Camelia, the Perl 6 bug

IRC log for #mojo, 2010-11-06

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

All times shown according to UTC.

Time Nick Message
00:29 und3f is now known as und3f[A]
00:33 spleenjack left #mojo
00:50 daviddelikat joined #mojo
02:52 hatman joined #mojo
02:52 hatman can you help  me, how to change session domain for mojo
02:53 hatman ?
02:57 hatman i cant get this to work, i need session to work with cookie in www.example.net and with example.net
03:16 daviddelikat left #mojo
03:17 hatman i get it mojo sucks
03:17 hatman left #mojo
03:50 alnewkirk lol
04:03 tl joined #mojo
04:05 tholen42 left #mojo
04:21 kimoto left #mojo
04:21 garu too bad hatman is gone, I believe Mojo::Cookie::Response's domain() would solve his problem
04:39 achromic left #mojo
04:40 achromic joined #mojo
05:17 GabrielVieira2 joined #mojo
05:23 GabrielVieira left #mojo
06:40 perlrocks Twitter: "@shadowcat_mst RE:Mojolicious Perl Sane Versioning ++ ... maybe Semantic would be better?" --xenoterracide http://twitter.com/xenoterra​cide/status/799607437135872
07:49 und3f[A] is now known as und3f
08:02 * sri yawns
08:02 sri lol, i don't even understand hatmans question
08:03 sri but garu is prolly right
08:09 achromic left #mojo
08:09 achromic joined #mojo
08:12 perlrocks Twitter: "#mojolicious 1.0 is getting closer and closer. :) #perl" --kraih http://twitter.com/kraih/status/822818363146240
08:45 und3f Is it possible to get config of Mojolicious in tests?
09:02 sri what?
09:03 sri und3f: be more specific with your questions
09:04 sri you're always very hard to understand
09:05 und3f sri, i need to get result of Plugin::JsonConfig in tests. Is it possible?
09:05 sri everything is possible
09:06 und3f is it possible in easy way?
09:06 sri all you have to do is follow the tutorial
09:06 sri i bet you just ignored MOJO_HOME
09:07 und3f sri, i need to get json_config result right in test code
09:07 und3f sri, application uses json_config without any problems
09:07 sri read the tutorial
09:08 sri there's a perfectly fine example for how to correct MOJO_HOME
09:09 und3f i don't have problems with MOJO_HOME. I just asking how to get json_config result in tests
09:09 perlrocks Twitter: "[Perl] / Mojolicious:: Lite you can create a large-scale Web apps - Getting Started with Perl code samples http://htn.to/v1kNxi" (ja) --yasutaka http://twitter.com/yasutaka/status/837163595862016
09:09 sri und3f: then i don't understand the question :(
09:10 perlrocks Twitter: "RT! Yasutaka: [perl] / Mojolicious:: Lite you can create a large-scale Web apps - Getting Started with Perl code samples http://htn.to/v1kNxi" (ja) --perlism http://twitter.com/perlism/status/837304595783680
09:33 marcus_ joined #mojo
09:38 marcus_ left #mojo
09:40 marcus_ joined #mojo
09:42 sri english is also not my native language, so always make your questions as specific as possible
09:55 marcus_ Back on znc
09:56 perlrocks Twitter: "RT @ kraih: # mojolicious 1.0 is getting closer and closer. :) # Perl" (de) --SaschaDroste http://twitter.com/SaschaDr​oste/status/848997069627392
10:05 baton8 left #mojo
10:10 perlrocks Twitter: "RT @kraih #mojolicious 1.0 is getting closer and closer. :) #perl" --xomaa http://twitter.com/xomaa/status/852436797366272
10:11 baton8 joined #mojo
10:11 dotan_ joined #mojo
10:15 perlrocks Twitter: "Thinking about switching to a #rakudo like $year$month$day versioning scheme for #mojolicious with cutesy names for major releases. :) #perl" --kraih http://twitter.com/kraih/status/853655985070080
10:18 perlrocks Twitter: "Traditional $major.$minor.$patch versioning is broken in #perl, so we might just as well take two steps forward and innovate. #mojolicious" --kraih http://twitter.com/kraih/status/854624865099776
10:33 perlrocks Twitter: "Just look at those horrible #catalyst version numbers during the last few years. http://bit.ly/cgXFsx #perl #mojolicious" --kraih http://twitter.com/kraih/status/858296097443840
