Camelia, the Perl 6 bug

IRC log for #mojo, 2012-03-27

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

All times shown according to UTC.

Time Nick Message
00:19 zivester joined #mojo
00:32 SmokeMachine joined #mojo
00:47 xaka joined #mojo
01:00 Gedge joined #mojo
01:07 tempire console javascript test framework?
01:13 scott joined #mojo
01:45 GabrielVieira joined #mojo
01:53 inokenty joined #mojo
02:25 Psyche^ joined #mojo
02:33 xaka joined #mojo
02:52 asarch joined #mojo
02:53 asarch I saved the example from the site in the 'mojo.pl' file and I get this when I do 'morbo mojo.pl'
02:53 asarch Can't load application "mojo.pl": Can't locate object method "text" via package "response" (perhaps you forgot to load "response"?) at mojo.pl line 32.
02:54 tempire pastie it
02:55 asarch http://paste.scsys.co.uk/189061
02:55 tempire which line is line 32?
02:56 tempire (gist is recommended)
02:56 tempire guess I should say gist it from now on
02:57 tempire it's probably under Mojolicious::Lite, though
02:57 tempire you're missing a # in front of what's probably a comment
02:59 asarch Thank you
02:59 asarch Maybe it was deleted when I did :31,$s/^#// in vim
02:59 asarch Thank you very much :-)
03:07 SmokeMachine joined #mojo
06:17 Cyrus left #mojo
06:31 taryk joined #mojo
06:35 spleenjack joined #mojo
06:46 lbr joined #mojo
06:50 lbr it seems I need both get '/' and get '/*' to catch all get-urls, is there really no way to catch them with only one get?
06:50 lbr (M::L)
07:07 batman joined #mojo
07:14 Gedge joined #mojo
08:09 judofyr joined #mojo
08:15 dolmen joined #mojo
08:39 tempire joined #mojo
08:39 tempire oh snap
08:39 tempire forgot to release
08:43 AmeliePoulain joined #mojo
08:45 hugh joined #mojo
08:53 tempire I'm wondering if we really need to tweet releases anymore
08:53 tempire Every release, I mean.
08:54 tempire Hmm, I guess the past couple of releases haven't been tweeted anyway.
08:58 red_ joined #mojo
08:59 red_ Anyone having nothing better to do than helping a newbie?
08:59 tempire If it can be done in 3 minutes.
08:59 tempire go
08:59 red_ I'm in the ::Lite tutorial.
08:59 red_ At the beginning...
09:00 red_ here:
09:00 red_ get '/' => sub {
09:00 red_ my $self = shift;
09:00 red_ $self->render(text => 'Hello World!');
09:00 red_ };
09:00 red_ what does the "shift" do?
09:00 tempire ack.  for future reference, don't paste into the channel.  use gist.github.com
09:00 red_ (oops.)
09:00 tempire my $self = shift is equivalent to my ($self) = @_;
09:01 tempire well, sort of equivalent.  shift takes the first position of @_ and puts it into $self
09:01 red_ Ah. Ok.
09:01 tempire You could also do shift->render(text => ...)
09:01 tempire which is much more hip
09:01 tempire hipster perl
09:02 red_ Hm. I am not experienced in oo programming. Should I learn that first and then come back to mojo or do you recommend to learn it via mojo?
09:03 tempire that's a tough call.  you might be able to pick up the basics just by reading the tutorials
09:04 tempire actually, there's no reason you couldn't learn them hand-in-hand
09:05 tempire make sure and check out the learning perl references on mojolicio.us/perldoc
09:05 tempire the learn perl in 2 hours and 30 minutes in particular would be good for you if you've never seen shift before.
09:06 red_ I only know shift like "$x = shift(@y)"
09:06 tempire it's the same thing.  the @_ is just implied.
09:06 tempire shift; == shift(@_)
09:07 red_ OK. I'll look through the pearldoc. Hopefully I am just lacking some words ;)
09:07 red_ Weird things like pointers, subroutines etc waiting for me!
