The web in a box - a next generation web framework for the Perl programming language

IRC log for #mojo, 2014-03-03

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

All times shown according to UTC.

Time Nick Message
00:06 jberger_ HARNESS_OPTIONS=j9 cpamn Mojolicious
00:06 jberger_ ^^ even faster
00:10 sri no clue how anyone could get work done without perlbrew
00:10 jberger_ I wish cpamn could take -j9
00:10 jberger_ sri agreed
00:11 sri i need at least one installation of the latest perl, and one of the lowest i have to support for the project
00:15 judofyr joined #mojo
00:18 sujithm joined #mojo
00:20 Akron joined #mojo
00:22 jack joined #mojo
00:23 skullbocks joined #mojo
00:24 Akron bpmedley: It's getting better. The less code you need for demonstrating what you want, the better it is.
00:25 bpmedley Akron: Thanks.  I agree it needs to be streamlined; however, I also want the user to interface with a stylish interface.
00:26 sri i don't think the style of the actual app matters as much as the surroundings
00:26 sri if you put a jsfiddle-ish border around it everything looks stylish
00:26 bpmedley I used the wrong word, I think.  I would like the output they generate in the tutorial to be somewhat stylish.
00:27 bpmedley Yes, that's on the todo list; however, I've been working on a different project for a few days.
00:27 basiliscos joined #mojo
00:31 Akron bpmedley: But if you have globaly something nice for headlines, paragraphs etc. it should be enough.
00:31 Akron globally
00:33 Akron To all the native english speakers out there: In case I'll give a talk about mojo plugins in english, how does "Hook into the Rainbow: Writing Plugins for Mojolicious" sound?
00:34 Akron bpmedley: I thought today about doing a live demo using liveperl.us - but I think I'm typing too slow. So I'll just present it afterwards I guess.
00:38 jberger_ I've been trying to decide what to talk about at mojoconf
00:39 jberger_ I could do some nonblocking
00:39 jberger_ Or the command system
00:40 punter joined #mojo
00:43 Akron No public reading of some aphorisms from your upcoming Mojo book?
00:48 jberger_ Haha if only it had progressed that far
00:49 laouji joined #mojo
00:50 jberger_ With any luck $work will be a little slower now that we are in a testing code freeze
00:50 Akron Are you doing perl at the job?
00:50 jberger_ Versus the mad coding rush of the last few weeks
00:51 jberger_ Sadly no
00:51 jberger_ Python
00:51 Akron Not bad either.
00:51 jberger_ There is perl in other groups though
00:51 jberger_ Even some mojo
00:51 * sri excommunicates jberger
00:52 Akron Then I assume it's no scientific postdoc job.
00:52 jberger_ If it were up to me!
00:52 jberger_ No
00:52 jberger_ Financial firm
00:52 * Akron (won't tell sri he is doing java for living currently)
00:52 jberger_ The sciences are dead in the us
00:52 * sri faints
00:52 jberger_ Funding is cut everywhere
00:53 jberger_ I have my eyes open
00:53 Akron The funding is fine in germany at the moment.
00:53 jberger_ But the near term prospects for a scientific job are not good
00:53 Akron But you have to find new projects regularly ... so no fun.
00:53 * jberger_ is jealous
00:54 Akron Yeah.
00:54 Akron But you are making money!
00:54 jberger_ I'm being paid really well though
00:54 jberger_ Haha
00:54 Akron See!
00:54 jberger_ I have learned a lot by learning python though
00:55 jberger_ A very different paradigm on a very similar platform
00:56 jberger_ See Generator::Object
00:56 * sri wonders if mojolicious should use the "reactive programming" buzzword
00:56 jberger_ wtf is that
00:56 purl i heard that was the issue.  I would like to have the catalyst stuff just as adaptors for the plack stuff but there's enough trouble there to not encourage the work to do so :( or fine, although I thought it had some other issue?  Likely correctable.
00:56 Akron Really? I thought Python was quite similar when I learned it. Well ... compared to Prolog and XSLT I did before.
00:56 sri EventEmitter is reactive programming ;p
00:57 jberger_ The object model is very different
00:57 sri it's a fancy name for the observer pattern
00:57 * Akron thought reactive programming is when my boss looks over my shoulder ...
00:58 Akron The object system is everywhere, yes.
