Camelia, the Perl 6 bug

IRC log for #mojo, 2010-08-21

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

All times shown according to UTC.

Time Nick Message
00:27 xantus hm
00:38 sri xantus: feel free to add it properly again
00:40 sri reading the websocket mailing list is really demotivating, everything will just break again :(
00:43 xantus k
00:44 xantus I have an upcoming mojo paste rewrite, and I want as many supported ws clients as I can support
00:47 sri sure, i removed it after getting too many bug reports for draft 75 support
00:47 sri having no tests makes it pretty much unsupportable long term
00:48 sri if i can forward complaints to you thats fine with me :D
00:49 sri now i think most reports are because of bugs in the safari implementation
00:49 sri guess we could also call it deprecated and unsupported draft 75 support
00:50 sri your biggest problem will be headers, they need to be sorted for the handshake
00:51 sri anyway, really need to sleep
00:51 sri nn
00:54 sri btw. i think it's only chrome 5 that has working draft 75 support, safari 5 is just broken and won't work with anything
02:35 janus left #mojo
02:42 janus joined #mojo
03:44 alnewkirk left #mojo
03:44 alnewkirk joined #mojo
04:03 tl joined #mojo
04:05 tholen42 left #mojo
04:18 flu_ joined #mojo
07:14 Alias__ left #mojo
07:25 alnewkirk left #mojo
07:25 [1]awnstudio joined #mojo
07:31 dotan joined #mojo
07:31 dotan left #mojo
07:32 dotan joined #mojo
07:44 yko joined #mojo
08:11 dotan left #mojo
09:11 yko left #mojo
09:11 yko joined #mojo
09:31 [1]awnstudio left #mojo
09:32 [1]awnstudio joined #mojo
09:52 [1]awnstudio left #mojo
10:01 Foxcool joined #mojo
10:02 MojoGuest580 joined #mojo
10:02 MojoGuest580 From: http://www.google.com/search?hl=en&rlz=1C1​GGLS_enUS359US360&q=websocket+irc+html5&am​p;aq=f&aqi=&aql=&oq=&gs_rfai= (1 hits)
10:05 MojoGuest580 test
10:05 MojoGuest580 hrm wild irc chat client
10:06 MojoGuest580 is now known as testing
10:07 testing left #mojo
10:08 perlrocks Twitter: "Saturday shoutout to my #Mojo-licious moviemaker friend! "@danperezfilms: One word: "Mojo-licious": @ItaliaTrent #Film #Writer #Mojolicious”" --ItaliaTrent http://twitter.com/ItaliaTrent/status/21738557542
10:35 * sri yawns
10:42 perlrocks Twitter: "ー → blog written by [perl] Mojo + DBIx:: Skinny + Test:: mysqld can http://bit.ly/cHJoDm Introduction" (ja) --toritori0318 http://twitter.com/toritori0318/status/21739981768
10:42 perlrocks Twitter: "RT @ toritori0318: → ー blog written by [perl] Mojo + DBIx:: Skinny + Test:: mysqld can http://bit.ly/cHJoDm Introduction" (ja) --AE_PerlIsm http://twitter.com/AE_PerlIsm/status/21739981986
10:45 perlrocks Twitter: "RT! Toritori0318: → ー blog written by [perl] Mojo + DBIx:: Skinny + Test:: mysqld can http://bit.ly/cHJoDm Introduction" (ja) --perlism http://twitter.com/perlism/status/21740071660
10:53 janus yey, first time used the git stash to keep the kqueue tests enabled :)
10:57 Foxcool left #mojo
11:05 sri :)
12:30 perlrocks Twitter: "Mojo + DBIx:: Skinny + Test:: mysqld such that - perl http://htn.to/c7TB1T study notes" (ja) --koba04 http://twitter.com/koba04/status/21744991563
12:30 perlrocks Twitter: "RT @ koba04: Mojo + DBIx:: Skinny + Test:: mysqld こと はじめ - perl 勉強 メモ http://htn.to/c7TB1T" (id) --AE_PerlIsm http://twitter.com/AE_PerlIsm/status/21744991890
12:47 baton8 left #mojo
12:47 baton8 joined #mojo
12:56 perlrocks Twitter: "Extended long polling support in #mojolicious again. :) http://bit.ly/9lUxbM #perl" --kraih http://twitter.com/kraih/status/21746381017
13:11 sri http://github.com/jezdez/textmate-missingdrawer # now thats a cool project sidebar
13:21 janus hm, is "Host: foo\r\n :3000\r\n" really acceptable?
13:23 janus not that i've seen that in the wild, just trying to confuse parsers :>
13:31 sri janus: yes it is
13:31 sri it might not be an acceptable Host header though
14:11 crab sure it is.
14:21 janus at least it doesn't match the host header spec which doesn't seem to allow spaces there
14:31 sri but thats another layer
14:41 janus apache and lighttpd reply with 400, nginx doesn't seem to care
14:43 janus depends on the abstraction :)
14:48 sri implementing rfc2616 is an art :)
14:58 janus as is writing parsers and lexers :)
15:07 crab janus: you cannot read the rfc2616 abnf that way
15:07 crab janus: looks for "implied LWS". it is allowed between any two tokens or between a word and a separator, even though the ABNF for that production may not mention it explicitly.