09:09 tempire definitely check out the 2.5 hours link.  and be aware that the modern perl book is a free download.
09:09 tempire Rhaen: where's the training course you were working on?
09:10 red_ ?
09:10 red_ you mean the mojo::lite tut?
09:10 red_ already reading ;) Could you help me focusing on the 2.5 hours by explaining me what the subroutine does?
09:10 tempire red_: http://mojolicio.us/perldoc#LEARNING_PERL
09:11 * tempire is off to sleep
09:11 red_ good nightexit
09:11 red_ exit
09:31 BlueT_ joined #mojo
09:51 daxim joined #mojo
10:02 diphda joined #mojo
10:13 diphda left #mojo
10:58 memowe \o
11:19 ki0 joined #mojo
11:29 abra joined #mojo
11:32 hugh joined #mojo
12:21 * sri yawns
12:24 sri tempire++
12:25 d4rkie joined #mojo
12:28 sri hmm
12:29 sri it looks like tempire forgot to upload the release :S
12:29 * judofyr feels like sri has been yawning all day long
12:30 judofyr or was it yesterday?
12:31 sri https://metacpan.org/recent # hmm, looks like he really did :S
12:34 GabrielVieira joined #mojo
12:39 dolmen joined #mojo
12:46 Ulti joined #mojo
12:47 marcus karma for nothing!
12:47 purl nothing! has neutral karma
12:48 Ulti if I have a route setup   $r->route('/controller/method​')->to('controller#method');   and then I'm doing   http://site.com/controller/method?query='text'  am I meant to be seeing query in the stash?
12:51 GitHub104 joined #mojo
12:51 GitHub104 [mojo] kraih pushed 1 new commit to master: http://git.io/fPnuDA
12:51 GitHub104 [mojo/master] removed X-Forwarded-Host support since it is redundant for well configured reverse proxies - Sebastian Riedel
12:51 GitHub104 left #mojo
12:53 judofyr Ulti: no. use $self->param('query')
12:54 judofyr Ulti: or $self->req->params->param('query')
12:55 Ulti yeah sorry I just found it on the google group
12:55 judofyr :)
12:57 sri oh well... i uploaded 2.69
12:57 judofyr sri: question for the HTTP implementation master: does multipart use Content-Length or just boundaries?
12:58 judofyr so… many… RPCs…
12:58 sri content-length unless it is chunked
12:58 judofyr for multipart too?
12:58 sri yes
12:58 judofyr what's the point of the boundaries then?
12:58 sri you can only leave out content-length if you're using chunked or closing the connection after the response is finished
12:59 judofyr sri: oh no, I mean *inside* each parts of the multipart
12:59 sri no content-length for parts
12:59 judofyr they have Content-Disposition and some other stuff
12:59 judofyr good. then I've understood it correctly
13:00 judofyr and multipart messages can be chunked too, right?
13:00 judofyr chunking is on the HTTP-level, while multipart is more on the HTML/form-data-level?
13:00 judofyr (or, application-specific-level)
13:01 Akron joined #mojo
13:04 GitHub146 joined #mojo
13:04 GitHub146 [mojo] kraih pushed 1 new commit to master: http://git.io/OAnnNA
13:04 GitHub146 [mojo/master] documentation tweaks - Sebastian Riedel
13:04 GitHub146 left #mojo
13:05 sri sounds about right
13:09 GitHub75 joined #mojo
13:09 GitHub75 [mojo] kraih pushed 1 new commit to master: http://git.io/JjsJfw
13:09 GitHub75 [mojo/master] fixed start examples - Sebastian Riedel
13:09 GitHub75 left #mojo
13:10 Akron stephan48: Shortcuts in lite should work like this: (any '/:user/route')->shortcutname('parameter');
13:10 stephan48 ah cool
13:11 * sri thought that was obvious
13:11 stephan48 no
13:11 d4rkie joined #mojo
13:11 sri since you can also just do app->routes->route(...)->to(...)
