0.1 finally supports draft 76 - http://bit.ly/bE21W6
18:01 GitHub22 left #mojo
18:03 GitHub117 joined #mojo
18:03 GitHub117 mojo: master Sebastian Riedel * b99bb3d (2 files in 2 dirs): removed another draft 75 bandaid fix - http://bit.ly/9hOzzV
18:03 GitHub117 left #mojo
18:10 vel joined #mojo
18:11 marcus sri: guess that's why the websocket demo broke for people earlier today?
18:11 sri marcus: nope
18:11 sri didn't work with both
18:12 sri not sure why though
18:12 marcus http://10k.aneventapart.com/
18:12 garfield [ 10K Apart | Inspire the web with just 10K. ]
18:13 sri hehe
18:14 MojoGuest648 joined #mojo
18:22 Htbaa joined #mojo
18:26 sri no idea what breaks the irc client in safari
18:27 MojoGuest965 joined #mojo
18:27 sri hmm
18:27 sri weird thing is you can watch, just not talk
18:29 sri might not even be websocket related
18:30 sri could be encoding or so
18:30 sri "Not connected, use the command /connect to connect first"
18:30 sri thats the error message
18:31 sri my bet is on message encoding
18:39 MojoGuest812 joined #mojo
18:40 MojoGuest812 test
18:40 MojoGuest812 ok, safari bug
18:40 MojoGuest812 webkit nightly just works
18:41 MojoGuest812 no idea what apple is doing there
18:44 MojoGuest274 joined #mojo
18:44 MojoGuest274 hi
18:45 MojoGuest329 joined #mojo
18:45 MojoGuest648 joined #mojo
18:46 MojoGuest329 yo alex
18:46 MojoGuest648 Hey
18:46 MojoGuest648 oh it's that smart friend of mine
18:46 MojoGuest648 :D
18:46 drew :D
18:46 drew cool this
18:46 drew i want
18:46 MojoGuest648 nice, didn't realise you could do this with HTML5
18:46 sri you can do much more with it :)
18:46 drew neither did I, was just browsing for an alternative to java based irc clients
18:47 MojoGuest648 h
18:47 drew any idea where to get the source from?
18:47 MojoGuest648 Supermarkets
18:47 drew :|
18:47 MojoGuest648 :d
18:47 MojoGuest648 :d
18:47 MojoGuest648 :D
18:47 sri http://blog.kraih.com/perl-is-ready-for-html5 # there is a link
18:47 garfield [ Perl is ready for HTML5 - Sebastian Riedel about Perl and the Web ]
18:47 drew yay ty
18:48 MojoGuest648 PHP ftw!
18:48 drew hrm, i have to say this is amazing
18:49 drew should I just rip the code from my browser or will there be a proper distribution some place?
18:49 drew i read somewhere that the project was dropped?
18:49 MojoGuest648 so is this basically an iframe?
18:49 sri it depends on a websocket server
18:49 drew ah
18:49 sri which proxies to irc
18:49 drew :(
18:50 drew not standard stuff then
18:50 sri very much standard stuff
18:50 drew my host wouldnt have that kinda magically stuff
18:50 MojoGuest648 would've thought you'd know it, Drew lol
18:50 sri browsers will never be allowed to open sockets directly
18:50 sri websockets are a html5 standard
18:51 drew so, this is basically an ajax chat script but using a proxy to interface with irc?
18:51 sri similar to ajax, but much more powerful
18:51 MojoGuest648 more powerful you say?
18:52 sri bidirectional streaming
18:52 sri long poll in both directions
18:52 MojoGuest648 intriguing
18:52 MojoGuest648 I'm going to read more about these HTML5 sockets, I did read about HTML5 but I never came across sockets :/
18:53 drew long poll = no chance for php/apache standard setup
18:53 drew ?
18:53 sri correct
18:53 drew i was getting all excited then too :(
18:53 sri php/apache is toast
18:54 sri not future proof
18:54 drew :|
18:54 drew how so
18:54 MojoGuest648 because he stuck it in his toaster
18:55 sri because you need a dedicated process per connection
18:55 drew true, i understand at the moment, most configurations are crap when it comes to long polled ajax queries like what facebook uses. but there's no reason why this won't change in the future
18:56 drew but going off the uptake rate of php 5.3 it'll be a while!
18:56 drew yahoo, digg, facebook, photobucket, even wikipedia uses php... surely they won't just rewrite there entire code base... the language will evolve with them
18:57 sri they are using mostly python servers for their long poll needs
18:58 sri facebook is also using erlang
18:58 sri twitter is scala
18:58 sri they have all moved on
18:59 drew *click* wiki erlang
19:00 drew ewww
19:00 drew i still think php is nice to look at :D
19:00 MojoGuest648 lol
19:00 sri oh my
19:02 MojoGuest648 I'll stick with PHP
19:02 drew i've been visted by pizza claus
19:03 MojoGuest648 mmmm yum
19:03 drew python i definitely agree is a dam cool language
19:04 sri this is a perl channel btw :)
19:05 drew oh, sorry to offend :D
19:05 drew j/k
19:05 drew perl = good web development language?
19:05 drew going to do some reading...
19:05 sri yes
19:07 MojoGuest648 so if it turns you on you'll move to perl for your CMS then?
19:08 sri memowe/marcus: any updates?
