Camelia, the Perl 6 bug

IRC log for #mojo, 2013-05-15

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

All times shown according to UTC.

Time Nick Message
00:59 abqar joined #mojo
01:15 xaka joined #mojo
01:33 good_news_everyone joined #mojo
01:33 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/EAh7lw
01:33 good_news_everyone mojo/master c7f3e91 Sebastian Riedel: fixed a few typos
01:33 good_news_everyone left #mojo
01:33 moltar joined #mojo
01:33 * sri starts preparing a release announcement
01:33 sri STOP FINDING BUGS NOW!
01:39 d4rkie joined #mojo
01:47 D4RK-PH0ENiX joined #mojo
01:57 Meiermann joined #mojo
02:24 * jberger gets out his bug detector
02:25 egopro joined #mojo
02:28 * jberger is reading the list of 3.0 features
02:50 inokenty joined #mojo
02:52 * tempire has been prepared for days
02:55 * jberger hopes that sri uses tempire's picture for 🎩
03:27 preflex_ joined #mojo
03:29 rihegher joined #mojo
03:48 fildon_ joined #mojo
03:50 wruppert joined #mojo
04:15 Adura joined #mojo
04:47 d4rkie joined #mojo
05:09 ynonp joined #mojo
05:10 duncanthrax joined #mojo
05:19 hotsnow joined #mojo
05:19 * hotsnow $dat = { a => 1 };
05:19 * hotsnow return $c->render(json => $dat);
05:19 * hotsnow $dat = { a => 1 };
05:19 * hotsnow $json = $json->encode($dat);
05:19 * hotsnow return $c->render(text => $json);
05:20 hotsnow what's the difference of these two reutn json  strings ?
05:25 preaction i would expect the second to use a content-type of text/plain, instead of the correct content-type of application/json
05:34 hotsnow thanks
05:51 ver joined #mojo
06:01 dod joined #mojo
06:05 rihegher left #mojo
06:07 basiliscos joined #mojo
06:13 dod joined #mojo
06:24 rem_lex|pivo joined #mojo
06:36 dod joined #mojo
06:36 Mike-PerlRecruiter_ joined #mojo
06:41 rihegher joined #mojo
06:42 rihegher left #mojo
07:06 mattastrophe joined #mojo
07:09 suy joined #mojo
07:33 mire_ joined #mojo
07:37 jzawodn joined #mojo
07:58 egopro joined #mojo
07:59 egopro_ joined #mojo
08:06 Vandal joined #mojo
08:06 mugenken joined #mojo
08:21 dpetrov_ joined #mojo
08:50 alnewkirk joined #mojo
08:50 cosmincx joined #mojo
09:03 Dandre joined #mojo
09:12 nelio joined #mojo
09:25 Dandre hello
09:26 Dandre I have a couple of questions about mojo
09:27 Dandre is there any tutorial to help implementing some "discovering api" feature?
09:28 Dandre second how can I have some architecture like this:
09:28 Dandre http://foo.com/plug1 handled by plugin (or app) 1
09:28 Dandre http://foo.com/plug2 handled by plugin (or app) 2
09:37 yakudza joined #mojo
09:56 fhelmber_ joined #mojo
10:00 xaka joined #mojo
10:25 moltar joined #mojo
10:42 maxhq joined #mojo
10:48 hotsnow joined #mojo
10:49 maxhq joined #mojo
10:56 hrupp joined #mojo
11:28 lukep joined #mojo
11:44 mire_ joined #mojo
12:09 andrefs joined #mojo
12:41 inokenty left #mojo
12:42 d4rkie joined #mojo
13:00 sri \o\
13:00 sri /o/
13:01 sri no new bugs so far
13:03 good_news_everyone joined #mojo
13:03 good_news_everyone [mojo] kraih deleted portable_commands at 6ecadc2: http://git.io/FvlIBA
13:03 good_news_everyone left #mojo
13:03 dod joined #mojo
13:03 stephan48 Dandre: abou the second q, i think you migh want to have a look at shortcuts
13:03 good_news_everyone joined #mojo
13:03 good_news_everyone [mojo] kraih deleted timers_again at ff28702: http://git.io/wqZ48Q
13:03 good_news_everyone left #mojo
13:05 jberger_ Dandre, or check out the Mount core plugin
13:05 jberger_ https://metacpan.org/module​/Mojolicious::Plugin::Mount
13:06 Dandre stephan48: ok but what I want is that the plug1 andp plug2 applications are independants are can be distributed with 2 different archive files.
