Camelia, the Perl 6 bug

IRC log for #mojo, 2012-05-08

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

All times shown according to UTC.

Time Nick Message
00:11 andrefs joined #mojo
00:13 MisterHatt BlueT_: sorry for the delay, had to crash, using AnyEvent::IRC::Client and AnyEvent::MPRPC
01:13 scott joined #mojo
01:22 vel joined #mojo
01:42 d4rkie joined #mojo
01:50 xaka joined #mojo
01:58 arpadszasz joined #mojo
02:13 mire joined #mojo
02:29 * tempire hearts heroku
02:34 D4RK-PH0ENiX joined #mojo
03:23 d4rkie joined #mojo
03:35 scott_ joined #mojo
03:39 scott joined #mojo
03:46 scott_ joined #mojo
04:14 BlueT_ MisterHatt: the problem is that you can't pass the $res_cv to $irc->send_srv's callback to generate correct response, is it?
04:15 MisterHatt it's not the condvar, rather I want to define my own events that a user can register a callback for, and I don't know how that actually works
04:15 MisterHatt I suppose it's not build into AnyEvent anyway given most modules rely on Moose or Object::Event, but I see no Object::Event docu on such a thing
04:46 mire joined #mojo
04:47 BlueT_ MisterHatt: just get the code ref, put it somewhere (maybe hash), and call $hash{$name}->(@args) when you need it?
04:53 MisterHatt hmm, pretty sure that's what I'm already doing
04:53 MisterHatt like, I can register/call the callback, that's no issue, I just dont get how to trigger it
04:53 MisterHatt probably missing some fundamental understanding of what an event actually is...
05:10 scott joined #mojo
05:28 crab i think the waypoint deprecation warning should be worded more strongly
05:28 crab since the behaviour of existing code actually changes
05:30 crab (whether intended or not, i haven't tried to figure out)
05:31 * crab rewrites another app to avoid waypoints
05:54 crab fortunately it's easy to do so.
05:55 crab i wish i could create multiple sub-routes under a single route without thinking up a variable name for the parent
05:56 crab (but, not very strongly)
06:28 mire joined #mojo
06:46 Htbaa joined #mojo
06:49 espent joined #mojo
06:51 cosmincx joined #mojo
07:02 taryk joined #mojo
07:09 batman joined #mojo
07:09 pau4o joined #mojo
07:18 MisterHatt anyone know if setting  status code in render_json actually sets the response code to the same?
07:23 inokenty joined #mojo
07:24 fhelmber_ joined #mojo
07:53 gangabass joined #mojo
07:56 gangabass hi there! i'm just starting to use mojo::useragent and i think it's cool
07:57 gangabass but can it send forms for me like www::mechanize submit_form method? i mean i just set couple of fields i need but all rest fields (including hidden) will be sended with default values...
08:01 crab no, it can't do that.
08:07 Vandal joined #mojo
08:42 * tempire floats
08:43 nic tempire: Can you use perl on heroku (eg mojolicious)?
08:44 Foxcool joined #mojo
09:27 alexeiras joined #mojo
09:32 * sri yawns
09:36 * crab tosses beans at sri
09:37 * sri starts coughing
09:56 heytrav joined #mojo
10:27 purl joined #mojo
11:06 GabrielVieira joined #mojo
11:24 GitHub148 joined #mojo
11:24 GitHub148 [mojo] kraih pushed 1 new commit to master: http://git.io/7egp1Q
11:24 GitHub148 [mojo/master] fixed small cookie formatting bug - Sebastian Riedel
11:24 GitHub148 left #mojo
11:28 andrefs joined #mojo
11:33 crab sri: what do you think about printing a stronger warning about the use of waypoints?
11:34 sri crab: i think you might be the only one using waypoints
11:35 sri it's been a month, and nobody even asked once
11:35 crab ok, that's good then.
11:37 sri i'm more concerned about cookie formatting atm
11:39 sri it seems everybody lowercases expires, domain, path and secure
11:39 crab yes
11:39 sri i wonder if that's cargo culting or if there really are user agents depending on it
11:41 crab i've never found a user agent that would be affected (this, after looking at many cookie parsers to see what they did, though admittedly it was years ago).
11:50 andrefs joined #mojo
12:04 GitHub153 joined #mojo
12:04 GitHub153 [mojo] kraih pushed 1 new commit to master: http://git.io/NZmuYQ