00:58 * sri actually starts to enjoy doing more JavaScript these days :o
01:05 jb360 joined #mojo
01:10 judofyr joined #mojo
01:24 good_news_everyone joined #mojo
01:24 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/UYTleg
01:24 good_news_everyone mojo/master ddcce0b Sebastian Riedel: improved Mojo::IOLoop::Delay to allow more method chaining
01:24 good_news_everyone left #mojo
01:25 sri you can actually do $delay->begin(0)->('foo')->begin(0)->('bar'), not sure yet if that's a feature or an implementation detail :D
01:26 sri not actually documented, but kinda implied by ->pass() begin equal to ->begin(0)->()
01:26 * sri shrugs
01:29 Akron joined #mojo
01:39 bpmedley Akron: The title sounds fine - fun even.
01:41 good_news_everyone joined #mojo
01:41 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/t6KB8w
01:41 good_news_everyone mojo/master a7ed2b8 Sebastian Riedel: more delay tests
01:41 good_news_everyone left #mojo
01:46 d4rkie joined #mojo
01:54 sri yay, managed to fix the atom-perltidy bug where the cursor would jump around after an undo/redo
01:54 sri cursor position seems to be rock solid now :) https://github.com/kraih/atom-perltidy/commit/f9a8bbf0634cc18c1f75308ccc6250315162586e
02:03 sujithm joined #mojo
02:04 judofyr joined #mojo
02:11 ka2u joined #mojo
02:11 judofyr joined #mojo
02:21 klapperl_ joined #mojo
02:31 Akron bpmedley: Thanks - yeah, I planned it to be a bit entertaining.
02:32 Akron First draft for a potential intro slide: http://batteriehuhn.de/temp/mojo-hook-draft-1.jpg
02:33 jberger_ Hahaha excellent!
02:33 Akron Thanks - I should integrate a raptor at least ...
02:38 sri rofl
02:38 sri that is awesome
02:41 jberger_ Akron are you going to go as far as OO plugins?
02:41 Akron I don't beliebe in OO ... or ... what do you mean?
02:42 Akron Damn - I am no belieber at all - beliebe I meant!
02:44 Akron believe!
02:44 purl I believe in the Beatles except for Paul.
02:44 Akron What purl says.
02:44 jberger_ At least I have been trying to work on some best practices for persistent object plugins
02:44 Akron jberger_: You mean plugins with returning objects and their own methods?
02:44 jberger_ And I know that some people have copied the style
02:44 jberger_ More like you install a helper that closes over the plugin object
02:45 Akron I would be interested in your best practices on that - Unfortunately I think I won't be able to go too much into detail for the talk. It seems to be the only Mojo talk on the conference, so I will have to do some intro stuff and then I want to show some of the possible extension points of Mojolicious.
02:45 jberger_ https://metacpan.org/source/JBERGER/Mojolicious-Plugin-Humane-0.05/lib/Mojolicious/Plugin/Humane.pm#L45
02:45 Akron Helpers, Hooks, Conditions, Validators, Shortcuts, Commands ... etc.
02:46 jberger_ Oh this isn't for mojoconf?
02:46 Akron Nope - German Perl Workshop.
02:46 jberger_ In that case then no need to muddy it up
02:47 Akron Ah - I even use humane. Yes - it's what I thought you mean. There's one plugin I wrote that does something similar ...
02:47 jberger_ If this were for mojoconf then I would encourage you to take it further
02:47 jberger_ But for a non specialized conference, keep it simple
02:47 skullbocks joined #mojo
02:48 Akron Although - it's not the plugin object: https://metacpan.org/source/AKRON/Mojolicious-Plugin-BlogSpam-0.06/lib/Mojolicious/Plugin/BlogSpam.pm
02:48 Akron jberger_: I agree. Don't know if I'll make it to Mojo Conf.
02:49 jberger_ Yeah that's a similar concept
02:50 Akron I'll try to demonstrate most possible extension points by writing one useful plugin tutorial style ... that's pretty hard.
02:50 * jberger_ kidnaps Akron and brings him to mojoconf
02:51 * Akron has Stockholm syndrome.
02:51 jberger_ I encourage you to consider using mojo eval for demos too
02:52 jberger_ or ./myapp eval more specifically
02:52 Akron My typing is too slow for live demos!
02:53 Akron I try to make the audience simply believe what I tell them by showing colorful pictures!