13:08 Dandre jberger_ so if I understant correctly, two different applications can mount themselve to a different prefix
13:09 jberger_ thats the idea
13:09 stephan48 Dandre: You would create 2 different plugin distributions then each containing one or more plugin module
13:10 Dandre ok
13:15 sri gonna give it a few more hours, if nothing happens we can release 4.0 :)
13:16 dod joined #mojo
13:32 cosmincx1 joined #mojo
13:35 jberger sri: last minute idea, can totally be ignored
13:35 jberger I just remembered many months ago I asked if the memorize helper could be tweaked to allow for expiring a memorization
13:36 jberger I ended up implementing it for galileo: https://github.com/jberger/Galileo/blob/m​aster/lib/Galileo/Plugin/FlexMemorize.pm
13:36 jberger the only difference in the helper is that with no args it returns the hashref to the cache
13:37 jberger s/to/of/
13:37 jberger I'm not asking to add the expires helper, but any chance that the official memorize could take on that little tweak?
13:38 jberger again just an idea
13:41 sri jberger: what did i say last time?
13:41 denisboyun joined #mojo
13:41 jberger I think it was that you didn't like the idea of returning the hash
13:41 jberger I can try to search the log
13:42 sri would make sense, since i hate the idea now too, the hash is not a clean data structure like the stash and we would expose an implementation detail
13:44 sri on the other hand, maybe the whole helper should be removed from core
13:45 sri the test is a bit fragile and has actually been disabled by default
13:46 ppp joined #mojo
13:47 jberger shoot! now I remember you suggested that too!
13:47 jberger btw, I don't think the log goes that far back
13:48 jberger it came up when I was working on galileo last summer
13:48 * sri would only hate having to remove the fun example from the guides :)
13:48 jberger but the first hit on my nick was october
13:48 sri http://mojolicio.us/perldoc/Mojolicious/Gu​ides/Rendering#Memorizing_template_blocks
13:49 btyler joined #mojo
13:49 jberger sri: its a good example of code too
13:49 sri it might be better off in an external plugin though
13:49 sri evolve in new directions
13:50 sri and disabled tests is not nice at all
13:52 * jberger may have found the original discussion
13:53 sri not a unique example either, content_for is the better one in that category
13:54 good_news_everyone joined #mojo
13:54 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/xYpLbA
13:54 good_news_everyone mojo/master 0289626 Sebastian Riedel: removed memorize helper
13:54 good_news_everyone left #mojo
13:54 sri bye bye memorize
13:55 jberger yikes, sri with the machete
13:55 jberger guess that means its time for a CPAN version
13:55 sri i didn't mind the helper itself, but the fragile test annoyed me
13:56 sri was our only test that had to be disabled simply because it was fragile
13:57 sri if there's a rock solid and elegant reimplementation available, i wouldn't vote against reintroducing a similar helper in the future
13:58 jberger why not just store the time, render with that time, then render again and make sure the time is the same
13:58 jberger I guess it wouldn't change nevermind
13:58 * jberger didn't think it though
14:00 jberger http://irclog.perlgeek.de/​mojo/2012-07-27#i_5851882
14:00 jberger I was using my whole name as the nick
14:00 * jberger was so young then
14:01 jberger its funny to read now, especially knowing how Mojolicious development goes, to see me suggest adding a DefaultHelper
14:01 jberger ha!
14:02 sri haha
14:03 sri ANY OTHER LAST MINUTE REMOVALS?!
14:06 nicomen hm, is there an elegant way to use url_with but skip one or more query params?