10:37 perlrocks Twitter: ""use Mojolicious 20101104;" with a cutesy name for announcements would surely be an improvement. #perl #mojolicious" --kraih http://twitter.com/kraih/status/859345730404352
10:39 sri thoughts? :)
10:43 perlrocks Twitter: "@und3f Since the date is the major version you can just append a minor one for multiple releases per day. ;) "use Mojolicious 20101104.1;"" --kraih http://twitter.com/kraih/status/860767867895808
10:47 perlrocks Twitter: "@und3f Or you take it one step further and make a new release every second. :D "use Mojolicious 20101104.221013"" --kraih http://twitter.com/kraih/status/861854683365376
10:56 perlrocks Twitter: "RT @kraih #mojolicious 1.0 is getting closer and closer. :) #perl" --sharifulin http://twitter.com/sharifu​lin/status/864173164601344
11:05 perlrocks Twitter: "@kraih do you plan to stabilize all experimental features and modules in #mojolicious 1.0?" --und3f http://twitter.com/und3f/status/866244475494400
11:05 und3f twitter is nice
11:05 sri und3f: about experimental features, i won't stabilize all for 1.0 i think
11:06 sri but most
11:06 und3f sri, i've used first version of taghelpers in 2 or 3 projects
11:07 sri i don't get much feedback for taghelpers, those are unlikely to go non experimental so soon :/
11:07 und3f sri, and rewriting that projects  with new api wasn't the best work i did in day of release :)
11:07 sri DOM on the other hand will
11:08 und3f sri, what about resolver?
11:08 sri i like the api, but it is way too early to say
11:08 sri could still have huge problems like the unpack bug
11:09 und3f sri, before unpack bug were found some of server on check-host didn't returned result for some hosts
11:09 und3f sri, but i had no time for searching that bug
11:10 und3f sri, now everything seems to be fine
11:10 sri but we won't know for sure in time for 1.0
11:10 und3f right..
11:11 sri only good feedback can decide about an experimental feature going stable
11:14 und3f isn't check-host a good feedback? :)
11:14 sri sure, one :)
11:14 und3f :)
11:14 sri in this case i would also be happy with no feedback for a few months
11:15 sri if people don't notice the resolver it does its job :)
11:15 und3f right
11:25 dotan_ left #mojo
11:25 marcus left #mojo
11:25 marcus_ is now known as marcus
11:26 und3f sri, btw, is it right that socket write operation performed before on_connect event will not be sent (it will be sent just after another write perform)
11:26 perlrocks Twitter: "RT @ kraih: # mojolicious 1.0 is getting closer and closer. :) # Perl" (de) --PerlHowto http://twitter.com/PerlHow​to/status/871675067371520
11:39 MojoGuest366 joined #mojo
11:39 MojoGuest366 From: http://www.google.ru/search?hl=en&clien​t=firefox-a&hs=y9u&rls=org.mozilla%​3Aen-US%3Aofficial&q=html5+irc&aq=f​&aqi=g-c3&aql=&oq=&gs_rfai= (1 hits)
11:40 MojoGuest366 left #mojo
11:46 sri und3f: that would be wrong
11:47 sri but i think you might just be using ioloop wrong
11:48 sri like my $id = $loop->connect(...); $loop->write($id => ...);
11:48 sri that would be totally wrong
11:50 und3f sri, can you take a look? https://github.com/und3f/mojox-redis​/blob/master/lib/MojoX/Redis.pm#L79
11:51 sri too much code
11:52 sri you are simply not supposed to start writing before the on_connect event fired
11:52 sri well, actually it doesn't matter much, but the on_connect event might happen later
11:53 sri i'm not sure but i suspect the socket is writable even before it is connected
11:54 sri you just fill the write buffer
11:54 sri thats operating system specific though
11:55 daviddelikat joined #mojo
11:55 polvo sri, i want to load a 3rd party cookie. which module should i use?
11:56 sri polvo: more context
11:56 sri und3f: don't write before you have a connection
11:56 sri und3f: looks like you start writing right after connect returns
11:56 polvo sri, i'm setting a cookie with jquery and i want to load it into my controller
11:56 sri thats wrong
11:56 und3f sri, before i got on_connection?