13:12 stephan48 i would even say that the "any '/'" and so on is quite a bit of magic when you dont know the internals
13:12 stephan48 should probably documented a bit by someone understanding it a bit more :)
13:12 sri http://mojolicio.us/perldoc/Mojolicious/Lite#any
13:12 stephan48 ah then ofcourse its obvious, didnt look at that for some time :)
13:12 judofyr sri: doesn't mention that any() returns a Route object though
13:13 stephan48 my $route = any '/:foo' => sub {...};
13:13 stephan48 it does
13:13 judofyr it does? oh, nvm then
13:13 stephan48 judofyr: i thougth the same
13:13 judofyr oh, I see
13:13 judofyr kinda subtle though
13:13 Akron Yes, it does. But sri gave me the hint with ()->shortcutname() a while ago. Looks pretty enough to me.
13:16 GabrielVieira joined #mojo
13:19 GitHub96 joined #mojo
13:19 GitHub96 [mojo] kraih pushed 1 new commit to master: http://git.io/qU1jkQ
13:19 GitHub96 [mojo/master] link to Mojolicious::Routes::Route from Mojolicious::Lite - Sebastian Riedel
13:19 GitHub96 left #mojo
13:21 judofyr sri: if you want $self->req->params to include multipart content, you could add 'has "filename"', and override "" to return the filename. that's consistent with how HTML forms handles file uploads without setting the correct enctype…
13:21 judofyr (so $self->req->params could include Mojo::Content-objects)
13:27 sri the form value/file upload separation was inherited from catalyst, which inherited it from mod_perl
13:31 judofyr sri: how does Mojo handle long form values btw? does it read everything into memory (e.g. is it only uploads that get backed to disk?)
13:39 GabrielVieira joined #mojo
13:55 stephan48 Akron: i rewrote my plugin with shortcut now you will find it later today(will send you updated link to my repo)
13:55 Akron stephan48: Great!
13:55 sri judofyr: no, if an asset gets too big it is written to disk
13:56 sri but $req->param() can read it back into memory, no matter how big it is
13:56 judofyr ah
13:56 sri that's what makes allowing big uploads a little risky if you don't check requests
13:57 mattastrophe joined #mojo
13:57 sri mindless $self->param('foo') can easily be exploited if MAX_MESSAGE_SIZE gets set to a very large value
13:58 judofyr shouldn't MAX_MESSAGE_SIZE be set per-request?
13:58 sri for some time i prevented all assets that have been written to disk from being parsed, but that caused too much trouble
13:58 sri judofyr: it can be
13:58 sri you can tweak almost all limits per request if you wanted to
14:14 sri http://www.amazon.co.jp/dp/4774150258 # <- contains a chapter about mojolicious
14:18 judofyr nice
14:24 Netfeed why does each in M::Collection return $self instead of the result from the sub you send it?
14:30 judofyr Netfeed: use ->map for that
14:30 judofyr Netfeed: and the answer to your "why": because we have map :)
14:30 Netfeed oh, yeah, that makes a lot more sense
14:33 Netfeed thanks
14:35 black joined #mojo
15:02 kaare joined #mojo
15:21 Akron joined #mojo
15:23 GitHub192 joined #mojo
15:23 GitHub192 [mojo] kraih pushed 1 new commit to master: http://git.io/_cJuaQ
15:23 GitHub192 [mojo/master] added a few route examples - Sebastian Riedel
15:23 GitHub192 left #mojo
15:29 inokenty joined #mojo
15:31 xaka joined #mojo
15:32 GitHub70 joined #mojo
15:32 GitHub70 [mojo] kraih pushed 1 new commit to master: http://git.io/CJLDBg
15:32 GitHub70 [mojo/master] more realistic route examples - Sebastian Riedel
15:32 GitHub70 left #mojo
15:47 Psyche^ joined #mojo
15:47 abra joined #mojo
15:58 asarch joined #mojo
16:04 taryk joined #mojo
16:35 sri http://blog.dotcloud.com/dotcloud-a​nnounces-native-support-for-websock # for those disappointed with heroku not supporting websockets
16:43 sri it actually seems rather simple
16:54 inokenty Why they don't include a Perl example of using websocket?
