Camelia, the Perl 6 bug

IRC log for #mojo, 2013-09-06

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

All times shown according to UTC.

Time Nick Message
00:08 russum joined #mojo
00:57 jberger_ joined #mojo
01:03 laouji joined #mojo
01:03 whatitdo joined #mojo
01:09 btyler joined #mojo
01:22 basic6_ joined #mojo
01:27 btyler joined #mojo
01:45 jberger joined #mojo
01:57 Meiermann joined #mojo
01:58 d4rkie joined #mojo
02:17 russum joined #mojo
02:17 russum left #mojo
02:43 jberger_ joined #mojo
02:52 jberger joined #mojo
03:27 preflex_ joined #mojo
03:58 bzero joined #mojo
04:00 dvinciguerra joined #mojo
04:01 yko joined #mojo
04:06 btyler joined #mojo
04:18 KindOne joined #mojo
04:35 dqw joined #mojo
04:36 KindOne joined #mojo
05:10 pingvin joined #mojo
05:16 hummeleBop joined #mojo
05:34 arpadszasz joined #mojo
05:48 dod joined #mojo
05:53 dod joined #mojo
06:08 dotandimet joined #mojo
06:14 ab joined #mojo
06:19 Vandal joined #mojo
06:41 andrei_ joined #mojo
06:41 andrei_ hey guyz, how can I start my mojo app to run in background?
06:50 dod joined #mojo
06:53 denis_boyun joined #mojo
06:58 dotandimet joined #mojo
07:01 denis_boyun__ joined #mojo
07:03 Vandal joined #mojo
07:10 basiliscos joined #mojo
07:16 trone joined #mojo
07:25 gatitskiy joined #mojo
07:27 nicolaas hello ... I noticed that, when some of my ajax requests are too long, mojolicious re-executes it several times
07:27 nicolaas how can I prevent that ?
07:29 yakudza joined #mojo
07:40 batman nicolaas: sure there's not a new request from your javascript+
07:40 batman ?
07:40 gatitski_ joined #mojo
07:53 Adura joined #mojo
07:54 batman what do i do, when i have no idea how to fix a bug?
07:54 batman http://www.cpantesters.org/cpan/report​/337b0f5c-13d9-11e3-b8b2-96b61dda783b
07:54 batman should i simply mark the tests as TODO on Dragonfly BSD ?
07:57 mire joined #mojo
07:59 nicolaas batman yes, I made several tests
07:59 nicolaas I found a workaround, using the Mojo::IOLoop timeout param
07:59 nicolaas but I still think it's kind of a bug
08:00 nicolaas (or a feature ?)
08:02 batman what do you mean? tests for my module..?
08:02 fhelmber_ joined #mojo
08:07 nicolaas well, I don't know ..
08:09 omega batman: "can't locate strict.pm" sounds like a hosed smoker?
08:10 omega batman: try talking to bingos?
08:11 batman omega: oh! guess i'm more tired than i thought :(
08:11 batman BinGOs: comments?
08:29 inokenty joined #mojo
08:45 mtths joined #mojo
08:46 BinGOs Why is the assumption always that it is a hosed smoker?
08:46 batman i have no idea. that's why i'm asking for comments :)
08:51 BinGOs -bash-4.2$ /usr/home/cpan/pit/bare/perl-5.18.0/bin/perl -Mstrict -e1
08:51 BinGOs -bash-4.2$
08:53 BinGOs and having #!/usr/bin/env perl in postman will not use that ^^^^ perl (the one that is running the tests) it uses the system perl
08:53 BinGOs and the system perl reports
08:53 BinGOs -bash-4.2$ perl -Mstrict -e1
08:53 BinGOs Can't locate strict.pm in @INC (you may need to install the strict module) (@INC contains: /opt/perl-5.18.0/lib/site_p​erl/5.18.0/x86_64-dragonfly /opt/perl-5.18.0/lib/site_perl/5.18.0 /opt/perl-5.18.0/lib/5.18.0/x86_64-dragonfly /opt/perl-5.18.0/lib/5.18.0 .).
08:53 BinGOs BEGIN failed--compilation aborted.
08:55 batman BinGOs: so i need /usr/bin/perl ?
08:56 BinGOs that was /usr/bin/perl
08:56 BinGOs I have work to do now
08:58 BinGOs I have a Microsoft Direct Access server to configure yet again.
08:58 BinGOs /usr/bin/perl was linked to /opt/perl-5.18.0/bin/perl which I had deleted.
