Camelia, the Perl 6 bug

IRC log for #mojo, 2011-03-21

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

All times shown according to UTC.

Time Nick Message
00:00 damog joined #mojo
00:05 janus left #mojo
02:08 damog left #mojo
03:09 metaperl left #mojo
03:20 metaperl joined #mojo
03:28 metaperl left #mojo
03:41 yakudzo joined #mojo
03:48 metaperl joined #mojo
04:09 metaperl left #mojo
04:11 ok22 left #mojo
04:16 metaperl joined #mojo
04:36 metaperl left #mojo
05:01 giwi left #mojo
05:51 giwi joined #mojo
06:06 ysyrota joined #mojo
06:17 koban` joined #mojo
06:37 kaare joined #mojo
07:07 ver joined #mojo
07:12 minimalist left #mojo
07:18 trcjr left #mojo
07:20 cosmincx joined #mojo
07:25 fhelmber_ joined #mojo
07:26 fhelmber_ left #mojo
07:26 fhelmber_ joined #mojo
07:36 breaker313 joined #mojo
07:42 esskar joined #mojo
07:46 Christian joined #mojo
07:46 Christian morning
07:54 ok22 joined #mojo
07:57 Drossel left #mojo
07:58 Kulag joined #mojo
07:59 metaperl joined #mojo
08:01 zdk joined #mojo
08:04 koban` left #mojo
08:05 koban` joined #mojo
08:14 gabiruh left #mojo
08:14 gabiruh joined #mojo
08:47 crab hi
09:40 abra left #mojo
09:51 abra joined #mojo
10:10 Christian left #mojo
10:10 Christian joined #mojo
10:23 spleenjack joined #mojo
10:41 ok22 left #mojo
10:52 ok22 joined #mojo
10:53 breaker313 left #mojo
11:02 zdk left #mojo
11:15 sri good morning fellow evil supercomputers
11:54 GitHub177 joined #mojo
11:54 GitHub177 mojo: master Sebastian Riedel * 05a527b (1 files in 1 dirs): added test case - http://bit.ly/glAUOV
11:54 GitHub177 left #mojo
12:24 Debolaz Moin.
12:27 moritz good localtime()
12:43 daviddelikat left #mojo
12:59 marty joined #mojo
13:08 ispy_ joined #mojo
13:13 jfuller joined #mojo
13:16 jfuller Good morning. I'm hoping someone can help me out with a routing issue. Basically I want to have a default "404" route. And, I thought I had it when I defined it as "any '/(*)'" but it seems like this also intercepts my "post '/'" route, which I didn't want to do. What the best way to go about solving this issue?
13:17 moritz s/any/get/ ?
13:19 moritz what about other request types like PUT or DELETE? do you want them to take the default route too?
13:19 jfuller moritz: Well, I want to catch post requests for unknown routes to. Is there an equivalent for "one or more characters" in routes?
13:20 moritz $r-route('/:number', number => qr/\d+/) # from Mojolicious::Guides::Routing
13:21 moritz application to your problem is left as exercise...
13:21 sri /(*) is not a valid pattern
13:22 sri /(*foo) would be
13:22 sri it's very well documented
13:22 jfuller sri: What's the best way to say "if it doesn't match anything else go here"
13:23 sri any '/(*foo)'
13:24 * Debolaz wish Rails had better error messages than "undefined method `to_sym' for false:FalseClass (NoMethodError)" for indicating errors in a translation.
13:25 daviddelikat joined #mojo
13:27 sri Debolaz: wrong channel
13:31 Akron joined #mojo
13:41 Debolaz sri: What, badmouthing certain aspects of ruby web frameworks is a nono here?
13:43 giwi left #mojo
13:43 giwi joined #mojo
13:49 sri Debolaz: it's not really constructive, if you really wish those error messages were better you should tell them
13:50 GitHub30 joined #mojo
13:50 GitHub30 mojo: master Sebastian Riedel * 8e8cd86 (5 files in 4 dirs): fixed small Mojo::Util bug - http://bit.ly/glzjwc
13:50 GitHub30 left #mojo
13:58 Debolaz sri: Fair.
14:02 breaker313 joined #mojo
14:09 dekimsey joined #mojo
14:11 koban` left #mojo
14:11 koban` joined #mojo
14:13 koban` left #mojo
14:16 giwi left #mojo
14:18 koban` joined #mojo
14:24 koban` left #mojo
14:26 koban` joined #mojo
14:34 giwi joined #mojo
14:37 kaare left #mojo
15:03 damog joined #mojo
15:12 kaare joined #mojo
15:13 sri oh, lots of progress in the websocket working group
15:13 sri the latest spec doesn't look bad at all
15:13 sri and it might even be final :D
15:14 moritz so it will take only about 5 years before we can reasonably expect it to be in many browsers in that form...
15:14 sri guess it's time to break everything again and move to the -06 version
15:14 sri don't think so
15:15 sri common browsers auto update now
15:15 sri it will be out very quickly
15:15 sri even microsoft already has a patch for ie9
15:15 * moritz still uses debian, and the firefox that comes with it
15:16 sri firefox 4 auto updates
15:17 moritz so do older firefoxes... but not if you've installed as root, and then run as a user
15:21 sri that sucks for you