17:00 sri i don't think they employ any perl hackers anymore
17:00 sri miyagawa switched to another company
17:29 inokenty joined #mojo
17:29 inokenty http://browserquest.mozilla.org/ - cool
17:32 sri :D
17:39 tempire joined #mojo
17:39 tempire wat
17:39 tempire I did upload
17:40 tempire oh, apparently there was a pause error after I went to bed
17:40 tempire miyagawa moved away from dotcloud?
17:40 tempire wonder what they did wrong.
17:48 phluks joined #mojo
18:05 memowe sri: are you here?
18:05 purl Yes, I am, memowe.
18:05 memowe Or any other core dev?
18:05 memowe I think I found a small bug in the inflate command.
18:05 memowe And I got a fix, but there's no test file for inflate AFAICS.
18:06 memowe What now?
18:06 purl it has been said that now is the time for the Kwisatz Haderach to arise and conquer America on a sandworm army
18:09 memowe https://github.com/kraih/mojo/pull/308
18:10 sri memowe: commands don't have tests
18:10 memowe sri: Oh, I think it's ok then.
18:10 GitHub14 joined #mojo
18:10 GitHub14 [mojo] kraih pushed 2 new commits to master: http://git.io/8l1NPA
18:10 GitHub14 [mojo/master] fixed a minor inflate command bug with classes without __DATA__ section - Mirko Westermeier
18:10 GitHub14 [mojo/master] Merge pull request #308 from memowe/master - Sebastian Riedel
18:10 GitHub14 left #mojo
18:10 memowe Cool. :)
18:10 memowe Thanks!
18:10 memowe (Contenticious should be able to inflate again)
18:10 memowe The next release will be 2.7?
18:13 GitHub31 joined #mojo
18:13 GitHub31 [mojo] kraih pushed 1 new commit to master: http://git.io/hnpYnw
18:13 GitHub31 [mojo/master] updated Changes - Sebastian Riedel
18:13 GitHub31 left #mojo
18:13 sri 2.70
18:15 sri the bug also showed with "mojo inflate"
18:15 memowe oh, ok
18:19 sri memowe++
18:21 memowe :)
19:24 GitHub177 joined #mojo
19:24 GitHub177 [mojo] kraih pushed 1 new commit to master: http://git.io/HeGRsQ
19:24 GitHub177 [mojo/master] improved speed of version command by switching to the MetaCPAN API - Sebastian Riedel
19:24 GitHub177 left #mojo
19:25 sri it's *a lot* faster than search.cpan in my naive tests
19:25 sri please try :)
19:48 taryk joined #mojo
20:02 perlite_ joined #mojo
20:12 Shaeto joined #mojo
20:31 Shaeto left #mojo
20:41 lukep joined #mojo
20:52 tabbi joined #mojo
20:52 tabbi left #mojo
21:05 batman joined #mojo
21:15 batman sri: https://twitter.com/#!/dot_cl​oud/status/184685912628805632
21:15 batman so i guess your tweet is correct
21:15 batman anyone tested dotcloud and mojo+ws?
21:16 tempire mojo version speed is about the same for me
21:18 batman was that to me tempire?
21:18 tempire no, sri
21:18 batman good :)
21:25 sri tempire: interesting, it's instant here now, used to be 1-2 seconds
21:25 tempire it's around 600 ms here
21:26 tempire 600-700
21:26 purl -100
21:26 tempire 550-355
21:26 purl 195
21:26 tempire go purl!
21:26 purl You go, tempire!
21:26 sri 0.26s here
21:54 preflex_ joined #mojo
22:28 jwang joined #mojo
23:24 dwierenga joined #mojo

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