11:57 sri und3f: yes
11:57 sri thats the whole point of on_connect
11:57 und3f sri, that what the code does now
11:57 und3f sri, but before code with writing before on_connect worked :|
11:57 und3f dunno why
11:57 und3f sri, thank
11:57 sri und3f: ok, well, too much code for me to understand now
11:58 sri polvo: $self->cookie('name')?
12:00 polvo sri, thanks
12:00 sri polvo: if in doubt, there is always $self->req thats gives you access to absolutely everything
12:04 sri und3f: actually i don't understand the question again :(
12:04 sri "is it right that socket write operation performed before on_connect event will not be sent (it will be sent just after another write perform)"
12:04 sri what do you mean with write operation performed?
12:04 und3f sri, $loop->write( $id ... )
12:04 sri before on_connect it should not have anything in its buffer to write
12:05 und3f ok
12:05 sri but *you* perform that operation!
12:05 sri not the loop
12:05 sri *you* trigger ->write
12:06 sri und3f: if you want a blocking connect you need to make it blocking yourself
12:07 sri your api is wrong atm
12:07 sri $r->execute(ping => sub {...
12:07 sri if that one operation does connect *and* write you need to block yourself
12:08 und3f sri, why?
12:08 und3f sri, on connection error i return error to operation callback
12:08 sri ok, wrong words
12:09 sri you have to chain callbacks internally
12:09 sri your first write needs to be triggered by on_connect for a new connection
12:10 sri the api actually works
12:11 Htbaa joined #mojo
12:11 und3f sri, understand
12:11 und3f sri, actually it works in that way now. checkout $self->{_connection}
12:12 und3f sorry. $self->{_connecting}
12:12 und3f oh, right
12:12 und3f can make it easy
12:16 perlrocks Twitter: "@rainboxx Thats why i want cutesy names to go with them, Mojolicious Snowflake (20101224). ;)" --kraih http://twitter.com/kraih/status/884248651239424
12:28 marcus left #mojo
12:29 marcus joined #mojo
12:48 spleenjack joined #mojo
13:01 sri i guess in the end i will just stick with 1.0 and 2.01... version numbers :)
13:01 sri 99 minor releases for every major should be fine
13:04 perlrocks Twitter: "In the end i think we'll just go with minimalism for #mojolicious version numbers again after 1.0 and use 1.01. #perl" --kraih http://twitter.com/kraih/status/896292381270016
13:15 y0j left #mojo
13:29 y0j joined #mojo
13:44 MojoGuest354 joined #mojo
13:44 MojoGuest354 From: http://www.google.com.tw/search?hl=zh-T​W&q=websocket+Demo&aq=f&aqi​=g1g-m1&aql=&oq=&gs_rfai= (1 hits)
13:44 MojoGuest354 lkjlkj
13:45 MojoGuest354 What is it ?
13:45 MojoGuest354 sdfsdf
13:45 MojoGuest354 left #mojo
13:47 marcus woot
13:47 sri moo
13:48 marcus sri: do you irc with znc?
13:48 sri for almost a year now
13:48 marcus I just set it up on my new mini
13:49 sri i have it on the mojolicio.us box
13:49 marcus but I'm getting 'Cannot connect to IRC (No route to host). Retrying..' for both irc.freenode.net and my efnet server
13:49 marcus I can telnet to 6667 on the command line from the box
13:49 marcus ever had trouble like that?
13:49 sri nope
13:50 marcus :/
13:50 sri irc.freenode.org 6667
13:50 marcus hmm, I'll try changing it.
13:51 sri umm
13:51 sri .net
13:51 marcus irc.freenode.org and irc.freenode.net resolves to the same
13:51 sri yea, meant to say it's the same -.-
13:51 sri /fail
13:52 marcus chat.freenode.net has IPv6 address 2001:19f0:feee::dead:beef:cafe
13:52 marcus maybe it's trying to use ipv6 and failing
13:52 sri haha
13:56 und3f sri, just take a look at that pipeline command parsing i did and how "easy" it is with mojo https://github.com/und3f/mojox-redis​/blob/master/lib/MojoX/Redis.pm#L127
13:57 sri that code needs a serious cleanup :)
13:57 JonChisciotte joined #mojo
13:58 und3f sri, but what about main idea
13:58 und3f sri, of parsing
13:58 sri you mean switching the on_read callback?