15:08 crab s/looks/look/. section 2, notational conventions, iirc.
15:08 janus hmm, good point
15:08 crab janus: this is one of the most confusing things about http, which many implementors fail to understand
15:08 crab also, it is horrible, so some of the ones who understand it pretend they don't.
15:09 crab http://www.and.org/texts/server-http <- a description of how horrible things are
15:14 janus yikes... now that's a ton of arguments justifying the additional parsing buffer i thought of, thanks a lot!
15:16 crab always happy to help. :-)
15:21 janus so far still no showstoppers for my http / lalr(1) implementation...
15:23 crab where is it?
15:25 janus still not public, really not very much more than a proof of concept as of now
15:26 janus no replies or anything, only parsing requests. i'm still checking where the parser and the tokenizer need to interact more...
15:27 crab oh. too bad. i'd've liked to see it, just out of curiosity.
15:28 janus well, it's highly experimental, but if you don't fear incomplete and leaky code i can send it to you
15:28 crab sure.
15:29 crab ams@toroid.org
15:43 janus sent, but without warranty :)
15:45 janus and no comments on anything but the idea please ;)
15:45 crab have you seen ryan dahl's http parser library?
15:49 janus yep, definitely one of the nicer implementations
15:53 janus yet no separation of code and rules, which is one of the reasons why i like yacc and lemon very much :)
15:58 koban joined #mojo
16:02 janus crab: let me know what you think
16:07 crab janus: will do.
16:07 koban left #mojo
16:13 vel joined #mojo
16:23 vel left #mojo
16:51 vel joined #mojo
16:54 sri http://compbio.cs.uic.edu/~mayank/4sq.html # haha, perl -Mojo oneliner anyone? :)
17:14 vel left #mojo
17:14 vti beer time
17:26 sri my favorite time of the day
17:36 vti broke anything today? ;p
17:50 sri nope, i think we are done :)
17:51 sri unless someone here can point out any other design limitations that will prevent future optimizations
17:51 sri but my list is empty atm
18:02 vel joined #mojo
18:21 achromic left #mojo
18:22 mib_fmkqbk joined #mojo
18:22 achromic joined #mojo
18:26 mib_fmkqbk left #mojo
18:29 perlrocks Twitter: "♺ @kraih: #perl developers really need to get rid of the cgi mindset, this is the #html5 decade. #mojolicious" --arpadszasz http://twitter.com/arpadszasz/status/21768956608
18:35 GitHub182 joined #mojo
18:35 GitHub182 mojo: master Sebastian Riedel * 3257e5e (1 files in 1 dirs): pod cleanup - http://bit.ly/dCEAHV
18:35 GitHub182 left #mojo
18:44 sri doesn't get much easier than this :)
18:48 GitHub50 joined #mojo
18:48 GitHub50 mojo: master Sebastian Riedel * 83e2e50 (1 files in 1 dirs): pod cleanup - http://bit.ly/dvs2tm
18:48 GitHub50 left #mojo
18:49 yko One day ill get some beer, lock myself in my room and learn how all that stuff works...
18:49 sri it's actually quite simple :)
18:50 sri beer is the key to understanding it though
18:51 sri !tweet the key to truly understanding #mojolicious is beer. #perl
18:51 perlrocks sri: http://twitter.com/perlrocks/status/21770233385
18:51 perlrocks Twitter: "the key to truly understanding #mojolicious is beer. #perl --sri" --perlrocks http://twitter.com/perlrocks/status/21770233385
18:52 yko sri++
19:02 sri btw. write/write_chunk work across all server bindings
19:02 sri even the drain callback
19:02 sri most servers will buffer all cgi output though
19:02 sri but if it's supported we can do it :)
19:05 sri http://pastie.org/1106427
19:05 sri same for fastcgi and psgi
19:21 GitHub152 joined #mojo
19:21 GitHub152 mojo: master Sebastian Riedel * 3ef7fca (1 files in 1 dirs): cleanup - http://bit.ly/b1IBTL
19:21 GitHub152 left #mojo
20:05 DaTa so much cleanup, the source code must be a mess ;)
20:05 sri not anymore ;)
20:36 fod joined #mojo
21:14 yko left #mojo
21:20 MojoGuest316 joined #mojo
21:20 MojoGuest316 From: http://www.google.co.uk/search?hl=en&am​p;safe=off&amp;q=websocket+irc&amp;aq=f​&amp;aqi=&amp;aql=&amp;oq=&amp;gs_rfai= (1 hits)
21:21 MojoGuest316 left #mojo
22:41 vel left #mojo
22:42 vel joined #mojo
22:49 dotan joined #mojo
23:08 su-bzero_ joined #mojo
23:13 su-bzero left #mojo
23:23 dotan left #mojo
23:31 jawnsy left #mojo
23:37 jawnsy joined #mojo
23:57 vel left #mojo

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