14:07 mattastrophe joined #mojo
14:11 dod1 joined #mojo
14:13 Dandre can you tell me what is a stash?
14:15 nicomen Dandre: a place you can store things that will last through out the request
14:16 Dandre is it specific to mojolicious?
14:16 nicomen no
14:16 nicomen it's a common term
14:17 Dandre I haven't found on perl.org
14:17 btyler joined #mojo
14:17 nicomen Dandre: it's just a name, just like model or controller or view or session
14:18 nicomen in practice a variable accessible throughout the request, the next request that is processed will not have access to the same values, (the stash is empty)
14:18 Dandre ok so basically it is a hash or a more clever object?
14:20 gryphon joined #mojo
14:24 nicomen Dandre: it can be whatever you want it to be, in mojolicious' case it is a hashref, but you can store objects inside the stash if you want: $self->stash('my_object' => $some_object);  and later get it with $self->stash('my_object');
14:33 sri how do we feel about Test::Mojo::json_content_is? redundant with Test::Mojo::json::is('/'...)?
14:39 mire_ joined #mojo
14:39 sh4 joined #mojo
14:41 jberger sri: json pointers will always be strings right?
14:42 jberger could json_is (and json_message_is for that matter) assume '' if the first arg is a reference?
14:42 jberger kinda merges json_content_is and json_is that way
14:58 maxhq joined #mojo
15:06 xaka joined #mojo
15:09 * jberger hears crickets
15:10 crab i don't like it
15:11 jberger I'm not sure its my favorite either, however if someone has hundreds of json_content_is tests, having to add the effectively null pointer to all of them seems a little silly doesn't it?
15:25 mattastrophe joined #mojo
15:26 btyler joined #mojo
15:33 Akron joined #mojo
15:41 bluescreen joined #mojo
15:44 sri jberger: i don't like it
15:46 Akron I just went online to say: I like it. ;)
15:46 Akron Too late I guess.
15:47 sri what i dislike most is that it makes the description of the method complicated
15:48 jberger "test a retrieved json structure, with an option leading json pointer"
15:48 jberger yeah, its awkward
15:49 sri documentation driven design :)
15:49 jberger documentation is important
15:49 jberger still I think the most common use case will be to test the entire returned structure
15:50 jberger so the most common use will be case that is the least good looking
15:51 jberger json_is( '/some/deep/key/2' => { foo => bar } ) looks really slick
15:51 jberger json_is( '' => { foo => [qw/ bar baz/] )
15:52 jberger s!)!})!
15:53 jberger I guess I liked the root case better when we thought the root was '/'
15:54 jberger "omitting the pointer assume '', the document root"
15:54 jberger assumes
15:56 Akron I agree with jberger. Saying, a prepended jsonpointer is optional seems to be fine in my mind.
16:05 * sri changed his opinion
16:06 sri and json_message_is can get the same treatment
16:07 jberger \m/
16:10 * jberger would submit a pull-req, but I have a feeling sri would change every part of my implementation on this one :-)
16:12 * sri already has written everything
16:14 SmokeMachine joined #mojo
16:15 * jberger assumed as much, but likes to offer when its a functionality he's fighting for
16:15 sri you could have proposed it with a patch ;p
16:16 sri it's like a 4 line change with everything :)
16:16 sri i'm actually struggling with expaining it in the changelog xD
16:16 jberger haha
16:17 jberger "merged json_content_is into json_is"
16:17 jberger unified
16:17 denisboyun_ joined #mojo
16:19 good_news_everyone joined #mojo
16:19 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/FCkeHQ
16:19 good_news_everyone mojo/master ef7170b Sebastian Riedel: removed json_content_is method from Test::Mojo
16:19 good_news_everyone left #mojo
16:20 sri sweet sweet consistency
16:20 marcus sri: https://www.catincan.com/ - seek funding for new features?
16:21 sri hard to define features in advance
16:22 sri at least for mojolicious, it's a lot easier with mango
16:24 sri you can't really fund ongoing maintenance :/
16:26 jberger sri: so in json_is, $p might be the data structure?