15:21 ver left #mojo
15:22 sri i'm certain the majority will auto update though, forcing you to update too or miss out on the latest features
15:23 Drossel joined #mojo
15:23 Kulag left #mojo
15:24 tempire sri: the websocket spec stuff, you're getting that from the mailing list, right?
15:25 sri right
15:25 sri i'm playing with the new handshake now
15:25 sri looks solid
15:25 tempire where's the web front end for the mailing list?
15:25 tempire (url)
15:25 sri google for hybi
15:25 tempire are they still trying to re-engineer a message queue?
15:26 sri think i'll update mojolicious to draft 06
15:26 sri nope they are pretty much done now
15:26 sri the spec is expected to be finalized during the meeting in a few days
15:29 ryanc left #mojo
15:29 ryanc joined #mojo
15:32 tempire so...the latest spec was released on feb 25?
15:35 * sri nods
15:37 tempire they should publish a test with every spec, so everyone can be sure their implementation is on target
15:38 sri not really
15:38 sri there are unofficial browser builds already
15:39 moritz test would be awesome, but probably not so easy to write
15:39 breaker313 left #mojo
15:39 sri there are dozens already
15:39 sri jetty for server side, firefox minefield for client
15:39 sri as an example
15:40 sri but you don't have to care anyway :)
15:40 sri i'll have a good test suite soonish
15:42 minimalist joined #mojo
15:46 Christian left #mojo
15:49 sri yay 06 handshake already working :)
15:49 sri requires perl 5.10 though
15:54 duff joined #mojo
15:55 duff how do I turn logging off completely?
16:01 koban` left #mojo
16:02 duff left #mojo
16:10 cosmincx left #mojo
16:18 crab sri: what did they change? all the earlier problems are resolved?
16:19 crab and what is it that requires perl 5.10?
16:19 crab (didn't you plan to require 5.10?)
16:19 sri absolutely everything changed
16:19 sri whole new protocol
16:20 sri it uses sha1 checksums
16:20 crab speaking of which, i have a friend who is forced to develop using 5.8.x because his sysadmins insist on using the redhat-packaged perl in some old version of rhel that still happens to be supported.
16:20 sri Digest::SHA is 5.10+
16:20 moritz isn't sha1 already mostly broken?
16:20 sri who cares?
16:20 purl not me
16:20 * moritz
16:21 sri still good for non crypto tasks and hmac
16:22 sri heck, even md5 is good with hmac
16:22 moritz ah, I kinda assumed it was used for something crypty
16:22 sri nobody sane would do that
16:22 crab i wish sha3 were out already. i'm not terribly happy with the idea of using sha2-256 and the like.
16:23 crab i like the idea of websockets. i think i'm going to use them as soon as i can, for *something*
16:24 crab unfortunately, i've been having lots of ideas for them in my current project(s), but i think "oh, they don't wor" and i don't write them down
16:24 sri i bet people will cry when evil me breaks websockets in mojolicious :)
16:24 moritz sri: just get a new release out which deprecates the old one, and then immediately a new, major release :-)
16:25 crab wow, moritz, that's amazingly evil.
16:25 sri naah, i'll just break them
16:25 sri websockets are by nature experimental
16:25 crab agreed
16:25 * moritz bows to crab
16:25 crab (of course)
16:25 purl (of course) is expecting to spend tomorrow doing detail-oriented programming work
16:25 crab purl, forget (of course)
16:25 purl crab: I forgot (of course)
16:26 giwi left #mojo
16:26 esskar left #mojo
16:26 sri and with a bit of luck it might be the last time
16:26 sri from what i heard 06 has very very good chances to go stable
16:27 giwi joined #mojo
16:28 stephan48 sri: will websockets work then without a flash policy server?
16:29 sri stephan48: that question makes no sense
16:30 sri websockets don't have anything to do with flash
16:30 stephan48 at the moment i recall needing some sort of flash policy server for working websockets under some browsers... will that still be required in future versions?
16:31 moritz depends on whether that browser continues to be broken, I'd guess :-)
16:31 sri that's wrong
16:31 purl sri is channeling thoth!
16:31 sri some flash workaround your app uses might need hacks
16:32 sri but websockets never did or will
16:32 stephan48 ah
16:42 jfuller left #mojo
16:54 minimalist Gentlemen, I have 1000 image files in my public folder and they load very slow because they go through the router. How can I mark them as static files?
16:57 moritz a very simple solution is a static.yourdomain.tld host with a separate server