08:58 BinGOs I have now removed /usr/bin/perl
08:58 * BinGOs &
08:59 batman let me rephrase my question: do i need to change my hashbang to make the test successful?
09:03 maxhq joined #mojo
09:04 marcus batman: I think bingos is saying it was a hosed smoker after all =]
09:04 batman sweet!
09:04 marcus "/usr/bin/perl was linked to /opt/perl-5.18.0/bin/perl which I had deleted."
09:13 mib_0uymk8 joined #mojo
09:14 BinGOs No I was saying the tests are being run by /usr/home/cpan/pit/bare/perl-5.18.0/bin/perl and you appear to be using a different perl to run your tests
09:14 BinGOs spot the disconnect.
09:14 BinGOs the fact that the system perl was hosed is immaterial
09:15 BinGOs I have removed /usr/bin/perl
09:15 BinGOs so env perl will now point to /usr/pkg/bin/perl instead
09:16 BinGOs which is still ne /usr/home/cpan/pit/bare/perl-5.18.0/bin/perl
09:16 batman so i need to change the hashbang to..?
09:16 mib_0uymk8 hi
09:16 * BinGOs &
09:16 mib_0uymk8 i just came up with one question: what are the benefits of mojo comparing to dancer?
09:17 batman mib_0uymk8: dancer is crap. you should at least consider dancer2.
09:17 nicolaas also, when you use mojolicious, you get free popcorn
09:18 batman mib_0uymk8: mojolicious is easier to move around, since it has fewer deps
09:19 BinGOs batman: if you want to run your tests with a different perl to the one that is testing your distribution then the shebang is fine I guess.
09:19 batman hm... ok.
09:20 batman mib_0uymk8: i just like the mojo syntax more. and i'm a huge fan of Mojo::IOLoop
09:21 mib_0uymk8 mh
09:21 batman mib_0uymk8: i also don't like template toolkit
09:28 mib_0uymk8 mh interesting
09:49 batman mib_0uymk8: what do you think?
09:50 moltar joined #mojo
09:53 BinGOs batman: okay, your cgi-bin scripts are just simple little things.
09:56 basiliscos joined #mojo
10:12 d4rkie joined #mojo
10:19 ladnaV joined #mojo
10:19 bowtie_ joined #mojo
10:20 bowtie_ joined #mojo
10:23 gatitskiy joined #mojo
10:25 denisboyun joined #mojo
10:26 maxhq1 joined #mojo
10:31 KindOne joined #mojo
10:37 jberger joined #mojo
10:40 jberger mib: mojo is non blocking (think node.js) and has lots of useful functionality built in
10:42 jberger Useragent, dom and json parsers, testing library, template engine
10:42 jberger All designed to work together
10:56 jberger_ joined #mojo
10:58 KindTwo joined #mojo
11:04 wsri joined #mojo
11:04 marcus joined #mojo
11:04 batman joined #mojo
11:13 dqw joined #mojo
11:41 Vandal joined #mojo
12:09 punter joined #mojo
12:20 BinGOs batman: I've linked /usr/bin/perl to /usr/pkg/bin/perl on that dflybsd box now.
12:35 batman thanks :)
12:39 asarch joined #mojo
13:01 btyler joined #mojo
13:11 Dandre Hello,
13:12 Dandre I am trying to use authentication using ssl certificat to authentify the client. I have found this pluging: https://github.com/tempire/​mojolicious-plugin-sslauth But I don't really know how to use it
13:14 Dandre if I open a connection with firefox, I have no way to select the proper certificat and if I dump the certificate from mojo I get this:
13:14 Dandre "Subject Name: undefined\nIssuer  Name: undefined\n"
13:14 Dandre can anyone help me?
13:14 Dandre tempire: ?
13:39 dotandimet joined #mojo
13:41 btyler joined #mojo
13:43 btyler batman, marcus: to clarify, you guys just don't like the name 'subscription_type', or you'd rather the attribute didn't exist and psub/sub were split more deeply?
13:43 batman btyler: don't like the name
13:44 batman that's it for me at least...
13:46 btyler alrighty, cool, that's an easy one :)
13:48 marcus yes
13:48 marcus :)
13:51 drzz joined #mojo
13:52 drzz left #mojo
14:02 sri btw. any thoughts on making delays capture exceptions?
14:04 gryphon joined #mojo
14:05 sri http://pastie.org/8303239 # stuff like this
14:09 mire joined #mojo
14:15 Vandal joined #mojo
14:22 ryozi joined #mojo
14:22 marcus sri: I for one support it.