13:59 und3f sri, the main idea was to create table based parsing
13:59 marcus left #mojo
13:59 sri can't really make much sense of it without reading everything :S
13:59 marcus joined #mojo
14:00 marcus left #mojo
14:00 sri looks ok at first sight though
14:01 und3f sri, and works too
14:01 sri und3f: but if you want me to understand it you need to start using perltidy and comments ;p
14:01 marcus joined #mojo
14:02 marcus yupp
14:02 marcus must have been ipv6
14:02 marcus changed to use ip instead of hostname
14:02 marcus and it works
14:02 sri yay for ipv6
14:02 marcus yupp, it's awesome
14:02 und3f sri, but the problem is that Mojo::IOLoop mechanisms are too limited. For applications like this need thing like in AnyEvent::Socket
14:03 sri und3f: what specifically?
14:04 sri i suspect you mean Handle and not Socket
14:04 sri which is just higher level abstraction
14:05 und3f sri, right! need something like ->push_read( chunk => 4 bytes )
14:06 und3f possible we can write another wrapper library for Mojo::IOLoop
14:06 und3f That will do such actions
14:06 sri thats trivial
14:06 sri like 10 lines of code or so
14:06 und3f sri, but will help a lot during developing low level apps
14:07 sri the hard part is actually writing tests for it ;p
14:11 sri well, for core this is too high level
14:11 sri but trivial to build around
14:13 sri on that layer i'm more interested in the node.js api
14:13 sri it is cleaner
14:16 sri api wise i think ioloop itself is pretty much complete
14:19 sri und3f: you bring up anyevent a lot, maybe you should just use anyevent? :)
14:20 perlrocks Twitter: "Released MojoX::Redis 0.2 #Redis client for #Mojolicious https://github.com/und3f/mojox-redis" --und3f http://twitter.com/und3f/status/915460967112705
14:21 und3f sri, i used anyevent a lot
14:21 sri i'm all for using the right too for the task, and sometimes mojolciious isn't ;)
14:21 sri *tool
14:22 sri you could also just mix both
14:23 marcus heresy!
14:23 sri the on_tick callback makes it rather trivial to use multiple event loops
14:24 sri ;p
14:27 sri what i don't like about push_read is that it discourages event loop agnostic protocol parsers
14:33 vel joined #mojo
15:10 JonChisciotte1 joined #mojo
15:11 JonChisciotte left #mojo
15:12 y0j left #mojo
15:15 spleenjack left #mojo
15:20 vel left #mojo
15:31 y0j joined #mojo
15:53 und3f is now known as und3f[A]
15:55 perlrocks Twitter: "Watching England v. New Zealand whilst trying to figure out how to use #mojolicious. @kraih, it needs something like http://bit.ly/c793EI" --purinkle http://twitter.com/purinkle/status/939392784928768
16:15 polvo can param() delete leading and trailing whitespaces from returned values?
16:17 sri no
16:49 * marcus can haz roaring fireplace
17:24 * sri can haz pizza
17:42 marcus can had pasta
18:08 jawnsy left #mojo
18:14 jawnsy joined #mojo
19:03 achromic left #mojo
19:04 kvorg joined #mojo
19:15 achromic joined #mojo
20:13 JonChisciotte1 Hi, I'd like to use Template Toolkit with Mojo and use the I18N mojo's plugin. Is it possible to use it as a tt's filter?
20:14 sri doubt that
20:14 JonChisciotte1 sigh
20:14 sri you prolly need a tt specific plugin for that
20:15 JonChisciotte1 ok, I'll google it
20:15 JonChisciotte1 is now known as JonChisciotte
20:15 sri not sure why you would want to use it as a filter though
20:15 sri seems rather ugly
20:17 JonChisciotte uhm, yep, I could find other ways (create a vmethod? moh...)
20:17 sri just call the helper
20:17 sri [% c.l('lalala') %]
20:19 JonChisciotte uh, it worked (not translating but this is another problem), why 'c'? Probably I've missed it in the doc..
20:19 sri controller instance
20:19 sri $self in ep templates
20:20 JonChisciotte ok, thanks!
20:21 JonChisciotte then another more important doubt: is bad to think to call a route inside another action? (like triggering an event) is it possible?