12:04 GitHub153 [mojo/master] be more conservative when generating cookies - Sebastian Riedel
12:04 GitHub153 left #mojo
12:12 migo joined #mojo
12:25 migo joined #mojo
12:32 GitHub45 joined #mojo
12:32 GitHub45 [mojo] kraih pushed 1 new commit to master: http://git.io/AiSOUg
12:32 GitHub45 [mojo/master] be even more conservative - Sebastian Riedel
12:32 GitHub45 left #mojo
12:33 andrefs joined #mojo
12:51 GitHub5 joined #mojo
12:51 GitHub5 [mojo] kraih pushed 1 new commit to master: http://git.io/3JJTVg
12:51 GitHub5 [mojo/master] fixed small bug in cookie parser - Sebastian Riedel
12:51 GitHub5 left #mojo
13:48 andrefs joined #mojo
14:07 vel joined #mojo
14:25 fhelmber_ joined #mojo
14:42 jwang joined #mojo
14:51 marcus hel0
14:52 rem_lex|pivo joined #mojo
15:03 crab hi marcus
15:15 OlegG joined #mojo
15:16 xaka joined #mojo
15:17 OlegG is it ok that html_escape("<") returns "&LT" without semicolon at the end?
15:18 crab it doesn't for me. what version are you using?
15:19 OlegG how can I check version :) ?
15:20 OlegG Found, version 2.92
15:24 OlegG So, should I submit bugreport for html_escape problem?
15:25 Andrei666 joined #mojo
15:28 crab please do.
15:30 marcus hi crab
15:40 OlegG Ok, issue 326 created
15:42 inokenty I think the 'problem' is in file 'entities.txt'
15:45 crab more specifically, it's in how $REVERSE is built from entities.txt in Mojo::Util
15:46 OlegG inokenty: I don't think so, entities without semicolon used by html_decode
15:46 inokenty ...
15:46 inokenty 1058   LT; U+0003C
15:46 inokenty 1059   LT U+0003C
15:46 inokenty 1060   Lt; U+0226A
15:46 inokenty 1061   lt; U+0003C
15:46 inokenty Sorry
15:47 crab inokenty: right, and that's fine. it needs to recognise all of those.
15:47 crab (though i don't know what's up with Lt; being a different value than LT. weird.)
15:49 Psyche^ joined #mojo
15:54 alexeiras left #mojo
15:55 sri OlegG: that's not a bug
15:56 alexeiras joined #mojo
15:57 sri i'm open for suggestions how to handle entities without ; better though
15:58 sri crab: the html5 spec has a list of which entities are allowed without ;
15:58 crab it may not be illegal to generate &LT, but (a) we used to generate &lt; and it's an unexpected change, and (b) unlike cookie-case, there's plenty of code that breaks in parsing entities, so there's good reason to generate the ;
15:58 sri we extract the entity list directly from the spec
15:58 sri crab: you're welcome to improve it
15:59 crab i know. but the change oleg suggests (ignoring keys without /;$/ when building $REVERSE) seems sensible.
15:59 crab if you don't object or don't have a better idea, i'll do that.
16:00 sri no objections
16:00 crab doing now.
16:05 JerJer joined #mojo
16:10 pau4o left #mojo
16:12 GitHub184 joined #mojo
16:12 GitHub184 [mojo] amenonsen pushed 1 new commit to master: http://git.io/_eAzZA
16:12 GitHub184 [mojo/master] Generate &LT; instead of &LT for html_escape('<') (closes #326) - Abhijit Menon-Sen
16:12 GitHub184 left #mojo
16:12 crab i didn't bother to try to make it generate &lt; (lowercase)
16:13 sri that would be rather tricky
16:16 crab yes. doable by the equivalent of select array_agg(entity) group by value and picking the lowercase-est version in the array for %REVERSE. but didn't seem worthwhile.
16:20 GitHub132 joined #mojo
16:20 GitHub132 [mojo] kraih pushed 1 new commit to master: http://git.io/fJi1DA
16:20 GitHub132 [mojo/master] replaced expression with block - Sebastian Riedel
16:20 GitHub132 left #mojo
16:22 OlegG may be without grep :) ? smth like /;$/ and $REVERSE{$ENTITIES{$_}} //= $_
16:23 sri grep is fine, expressions are just bad style
16:24 batman joined #mojo
16:27 * crab 's expressions are stifled :-(
16:28 sri :p
16:29 stephan48 joined #mojo
16:30 sri blocks without comma are weird too, but i agree with damian that expressions are worse
16:33 crab i didn't know he said they were bad. why?
16:33 sri it's in PBP
16:34 sri "Mapping and Grepping"
16:35 crab ok.