14:23 sri marcus: the big question is what exactly should happen with the still active events
14:23 sri you may notice the example is flawed
14:24 marcus sri: oh? I think it should not execute the timer.
14:24 marcus or am I missing something?
14:24 sri it's impossible to stop the timer
14:25 marcus So it will be reachable?
14:25 sri for all we know that might be an anyevent timer we have absolutely no control over
14:25 marcus sri: could we wrap it in a sub that just returns if the delay has an error?
14:25 sh4 joined #mojo
14:25 sri we could do something like that
14:25 marcus I mean, we already wrap it, right?
14:26 sri no need to wrap, just some extra logic
14:26 marcus mm
14:27 Dandre is there any documentation to do client ssl authentification with mojolicious? or must I j=use apache? sorry for this question but I don't know how to do
14:27 marcus Dandre: your question doesn't make sense to me. client or server ssl?
14:27 marcus apache is not a ssl client...
14:28 Dandre I am runing mojolicious server and I need to authenticate client connecting to if using ssl sertificates
14:29 Dandre must I apache as replacement to mojo builtin server?
14:30 Dandre basically I have a route definition tha I must protect using ssl client authentification
14:32 Dandre marcus: I have found this plugin https://github.com/tempire/​mojolicious-plugin-sslauth that I can use to validate the certificate but I don't know how to request it from the client
14:34 marcus Dandre: I guess tempire is the man you want to talk to, he implemented this in mojo as well iirc.
14:35 Dandre ok I will wait for him
14:53 basiliscos joined #mojo
14:57 btyler joined #mojo
15:13 denis_boyun joined #mojo
15:13 btyler joined #mojo
15:18 whatitdo joined #mojo
15:27 Tiger joined #mojo
15:29 dqw joined #mojo
15:30 batman btyler: when will there be a new commit? :)
15:31 btyler in about 1.5 hours, when lunch starts
15:31 batman hehe, ok :)
15:31 * batman will have some dinner
15:31 btyler the problem sri pointed out with the documentation applies to the existing subscribe doc too, I think
15:31 btyler or perhaps just a confusing use of 'sub'
15:32 batman i will probably make a new release later today or tomorrow then...  (localtime: 5:30pm)
15:32 btyler awesome
15:37 batman $sub = object, not a sub... i understand it's probably quite confusing :)
15:37 batman i can't fix it, since i don't know what to write
15:38 btyler yeah, ok, then it is the same problem. I'll fix the other spot too
15:39 batman sweet
15:53 gatitskiy joined #mojo
15:54 ladnaV joined #mojo
15:55 * sri just uses $cb for all closures to avoid confusion
16:05 btyler sri: thanks for the doc fix, I learned a bit more about EventEmitter and unsubscribing a specific coderef
16:06 Vandal joined #mojo
16:07 sri btw. if you're thinking "i could just keep a reference myself, no need for a return value", talke a look at ->once :)
16:07 sri s/l//
16:08 sri this way it's all consistent
16:11 hummeleBop joined #mojo
16:21 denisboyun joined #mojo
16:21 sri marcus: what should happen here exactly? http://pastie.org/8303575
16:21 sri i think that test case covers all the edge cases
16:23 sri it's way more tricky than it might look at first, since ->wait is tied to the finish event
16:24 Dandre joined #mojo
16:24 sri tempire, jberger: you might be interested in this too
16:24 sri perhaps batman too
16:24 batman the error handling?
16:24 sri yes
16:25 sri and what happens to the finish event and ->wait
16:25 dod joined #mojo
16:25 batman sri: i don't see how finish can be triggered when error is triggered
16:25 batman i mean... it's not really finished
16:26 sri so, you would want them to be alternatives
16:26 batman yes. either you finish or you fail :)
16:26 sri and what should ->wait do?
16:27 batman i need to ask something before answering that...
16:28 sri i suppose ->wait could subscribe to the error event as well if they are alternatives https://github.com/kraih/mojo/blob/​master/lib/Mojo/IOLoop/Delay.pm#L25
16:29 batman i wrote a onliner instead of asking...
16:29 batman but what should wait() return on error? empty list..?
16:30 sri can it return anything else?
16:30 batman it could die..?
16:30 sri it could
16:32 batman http://p.thorsen.pm/0aa8ee0062c5
16:32 batman it should be // and not || inside the on(error ...)
16:33 sri question is, should it do something like that?