20:21 sri don't understand the question
20:26 JonChisciotte For instance, having $r->route('/login')->via('p​ost')->to('Session#login'); I'd like to call the same method within another action of another controller... I just though to use routes like events and trigger when needed without a request. Maybe you did not understand the question because what I want to do is virtually wrong :)
20:27 sri sounds very wrong
20:28 sri if you want internal redirects we don't do those
20:30 chansen woha, mojo got async dns queries  =)
20:30 sri no recursion though :)
20:31 JonChisciotte yep, I've misinterpreted Mojo thinking routes like commands in puremvc. My fault, I think I should use composition.. but is it correct to add a generic object as an helper and use it in various actions? For instance, $self->helper("model" => $dbic_object)?
20:31 sri we fallback to blocking if the dns server can't recurse for us
20:32 sri JonChisciotte: just set it as a default
20:32 sri $app->defaults(db => $dbic);
20:33 sri will be in the stash for every request then
20:33 chansen sri: better than only blocking synchronous resolves =)
20:33 sri indeed :)
20:34 chansen whats blocking recursive lookups?
20:34 sri adds a lot of complexity
20:34 kvorg left #mojo
20:35 chansen hmm
20:35 sri i'd rather add more record types first
20:36 sri currently we only have A, AAAA and TXT
20:36 chansen ok
20:37 JonChisciotte sri: ok, but is "defaults" still experimental as pod say?
20:38 sri thought about doing mdns natively instead of relying on an xs module... but lack of tuits :/
20:39 sri JonChisciotte: it's rather unlikely to change and will prolly lose experimental status soon
20:40 sri but yea, there is a slim risk
20:41 JonChisciotte sigh, I'm just afraid to refactor lots of code in the years to come, but I'll take the risk..
20:41 daviddelikat left #mojo
20:42 daviddelikat joined #mojo
20:42 sri always the risk with bleeding edge projects
20:43 sri no plans for big changes in the near future though ;)
20:44 sri 1.0 wants to be wrapped up
20:46 JonChisciotte glad to hear that, anyway in the worst case I'll camp here :)
20:47 sri :)
20:54 perlite joined #mojo
20:58 perlite_ left #mojo
21:01 esskar_ joined #mojo
21:02 esskar_ moin
21:03 polvo phoh9ooX
21:03 polvo wrong window :\
21:05 sri moin
21:05 sri now we know your password!
21:06 esskar_ t/mojolicious/websocket_tls_proxy_lite_app.t .. 1/16 <- just hangs forever (on windows)
21:06 polvo sri, aaaaah!
21:06 sri esskar_: latest?
21:07 esskar_ sri, just a fresh pull
21:07 esskar_ raises some meory error eventually
21:07 sri esskar_: considering the ode is pretty much equal to the non tls proxy test i guess your IO::Socket::SSL is screwed
21:07 sri *code
21:08 sri http://cpantesters.org/distro/M/Mojolicious.html # this looks rather good
21:10 sri not sure why but it seems rather common that people screw up their ssl modules :(
21:10 esskar_ :)
21:10 esskar_ well, so be it
21:11 sri we'll see, if it becomes too common the test becomes optional ;p
21:14 chansen sri: I think you need mx, srv and a in-order to do multicast-dns
21:15 sri chansen: ptr too i think
21:17 sri compression is especially annoying
21:17 chansen sri: yes, you are right, http://files.dns-sd.org/draf​t-cheshire-dnsext-dns-sd.txt
21:19 sri dns is really not as bad as i originally thought, but all the special cases add up
21:19 chansen true
21:19 sri recursion, compression, truncation...
21:19 chansen ;P
21:20 chansen the case for most software, +1 , +2, +3, -4, ... ;P
21:20 sri but at least it is all in the spec, not like http where you just have to go with trial and error xD
21:21 sri dnssec also looks like fun
21:21 chansen yeah, http is a big <insert favorite bad word> mess
21:22 chansen there is a spec and there is x implementations behaving in y was
21:23 chansen s/was/ways/
21:23 sri websockets going mainstream will make this so much worse
21:24 sri i can already see node.js server side dom implementations streaming dom events to the browser
21:25 sri we all will miss the good old screwed up but predictable http :D
21:25 chansen latest ws draft is a mess IMO, wtf are they introducing a "binary" protocol over HTTP?