16:26 sri oops, i guess i messed up the description :)
16:27 jberger yet $self->tx->res->json($p)
16:27 jberger works?
16:27 jberger no I'm looking at the implementation
16:27 sri hmm, i might have messed up more and it doesn't appear to matter
16:27 jberger https://github.com/kraih/mojo/commit/ef717​0b2559b5fdcb7b65eaf5c6d9d4c92450e43#L1R163
16:28 sri yes, i screwed up
16:30 good_news_everyone joined #mojo
16:30 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/efF08Q
16:30 good_news_everyone mojo/master 47cd205 Sebastian Riedel: fixed JSON tests
16:30 good_news_everyone left #mojo
16:31 xaka joined #mojo
16:31 sri i suppose the pointer worked because it didn't start with a slash or so :)
16:32 jberger was about to paste mine: http://pastie.org/7912146
16:32 jberger you got it though, and essentially the same as mine
16:32 jberger yours is one less line
16:33 jberger sri++
16:34 jberger ok, time go work out
16:34 jberger see yall in abit
16:34 sri ANY OTHER LAST MINUTE REMOVALS?! :)
16:37 basiliscos joined #mojo
16:37 mattastrophe joined #mojo
16:37 labrown joined #mojo
16:43 marcus alas poor Memoize, I knew him, Horatio
16:46 sri his name was memorize
16:49 moltar joined #mojo
16:50 hotsnow joined #mojo
16:59 chansen sri: did you fix the JSON serialization of Infinite and NaN's?
16:59 sri nope
17:01 alester joined #mojo
17:01 chansen why not?
17:01 sri not sure it's worth fixing
17:02 chansen Emitting ill-formed JSON isn't nice!
17:03 sri all implementation do it, nobody seems to care
17:03 sri *s
17:04 chansen so, let's Mojo::JSON superior! the fix is simple
17:04 chansen if B::svref_2object(\$value)->FLAGS & (B::SVp_IOK | B::SVp_NOK) && $value * 0 == 0;
17:05 chansen oops
17:05 chansen never mind, its correct
17:05 * marcus is confuzzled now
17:08 sri chansen: heh, that's clever
17:08 chansen =)
17:10 chansen marcus: $ perl -MMojo::JSON -le 'print Mojo::JSON->encode([9**9**9, -9**9**9, -sin(9**9**9)])'
17:16 good_news_everyone joined #mojo
17:16 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/hbLDfg
17:16 good_news_everyone mojo/master 7169454 Sebastian Riedel: fixed Mojo::JSON to encode "inf" and "nan" values as strings
17:16 good_news_everyone left #mojo
17:16 mire_ joined #mojo
17:18 chansen sri: no, no, don't rely on the actual stringification of Infinite or NaN, it can vary depending on compiler/system
17:19 sri :/
17:19 chansen only verify that the value is a serialized string
17:19 chansen http://en.wikipedia.org/wiki/NaN#Display
17:20 * chansen unless Perl makes is consistent
17:20 * chansen checks source
17:25 sri knowing perl... it prolly doesn't
17:25 sri seems fine on linux and os x at least though https://travis-ci.org/kraih/mojo
17:27 sri can anyone run perl -E 'say 9**9**9' on a windows perl?