02:54 jberger_ Ah. Good plan
02:54 jberger_ I like to make my db methods as helpers
02:55 jberger_ Then I can use eval on the command line to do quick queries
02:56 jberger_ https://github.com/jberger/MojoForum/blob/master/lib/MojoForum/Helpers.pm
02:56 Akron I have a separated account management system based on DBIx::Oro with a CLI for management - and I wrote a command for Mojolicious to directly manipulate the db of the app. I totally love the command system.
02:56 jberger_ Nice
02:56 Akron Oh - didn't saw MojoForum!
02:57 jberger_ Not a real project ... yet
02:57 jberger_ Tinkering with mandel
02:58 jberger_ I needed a real world use case to try with
02:59 Akron Nice start. I always wanted to have a good reason to dive into DBIx::Class ... never found one.
02:59 jberger_ That was Galileo for me :-)
03:00 Akron You are way more productive than me!
03:00 Akron ... and earn more money ...
03:00 jberger_ I think it's next iteration will be mandel based
03:00 jberger_ Haha
03:00 Akron ... well. Sounds fair.
03:01 Akron Oh - damn. 4am and I've got an appointment tomorrow. Have to go. See you!
03:01 Akron Good night!
03:01 jberger_ O/
03:02 * jberger_ wonders if he can turn off auto capitalization on android in irc apps
03:05 judofyr joined #mojo
03:10 jberger__ joined #mojo
03:12 sri the d3.js guy has his own async control flow library too https://github.com/mbostock/queue
03:14 sri only works when you can pass around methods though
03:16 sri my ($first, $second) = Mojo::Queue->new->defer($ua, 'get', 'http://mojolicio.us')->defer($ua, 'get', 'http://mojolicio.us')->wait;
03:16 sri dunno, that seems a little silly :)
03:16 * jberger_ dislikes
03:16 purl dislikes are normal. do whatever you can to make sure
03:17 * jberger_ smacks purl with a comically large fish
03:34 jberger back on convos
03:59 judofyr joined #mojo
04:11 laouji joined #mojo
04:27 laouji joined #mojo
04:53 judofyr joined #mojo
05:01 ka2u joined #mojo
05:41 good_news_everyone joined #mojo
05:41 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/7B__7w
05:41 good_news_everyone mojo/master ac9a001 Sebastian Riedel: improved WebSocket and long poll performance
05:41 good_news_everyone left #mojo
05:44 good_news_everyone joined #mojo
05:44 good_news_everyone [mojo] kraih tagged v4.86 at b1cb439: http://git.io/GesQmw
05:44 good_news_everyone left #mojo
05:48 judofyr joined #mojo
05:50 jb360 joined #mojo
05:51 sujithm joined #mojo
06:20 Vandal joined #mojo
06:28 davido_ Someone posted a writeup about an RSS aggregator based on Mojolicious::Lite on PerlMonks: http://perlmonks.org/?node_id=1076732
06:42 judofyr joined #mojo
06:51 sujithm_ joined #mojo
06:54 sujithm joined #mojo
07:00 firnsy joined #mojo
07:12 hrupp joined #mojo
07:13 gnephiak joined #mojo
07:13 firnsy joined #mojo
07:19 sujithm joined #mojo
07:30 sujithm_ joined #mojo
07:34 HtbaaPi joined #mojo
07:36 judofyr joined #mojo
07:36 basiliscos joined #mojo
07:43 dod joined #mojo
07:48 bpmedley joined #mojo
07:49 rawler joined #mojo
07:49 firnsy joined #mojo
07:55 dod joined #mojo
08:00 zz_Foxcool joined #mojo
08:19 judofyr joined #mojo
08:24 trone joined #mojo
08:34 mire__ joined #mojo
08:36 sujithm joined #mojo
08:53 zz_Foxcool joined #mojo
09:22 fhelmber_ joined #mojo
09:26 mire__ joined #mojo
09:30 Andy2 joined #mojo
09:35 maxhq joined #mojo
09:38 sujithm_ joined #mojo
09:39 sujithm joined #mojo
09:44 plenum joined #mojo
10:13 odc joined #mojo
10:17 hrupp joined #mojo
10:28 sujithm_ joined #mojo
10:39 Vandal joined #mojo
10:44 maxhq joined #mojo
10:48 dpetrov_ joined #mojo
10:55 dotandimet joined #mojo
11:06 rem_lex|pivo joined #mojo
11:10 denis_boyun joined #mojo
11:27 sujithm joined #mojo
11:29 sujithm_ joined #mojo
11:32 sujithm__ joined #mojo
11:42 zivester joined #mojo
12:13 Jonneh joined #mojo
12:14 Jonneh_ joined #mojo
12:15 Jonneh joined #mojo
12:22 bluescreen joined #mojo
12:58 malikai joined #mojo
13:03 zz_Foxcool joined #mojo
13:05 BenGoldberg joined #mojo
13:17 ioguix odc: je vais esasyer de faire ça un soir... :/
13:17 ioguix oops, sorry guys, wrong chan
13:17 hrupp joined #mojo
13:19 mattastrophe joined #mojo
13:30 Jonneh joined #mojo
13:37 sujithm joined #mojo
13:42 d4rkie joined #mojo
13:43 punter joined #mojo
13:44 punter Do I need to wait for an invitation from the authors to be able to download the Atom editor?