21:26 sri first thing me and xantus thought "damn, that looks totally like google spdy"
21:26 chansen \x00 .. \xFF + handshake is god enough
21:26 chansen s/god/good/
21:26 sri yea, they should just have fixed the few remaining flaws and went with ietf-00
21:26 chansen yup
21:28 chansen it will en up as the uri template spec, ... nobody cares
21:29 sri all vendors except for microsoft have ietf-00 implemented now, so there is still a little hope
21:29 chansen s/cares/implement the extended spec/
21:29 esskar_ why does mojolicious have it's own dns resolver?
21:29 sri esskar_: non blocking
21:29 chansen sri: the handshake in current ws spec is IMO good, the framing and header part is not IMO
21:30 sri i'd like to see an http compliant handshake
21:30 sri good old upgrade
21:31 sri not sure why thats not enough
21:31 chansen sri: well I can understand the argument that WS isn't HTTP
21:31 sri not like it would allow more x protocol attacks than xhr
21:32 sri sure, but everybody will want to use the same server for both
21:32 chansen HTTP compliance would be simple, Content-Length header as mojo already does
21:33 sri last i looked headers were also still utf-8
21:33 chansen doesn't matter since latin-1 is within UTF-8
21:34 chansen 2616 says latin1 (ISO-8951)
21:34 chansen forgot a -1 at the end
21:35 sri oh right
21:39 sri dunno why i still follow the hybi list, it's rather depressing
21:40 sri json handshake was a funny idea though
21:40 chansen heh
21:41 chansen it wouldn't hurt if you provided a more robust validation/generation of sec keys
21:42 sri that part has been obsoleted in the latest draft anyway i think
21:42 sri all connect stuff now
21:42 chansen feel free to borrow/rewrite this http://nopaste.snit.ch/25356
21:43 sri thanks, but i think i'll wait for the next incarnation of the protocol
21:44 sri google folks are very keen on getting it adopted, chrome will break ietf-00 compatibility soonish i heard :S
21:45 chansen the handshake has been the same since 00, no need to wait, it will be the same, what's the current issue is the framing part
21:45 chansen or if you like -76
21:45 sri http://www.adambarth.com/e​xperimental/websocket.txt # thats -04 i think
21:46 sri draft upload is just disabled atm or so
21:46 chansen according to who, ietf?
21:47 sri think he's the editor
21:47 chansen no, Ian Hickson it the main author
21:48 vel joined #mojo
21:48 sri ian doesn't seem to be involved in the ietf process
21:48 chansen he is
21:49 chansen http://www.whatwg.org/specs/web-socket-protocol/
21:49 chansen he is employed by google
21:49 sri http://tools.ietf.org/wg/hybi/ # this is where the broser vendors meet atm
21:50 sri *browser
21:52 sri seems to be mainly ian fette and some other google folks editing the draft
21:52 chansen the only difference AFICS is the framing protocol, still the same
21:53 chansen ... handshake
21:54 sri i give it a few days before the connect handshake will be merged in ;p
21:54 chansen what is this: http://showmetheco.de/articles/2010/11/technical-d​etails-on-implementing-a-websocket-vnc-client.html
21:55 chansen ?
21:55 chansen is mojo to slow or to hatd to work around?