16:35 sri "But when the first argument to a map or grep is specified as an expression, it becomes harder to distinguish from the remaining arguments:"
16:35 sri his examples are rather convincing
16:36 crab ah.
17:17 stephan48 a actionless route mapping directly to a template.. howto accomplish that?
17:21 JerJer stephan48:   if i follow what you mean, one could simply have a properly named template
17:21 stephan48 ah ok
17:21 stephan48 that works
17:23 JerJer i use that a lot :)
17:31 batman joined #mojo
17:31 mire joined #mojo
17:35 tempire nic: yes, mojoexample and mojocasts are hosted on heroku
17:36 tempire https://github.com/judofyr/perloku
17:52 d4rkie joined #mojo
18:12 vel joined #mojo
18:22 andrefs joined #mojo
18:32 vel joined #mojo
18:47 mire joined #mojo
18:55 marcus hai.
18:55 * marcus watches house.
18:55 marcus 3 episodes to go.
19:02 mire joined #mojo
19:32 tempire I stopped watching house aeons ago
19:32 marcus They are actually making it a bit exciting towards the end :)
19:35 * sri bets house will be diagnosed with lupus in the final episode
19:41 dmn001 joined #mojo
19:42 dmn001 how do i add a html title in mojo?
19:43 JerJer joined #mojo
19:47 rem_lex| joined #mojo
19:49 dmn001 im reading this: http://mojolicio.us/perldoc/Mojoli​cious/Plugin/DefaultHelpers#title but i cant get it to work
19:51 dmn001 just going to put in plain text instead: <!DOCTYPE html>  <html>    <head><title>Search</title></head>
19:53 fibo joined #mojo
20:01 marcus Sri: actually, Wilson has cancer.
20:01 sri SPOILER!
20:01 purl i think spoiler is Hermione dies or ron dies or harry dies or hagrid dies or JK Rowling dies and never bothers us with her crap again or hermoine can't contain herself any longer and does ron's corpse with a strapon
20:02 tempire oh, this is the last season?
20:02 tempire I may have to tune in for that
20:02 sri yea
20:02 sri EVERYONE DIES!
20:03 tempire I'll wait until its done and then go through them.
20:03 tempire I just finished a run through of supernatural from episode 1
20:03 tempire ran out of shows :(
20:12 marcus Not a problem I have.
20:13 marcus Graphite looks neat.
20:26 tempire https://launchpad.net/graphite ?
20:29 JerJer1 joined #mojo
20:33 xaka joined #mojo
20:40 batman joined #mojo
20:41 quicoju joined #mojo
20:49 sri oh my, 55000 twitter passwords leaked :o
21:04 dmn001 thats only a v. small % of twitter accts
21:06 sri oh, then it's alright i guess -.-
21:07 dmn001 anyway was wondering, is it possible to render the response time it took?
21:08 dmn001 e.g.  22:03:05 2012] [debug] 200 OK (2.682748s, 0.373/s).
21:08 dmn001 put that into the browser window
21:08 dmn001 e.g. this Search took 2.6 seconds
21:09 dmn001 i guess one way would be to use Time::HiRes and time it manually
21:10 dmn001 and put the time in the stash
21:11 sri ooooooh, websocket inspector already works with webkit nightly, neat
21:16 dmn001 btw those twitter pws look like fake accts, generated by bots
21:52 lukep joined #mojo
22:16 JerJer joined #mojo
22:18 fibo hi all, 0.18 here
22:19 fibo i posted a question on the mailing list about overriding Mojo::Command namespace
22:19 fibo i've got no answer, can you tell me what you think?
22:23 sri fibo: http://mojolicio.us/perldoc/Mojolicious#commands
22:27 JerJer1 joined #mojo
22:28 fibo mmh, I override the namespace, not adding more namespaces : https://groups.google.com/forum/?from​groups#!topic/mojolicious/pEmz47EoR9s
22:30 fibo for instance I just left the "daemon" command
22:31 JerJer joined #mojo
22:40 fibo well, for me it works, I thought it could be interesting for somebody
22:45 sri custom scripts are not recommended anymore, just customization of app->commands
22:45 sri but whatever works for you
23:33 fibo ok, 10x, I will try in the recommended way
23:54 dmn001 trying to change it to port 80, but it says: Can't create listen socket: Address already in use
23:55 dmn001 Can't create listen socket: Address already in use at /usr/local/share/perl/5.10.1/Mojo/IOLoop.pm line 155
23:56 dmn001 ah. apache was running.

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