16:33 batman what is "that" ? if that = my paste, then i say ++ :)
16:33 sri yes
16:34 batman after all... wait() is blocking
16:45 Vandal joined #mojo
16:46 Vandal joined #mojo
16:47 sri oh, one more edge case
16:47 sri the first closure actually runs right away
16:48 sri my $delay = Mojo::IOLoop->delay(sub { die 'First step!' }, sub { die 'Second step!' });
16:49 sri so this dies right away, and couldn't be catched
16:53 sri hmm, that's actually fixable
16:53 sri could delay the first closure a little with a 0 timer
17:10 btyler batman: ok, pushed those fixes
17:18 dvinciguerra joined #mojo
17:23 ladnaV joined #mojo
17:30 asarch joined #mojo
17:46 jberger joined #mojo
17:47 * jberger is catching up on the log
17:48 jberger Interesting stuff
17:49 jberger I'll have to think about it
17:49 jberger I'll try to ponder and get back to you after work
17:50 jberger One more edge case: multiple begins in the same step
18:12 fuzzy8121 joined #mojo
18:12 fuzzy8121 hello folks
18:12 fuzzy8121 i'm sorry i've never asked for help before on IRC so I'm not nuanced on the protocol
18:12 fuzzy8121 but I do have a mysql connectivity issue
18:14 batman fuzzy8121: i think you're better off in #dbix-class - but please go ahead and ask...
18:14 fuzzy8121 well i have used dbix::connection
18:14 wsri joined #mojo
18:14 marcus joined #mojo
18:15 fuzzy8121 my issue is, if i lose connectivity to the mysql db, its ok, mojo continues to go along
18:15 batman joined #mojo
18:15 fuzzy8121 but it doesn't try to reconnect
18:16 fuzzy8121 i need the ability to reconnect
18:19 batman fuzzy8121: use DBIx::Class. MySQL sucks without it.
18:19 batman i also think there's some DBIx::Reconnector project, but i can't really remember the name...
18:25 fuzzy8121 does DBIx::Class reconnect on its on or does it just provide the ORM stuff (i'm not shy when it comes to SQL queries)
18:25 ladnaV I belive its DBIx::Connector
18:26 ladnaV fuzzy8121, DBIx::Class is ORM, yes
18:26 sri jberger: there is no difference between one and two begins
18:27 sri second would just reduce the counter, only gets problematic when the next step could be reached
18:27 jberger What if you die in one of the processes?
18:27 sri processes?
18:27 jberger For lack of a better term
18:28 jberger Deferred actions
18:28 sri you mean if the timer dies internally somehow?
18:28 jberger Yeah
18:28 sri out of scope, nothing we could do about that
18:28 jberger Or say it's not a timer
18:29 jberger Yeah I'm probably over thinking this
18:29 sri clients like Mojo::UserAgent have custom ways of handling exceptions for that reason
18:31 sri btw. low level errors inside Mojo::UserAgent for example propagate up the stack until they reach Mojo::IOLoop or Mojo::Reactor, where they will be caught
18:31 sri those are really a worst case
18:32 sri http://mojolicio.us/perldoc/Mojo/Reactor/#error
18:32 sri this is where it all ends, if you screwed up really really bad
18:35 jberger sri: Good to know
18:35 jberger Bender: spread the love
18:36 jberger This thing never loves me
18:36 sri there's also a recipe in the cookbook :) http://mojolicio.us/perldoc/Moj​olicious/Guides/Cookbook#Timers # towards the end
18:36 jberger Bender: share the love
18:36 sri Bender: trust jberger
18:36 Bender OK, sri
18:37 * jberger misses purl
18:37 * sri too
18:47 sri think i might have a patch for delay exceptions
18:47 sri just gotta write some more tests
18:58 mire joined #mojo
18:59 sri ok
19:00 sri http://pastie.org/8303949 # here's a first patch
19:00 * sri pokes marcus, jberger, tempire and batman
19:03 jberger sri: any chance you could push this as a branch, it's hard to read this as a diff
19:03 trone joined #mojo
19:04 good_news_everyone joined #mojo
19:04 good_news_everyone [mojo] kraih created delay_exceptions (+1 new commit): http://git.io/tX53CA
19:04 good_news_everyone mojo/delay_exceptions a0b612b Sebastian Riedel: added error event to Mojo::IOLoop::Delay
19:04 good_news_everyone left #mojo
19:04 sri jberger: there you go
19:05 jberger Thanks
19:05 sri and don't worry about the tests and documentation i removed, those were added by me yesterday ;p
19:10 jberger :)
19:10 jberger Looks good here
19:10 jberger I can't actually try it, but visually it makes sense
19:11 sri try thinking of more edge cases :)
19:15 * sri shakes marcus
19:31 batman sri: i'm not very poke-friendly atm. my very special lady is coming over :)
19:31 * sri shakes batman
19:31 batman ;)
19:32 batman die $err if defined $err; # +1
19:32 batman :)
19:32 btyler joined #mojo
19:32 sri fun thing... i just changed that line ;p
19:33 good_news_everyone joined #mojo
19:33 good_news_everyone [mojo] kraih pushed 1 new commit to delay_exceptions: http://git.io/aXAcOA
19:33 good_news_everyone mojo/delay_exceptions 4b8a860 Sebastian Riedel: small optimization
19:33 good_news_everyone left #mojo
19:33 batman haha
19:33 batman same shit, different wrapping ;)
19:33 batman makes sense to me
19:33 batman but i only scanned the code...