21:55 chansen s/hatd/hard/
21:55 sri i think he got annoyed we don't support draft 75 anymore
21:56 sri the static handshake messages were very annoying
21:58 sri in the end i don't think supporting obsoleted drafts is worth it yet, websockets are just not ready for prime
22:00 chansen I guess it depends what you want, iphone support latest draft handshake
22:01 sri draft 75 is mostly older android devices i think
22:01 spleenjack joined #mojo
22:01 sri and don't get me started on safari -.-
22:01 sri soooo many bugs
22:02 sri connections failing randomly in one direction is fun
22:04 MojoGuest963 joined #mojo
22:04 MojoGuest963 From: http://www.google.com/url?sa=t&amp;source=web​&amp;cd=5&amp;sqi=2&amp;ved=0CCYQFjAE&amp;url​=http%3A%2F%2Fdev.xantus.org%2F&amp;rct=j&amp​;q=websocket%20demo&amp;ei=y9DVTN-PJofqOdjBub​IJ&amp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (1 hits)
22:05 sri my bet is on "CONNECT 1C1BCE63-1DF8-455C-8235-08C2646A4F21.invalid:443 HTTP/1.1" as the new handshake
22:06 sri opera and goggle already accepted it
22:08 und3f[A] is now known as und3f
22:10 chansen well, dev.xantusorg doen't work on iphone 4.1
22:10 sri must be draft 75 then
22:10 sri apple doesn't care much about websockets ;p
22:11 sri after all they shipped safari 5.0 with 75 even though 76 has been in webkit nightly for quite some time
22:12 sri they just recently switched to 76 with 5.0.2
22:13 chansen dunno, not much do debug on unless I mirror my wlan port an sniff (using wireshark) which I don't have energy to do now
22:13 sri don't think there would be much of a point, everything will break within weeks anyway
22:14 esskar_ left #mojo
22:14 esskar_ joined #mojo
22:15 chansen I doubt Apple breaks things within weeks
22:15 sri who knows, they'll just inherit whatever webkit nightly supports
22:16 sri like they did with 5.0.2
22:16 chansen seems to contradict their policy
22:16 sri they prolly don't consider websockets a feature
22:16 chansen =}
22:16 sri like google
22:17 sri they might ven push a ietf-03 update with new framing and old handshake
22:17 sri *even
22:18 sri all they are worried about is draft 76 working for too long to be replaced again
22:18 sri which i think is a fair point
22:20 chansen I guess the future will tell us =)
22:20 sri indeed :)
22:21 sri i bet it will be at least two years before websockets go mainstream
22:21 chansen I don't agree, because industry needs it
22:22 sri i really hope i'm wrong, but i've followed this for almost a year now :S
22:24 MojoGuest963 left #mojo
22:24 sri what the hell
22:25 sri json in the handshake might actually happen
22:25 chansen what i think is that the current handshake will remain and the simple framing (\x00-\0xFF) will remain but the current framing (read support for other frame types))
22:25 chansen .. will drop
22:26 sri too obvious and simple solution xD
22:26 chansen no, just watch uri-templates ML, back to KISS
22:27 chansen it's better to have some supporters than none ;P
22:27 tempire hmm
22:27 tempire sri: check your paypal account
22:33 sri tempire: empty :/
22:34 tempire paypal fail
22:35 sri reminds me of that wepay marketing stunt
22:35 tempire ok, check now
22:35 sri http://techcrunch.com/2010/10/26/wepay-ice-paypal/
22:37 sri tempire: nope
22:37 tempire your donate form must be broken
22:37 tempire it says it was sent to sri@labs.kraih.com
22:38 sri oh
22:41 sri tempire: should be fixed
22:41 JonChisciotte is this the right way to pass config values to plugins? http://paste-it.net/private/v0f0b20/
22:42 sri JonChisciotte: plugin pod usually has examples
22:43 sri in the case of i18n there's also a good test
22:44 sri t/mojolicious/i18n_lite_app.t or so
22:45 sri it does look right though
22:45 JonChisciotte sri: it seems to me to following them but no one work for me, I'll check the one in t folder trying figure out why...
22:48 sri tempire: thanks :)
22:52 vel left #mojo
22:54 und3f is now known as und3f[A]
22:59 JonChisciotte For the TT plugin w/Mojolicious, options must be specified adding also the template_options key, ie: $self->plugin('tt_renderer' => {template_options => { FILTERS => ... For the I18N, still searching (and cannot understand why it relies on I18N::LangTags::Detect, moh)
23:00 sri it tries to detect the users language
23:00 Htbaa left #mojo
23:01 sri just a default you can override with the languages helper
23:03 JonChisciotte yep, but if my browser doesn't have that language my translation won't be accepted.. or so my firefox did: I had just english as language and had to add italian.. I think the problem is it doesn't get the default I specified. I'll keep searching...
23:08 sri set $self->languages('it') if you want to be sure
23:09 sri default is just something that gets appended to the detected languages, so whatever it finds will override the default
23:11 sri Zzz
23:11 sri nn
23:11 JonChisciotte where should I add the $self->lang? I've added it in startup method
23:12 JonChisciotte but the method languages is undefined
23:12 sri in the controller
23:12 sri or template
23:13 sri your mojolicious might be too old if it didn't work
23:14 sri anyway, now really Zzz
23:14 JonChisciotte ok, added in controller and languages has been set
23:14 JonChisciotte thanks sri, 'night
23:41 JonChisciotte left #mojo

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