16:59 crab minimalist: what i do is have an after_static_dispatch hook that sets an Expires/max-age far in the future so that my frontend server caches them
16:59 minimalist I'm using daemon --reload. Could this be the reason why all modules and config files load for each gif file?
17:03 minimalist moritz: although, this is a solution, it seems like it's running away from the problem
17:03 moritz minimalist: the first step towards a solution is not to use --reload
17:04 moritz which I've experienced to be not very robust
17:04 minimalist moritz: if this is the reason why this is happening, I can live with that. I'm using it during development only.
17:05 moritz minimalist: just try it witout --reload. If the effect is large, you'll immediately notice a difference
17:05 * moritz -> afk
17:06 minimalist moritz: I've tried it and it does make a difference. It would be nice to be able to use --reload and not have the entire site load for each single gif file.
17:22 sri umm
17:22 sri minimalist: you make no sense
17:22 crab a minimal amount of sense?
17:22 sri the whole purpose of --reload is to *reload*
17:23 minimalist sri: --reload seems to reload even for static files. That doesn't make sense.
17:23 sri umm
17:23 sri the static dispatcher is part of the app
17:24 sri hooks can actually interact with it
17:24 sri like crab said, you're using the wrong tool for the job
17:25 sri you're expecting normal web server behavior from an application server
17:26 minimalist I understand that. I was just wondering if there was a way around it, to make my life easier during development.
17:26 sri of course, don't use --reload
17:28 sri there are always tradeoffs
17:30 minimalist I'm still learning here, people. Bear with me!
17:30 ysyrota left #mojo
17:33 sri sorry if i did sound a bit rude, that always happens when i try to explain something :/
17:33 sri gonna stick with "impossible" next time ;p
17:36 minimalist sri: you didn't sound rude. thanks for your help. crab and moritz too.
18:17 Akron left #mojo
18:32 janus joined #mojo
18:34 spleenjack left #mojo
19:13 ok22 left #mojo
19:23 fhelmber_ left #mojo
19:24 ok22 joined #mojo
19:47 ispy_ left #mojo
19:49 daviddelikat left #mojo
20:38 grim_fandango joined #mojo
20:40 grim_fandango If I set up $self->defaults->({foo => 'bar'}); in startup() shouldn't my controllers be able to access this via $self->defaults->{foo} ?
20:40 grim_fandango Or am I missing something here?
20:43 minimalist grim_fandango: The way this is done is with a helper.
20:43 grim_fandango Oh wait ... defaults go into the stash, right?
20:44 grim_fandango So I really need to do $self->stash->...
20:44 crab grim_fandango: yes, the defaults are set in the stash
20:44 grim_fandango I am trying to put a global database handler for the controllers somewhere.  I think I read that this is usually done via defaults.
20:45 crab not $self->defaults->({foo => 'bar'}) but $self->defaults({foo => 'bar'})
20:45 crab but that was probably just a typo
20:45 grim_fandango Yep, it was.
20:45 crab grim_fandango: read http://toroid.org/ams/etc/mojolicious-db-handles
20:45 grim_fandango Okay, looks like $self->stash->{default} seems to work.
20:45 grim_fandango Will do, thanks crab.
20:47 minimalist crab: is defaults a good way to pass a database handle?
20:48 grim_fandango It was suggested so here by sri but that may have changed -> https://groups.google.com/group/mojolicious/​browse_thread/thread/6cd4ef35d84a24ef?fwc=1
20:48 crab minimalist: it's unnecessary
20:49 crab i mean, if you can say $c->db, why would you want to say $c->stash->{db} instead?
20:50 grim_fandango Makes sense to  me.
20:50 grim_fandango The less syntactic sugar the better.
21:01 stephen left #mojo
21:02 grim_fandango Huh. Can't locate object method "db" from $self in a controller method.
21:03 grim_fandango Of course.
21:03 purl Indubitably.
21:05 grim_fandango I missed $self->app, duh.
21:05 grim_fandango Okay, looks good, thanks folks.
21:26 mattastrophe joined #mojo
21:37 grim_fandango left #mojo
21:58 meraxes left #mojo
22:02 damog left #mojo
22:02 perlrocks Twitter: "Playing with #jetty and the latest #websocket spec. :) #mojolicious" --kraih http://twitter.com/kraih/status/49954151794343937
22:13 daviddelikat joined #mojo
22:16 stephen joined #mojo
22:34 janus openbsd port updated to 1.15: http://marc.info/?l=openbsd-ports-​cvs&m=130074430402998&w=2
22:44 kaare left #mojo
22:55 giwi left #mojo
23:50 damog joined #mojo

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