19:34 good_news_everyone joined #mojo
19:34 good_news_everyone [mojo] kraih pushed 1 new commit to delay_exceptions: http://git.io/3EO_AA
19:34 good_news_everyone mojo/delay_exceptions 8f76007 Sebastian Riedel: fixed a few typos in tests
19:34 good_news_everyone left #mojo
19:34 sri https://github.com/kraih/mojo/co​mpare/master...delay_exceptions
19:35 batman btyler: can you also update Changes ?
19:35 sri gotta love github
19:35 batman btyler: and add your name?
19:35 btyler oh sure
19:35 batman sri: yeah. i can't remember how life was before github, but i imagine it was awful :)
19:36 * sri vaguely remembers subversion... ewwww
19:36 batman btyler: don't add timestamp. just version and changes...
19:36 batman mypp will add timestamp
19:38 batman version = previous + 0.001 or something...
19:38 batman i think me and marcus are afraid of 1.0 ;)
19:38 btyler yep, got it, pretty clear looking at the previous ones :)
19:39 chansen sri: true, I also remember RCS and CVS ;P
19:39 sri heh
19:39 inokenty joined #mojo
19:40 * sri started with subversion and touched cvs only a few times
19:41 chansen sri: what do you think of this Ruby-ism <https://gist.github.com/chansen/6463123>
19:46 btyler batman: pushed
19:47 sri chansen: neat
19:48 sri Package::Anon is one of those things i wish were part of the core language
19:50 chansen I'm abusing Package::Anon in the above gist, but year Package::Anon in core would be nice
19:50 chansen s/year/yeah/
19:50 * chansen written to much datetime code
19:52 sri i wouldn't say abuse
19:53 chansen true, I'm just removing the P::Anon provided isa method
19:57 good_news_everyone joined #mojo
19:57 good_news_everyone [mojo] kraih merged delay_exceptions into master: http://git.io/p8yG5w
19:57 good_news_everyone left #mojo
19:58 sri nobody is going to say no anyway :)
19:59 good_news_everyone joined #mojo
19:59 good_news_everyone [mojo] kraih deleted delay_exceptions at 8f76007: http://git.io/ZIiB-Q
19:59 good_news_everyone left #mojo
19:59 batman btyler: it's on it's way to cpan now
19:59 batman thanks
20:03 btyler cool, glad to help
20:10 jberger sri that's why I had you push it to a branch :-P
20:10 jberger github++
20:10 jberger sri++ # cool feature
20:10 sri \o/
20:11 marcus \o/
20:31 good_news_everyone joined #mojo
20:31 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/zHUVeA
20:31 good_news_everyone mojo/master e7f42ef Sebastian Riedel: more delay tests and documentation
20:31 good_news_everyone left #mojo
20:47 rem_lex|pivo joined #mojo
20:55 d4rkie joined #mojo
21:01 moltar joined #mojo
21:17 dsteinbrunner joined #mojo
21:18 good_news_everyone joined #mojo
21:18 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/Y-PVVA
21:18 good_news_everyone mojo/master 4b52383 Sebastian Riedel: small optimizations
21:18 good_news_everyone left #mojo
21:19 good_news_everyone joined #mojo
21:19 good_news_everyone [mojo] kraih tagged v4.32 at c921bb5: http://git.io/1vGlMQ
21:19 good_news_everyone left #mojo
21:49 wsri joined #mojo
21:49 marcus joined #mojo
21:50 batman joined #mojo
21:51 mars joined #mojo
21:52 some_dude joined #mojo
22:15 mars joined #mojo
22:16 KindOne joined #mojo
22:27 whatitdo joined #mojo

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