19:08 drew alex, lol no
19:09 drew but if i wanted to, most languages are fairly interchangeable
19:09 drew wouldn't be too hard to pick up
19:09 drew same goes for you if you want to swap
19:09 drew but anyway, pizza calls
19:09 MojoGuest648 with no mouth
19:09 MojoGuest648 I see where you're coming from
19:09 drew i wish i could use this irc client :(
19:10 drew how is the proxy run?
19:10 MojoGuest648 Whyyyyyyyyyyyyyyyyyyy
19:10 drew just a web script?
19:10 spleenjack joined #mojo
19:11 sri perl script.pl daemon
19:12 sri built in web server
19:12 drew ah ok
19:13 MojoGuest648 Why's there so many users yet no one is chatting?
19:13 drew idling it's what the natives usually do on irc
19:13 sri everybody is playing starcraft 2 of course
19:13 drew lol
19:13 drew or that
19:13 MojoGuest648 mass kick
19:14 MojoGuest648 /me a
19:14 drew ?
19:14 MojoGuest648 ?
19:14 MojoGuest648 Only just realised you can move the window lol
19:15 drew sri = sebastian?
19:15 * sri nods
19:15 drew nice work! didn't realise you actually created it :)
19:16 MojoGuest648 I wouldn't have got that lol
19:18 yko wow. Look, mum, i'm op on #mojo!
19:20 ask joined #mojo
19:26 MojoGuest676 joined #mojo
19:35 xantus_ yoink!
19:35 xantus_ 'morning
19:35 MojoGuest950 joined #mojo
19:35 MojoGuest950 works
19:53 xantus_ joined #mojo
19:58 yko haha morning :)
19:59 * yko went to urbandictionary learn what is that 'yoink'
20:00 yko nice :)
20:12 y0j joined #mojo
20:26 xantus_ yko: yeah, usually said while taking something
20:28 yko aha, got it already, thanks
20:36 xantus_ darn, lost mine when I reconnected
20:37 sri xantus: the safari problems seem to b a bug in safari
20:37 sri you can even watch irc, just not send
20:37 sri same in 5.0 and 5.0.1
20:38 sri webkit nightly works flawless
20:38 ask joined #mojo
20:40 sri wonder how long it will take until we see websocket proxies
20:41 sri i can totally see nginx becoming some kind of message bus for websockets
20:44 xantus_ Use of uninitialized value in concatenation (.) or string at /home/zantus/mojo-websocket-example​s/script/../mojo/lib/Mojo/IOLoop.pm line 809
20:46 xantus_ an undef value is getting into @error from within one_tick
20:46 sri strange
20:46 xantus_ probably here push @error, $_ for $loop->handles(EPOLL_POLLERR); or
20:47 xantus_ the Poll one
20:47 xantus_ usually from a disconnect
20:47 sri looks harmless
20:48 xantus_ yeah
20:49 xantus_ setting $error to something in _error if not defined may be in order.  $error ||= 'unknown';
20:54 GitHub64 joined #mojo
20:54 GitHub64 mojo: master Sebastian Riedel * ef49686 (3 files in 2 dirs): cleanup - http://bit.ly/aosKVD
20:54 GitHub64 left #mojo
20:54 sri done
20:55 xantus_ :)
20:56 xantus_ propably harmless
20:56 xantus_ ;) I'll shut up now
20:56 sri ;p
20:57 sri it needs to be said, or people will get nervous
20:58 xantus_ its misspelled
20:58 sri :o
20:58 xantus_ was there a new futurama last night
20:59 GitHub178 joined #mojo
20:59 GitHub178 mojo: master Sebastian Riedel * 1df9166 (1 files in 1 dirs): fixed typo - http://bit.ly/bXwauo
20:59 GitHub178 left #mojo
20:59 sri don't think so
21:00 xantus_ tonight
21:08 y0j joined #mojo
21:16 xantus_ http://www.brookstone.com/​ar-drone-quadricopter.html
21:16 garfield [ AR Drone by Parrot Now Available at Brookstone! ]
21:16 garfield http://xrl.us/bhuqqd
21:39 ask_ joined #mojo
21:48 marcus ooog
21:48 marcus must have AR drone
21:48 marcus xantus: 404 :X
21:51 xantus_ joined #mojo
22:01 * sri wants a mojolicious textmate bundle
22:03 sri have working syntax highlighting for ep templates, need data section support now
22:11 ltriant joined #mojo
22:15 dotan joined #mojo
22:15 dotan has zzz been called yet?
22:16 xantus_ cool
22:17 marcus sri: nice
22:17 sri yay, data section works :)
22:17 marcus @w00p@w00p@
22:29 dotan left #mojo
22:34 sri http://github.com/kraih/mojo-tmbundle
22:34 garfield [ kraih's mojo-tmbundle at master - GitHub ]
22:34 sri just syntax for now
22:35 sri note that i have no clue what i am doing, patches welcome! ;p
22:41 sri oh, another positive side effect, you get free html highlighting in the DATA section :)
22:50 sri http://twitpic.com/29u217
22:50 garfield [ #Mojolicious syntax highlighting in TextMate. on Twitpic ]
23:15 MojoGuest48 joined #mojo
23:16 MojoGuest48 left #mojo
23:31 xantus_ http://www.27bslash6.com/i​mages/Chatroulette_03.jpg
23:31 garfield http://xrl.us/bhpqv6
23:31 sri oooooold ;p
23:31 xantus_ yeah
23:32 xantus_ but funny
23:33 Alias_ joined #mojo

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