17:27 chansen Perl relies on sprintf in stdio.h
17:27 chansen so, it's platform dependent
17:27 sri :S
17:28 good_news_everyone joined #mojo
17:28 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/o62FZA
17:28 good_news_everyone mojo/master caa1af3 Sebastian Riedel: made "inf" and "nan" tests a little more portable
17:28 good_news_everyone left #mojo
17:29 chansen But provide consistent behavior for the reverse, string to NV
17:29 denisboyun joined #mojo
17:30 chansen sri++
17:30 sri chansen++
17:36 good_news_everyone joined #mojo
17:36 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/I1WqBg
17:36 good_news_everyone mojo/master b95c9a2 Sebastian Riedel: fixed typo in test
17:36 good_news_everyone left #mojo
17:37 sri so many last minute changes, guess 4.0 will wait till tomorrow then :)
17:37 SmokeMachine joined #mojo
17:39 nelio joined #mojo
17:42 r0b3rt D:\>perl -E "say 9**9**9"    output = 1.#INF
17:43 r0b3rt strawberry perl 5.16
17:48 sri eeep, thanks
18:28 bowtie joined #mojo
18:35 Mike-PerlRecruiter_ joined #mojo
18:39 sri or i'll just get impatient and release in an hour or two ;p
18:40 jberger :-)
19:05 mire_ joined #mojo
19:25 salparadise joined #mojo
19:28 gryphon joined #mojo
19:40 dpetrov_ joined #mojo
19:40 * sri goes over the feature list one last time
19:42 sri oh, almost forgot we introduced the multi name form for $self->param
19:46 sri like $self->param(['foo', 'bar'])
19:46 sri http://mojolicio.us/perldoc/​Mojolicious/Controller#param
19:48 basiliscos sri:  cool!
19:50 mattastrophe joined #mojo
19:51 sri haha, even ->pluck is a post-3.0 feature... i have no clue how to mention that in the feature list though :)
19:53 good_news_everyone joined #mojo
19:53 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/iShsmg
19:53 good_news_everyone mojo/master ad31e23 Sebastian Riedel: fixed typo
19:53 good_news_everyone left #mojo
20:01 good_news_everyone joined #mojo
20:01 good_news_everyone [mojo] kraih tagged v4.0 at e455afa: http://git.io/qvcGTg
20:01 good_news_everyone left #mojo
20:01 sri \o/
20:01 njlg sri++
20:02 doby grats
20:08 mattastrophe joined #mojo
20:09 sri i messed up a little -.-
20:10 sri there are two .DS_Store files in the tarball... but i think they are harmless
20:13 jon joined #mojo
20:14 * marcus pops the champagne.
20:14 * marcus spills it on the .DS_Store files
20:15 yko hehe. congrats
20:16 yko mojolicious 4.0, perl 5.18.0 RC3... sweet month
20:28 mattastrophe joined #mojo
20:31 sri http://blog.kraih.com/post/50517069291/mojolic​ious-4-0-released-perl-real-time-web-framework
20:33 sri poke me if i forgot something
20:40 fhelmber_ joined #mojo
20:44 marty \o/  Congratulations on 4.0 to all the developers.
20:44 marty I'll download and test asap
21:08 suy joined #mojo
21:08 moltar joined #mojo
21:13 marty very smooth upgrade.  I'm amazed how many render_json() I still had around.  :)    Great release!
21:17 jberger sri+=2
21:18 * jberger grabs
21:18 * marcus is flying to dusseldorf for the weekend tomorrow.
21:19 marcus düsseldorf even.
21:20 dabudabu Grats
21:21 dabudabu I wanna go to Düsseldorf sometime
21:22 dabudabu I think I'll upgrade to 4.0 after I've submitted my bachelor's thesis next week, which is a web service written in Mojolicious ;p
21:32 lukep joined #mojo
22:11 salparadise joined #mojo
22:15 davido joined #mojo
22:22 salparadise joined #mojo
22:32 good_news_everyone joined #mojo
22:32 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/k3Ap6g
22:32 good_news_everyone mojo/master b117009 Sebastian Riedel: mention that not just GET/POST parameters can have multiple values
22:32 good_news_everyone left #mojo
22:34 sri the show must go on
22:47 jberger this is kinda fun: http://www.activestate.com/blog/2013/05/sm​all-simple-app?utm_source=feedburner&u​tm_medium=feed&utm_campaign=Feed%3A+ac​tivestate%2Fblog+%28ActiveState+Blog%29
23:00 * marty loves the new morbo reloading. o/
23:06 tempire new?
23:06 tempire whoa!
23:06 tempire release!
23:06 * tempire dances
23:29 sri \o\
23:29 sri /o/
23:38 d4rkie joined #mojo
23:52 suicidejack joined #mojo
23:55 Akron joined #mojo

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