13:46 Vandal joined #mojo
13:48 punter joined #mojo
13:50 marcus Akron: Awesome slide, you should submit your talk for mojoconf too ;-)
13:56 plenum joined #mojo
14:04 zackiv31 joined #mojo
14:07 dod joined #mojo
14:11 Coke crap, didn't find more time for mojo this past weekend.
14:14 skullbocks joined #mojo
14:14 Coke is there any point in porting things that are currently marked as deprecated?
14:14 Coke (i'm guessing: certainly not first)
14:19 jberger_ joined #mojo
14:20 jberger_ sri, regarding #597: it is the expected and necessary behavior, but should it be better documented?
14:21 sh4 joined #mojo
14:23 jberger_ header is not documented to have different return values by context
14:24 jberger_ In fact I don't see any return value documentation for that method
14:24 jberger_ I suppose it is in the example
14:26 jberger_ In fact I'm changing my tune: I think the example covers it
14:26 jberger_ Ignore me
14:30 nmac joined #mojo
14:36 mire__ joined #mojo
14:37 dotandimet joined #mojo
14:43 JSI joined #mojo
14:50 sjn joined #mojo
14:54 gryphon joined #mojo
14:55 JSI jberger: Hi, I opened #597, sorry for my hasty issue report! I think, the header() must be retuned undef, when I attempt to get a non-existent header. I got an uninitalized/unknown value instead of undef or a defined value. I think this is bogus, because in my example the hash keys and values are confused, and Perl says: "odd number of hash elements"
14:56 skullbocks joined #mojo
14:57 JSI When I assign an undefined variable to a hash key, then the hash key now exists and holds an undef value. And not "odd number of hash elements" error occured
14:58 JSI Can you tell me, I have misunderstand something or why become this "odd number" error?
15:01 Jonneh joined #mojo
15:03 sh3 joined #mojo
15:10 Coke sri: any input on stuff like Mojo::Util::trim ? (in perl6 it's just $str.trim)
15:15 sujithm joined #mojo
15:20 Coke (so I would tend to just not convert that.)
15:20 aleksey joined #mojo
15:21 dotandimet joined #mojo
15:24 dod joined #mojo
15:28 genio $str =~ s/\A\s+//; $str =~ s/\s+\z//;  <-- hasn't trim() been done to death and decided that two expressions are better than the one ORed expression?
15:36 genio I also prefer a return unless defined $str; above the regular expressions.   whether you want to return an empty string or the $str itself ini that case, that's up to you
15:36 genio it just gets rid of "Use of uninitialized value $str in substitution (s///)"  stuff
15:38 bluescreen joined #mojo
15:39 marty joined #mojo
15:45 cfedde joined #mojo
15:46 HERNAN joined #mojo
15:48 disputin joined #mojo
16:04 marty joined #mojo
16:05 aggrolite joined #mojo
16:10 marty_ joined #mojo
16:11 sk joined #mojo
16:12 marty joined #mojo
16:26 basiliscos joined #mojo
16:34 mire__ joined #mojo
16:35 btyler joined #mojo
16:37 sri there's a new JSON RFC... and it breaks Mojo::JSON
16:37 sri http://tools.ietf.org/html/rfc7159
16:38 tianon it's even from Bray over at Google, interesting
16:38 tianon are the changes any good?
16:38 tianon :)
16:39 sri charset detection is no more, BOM is not allowed, and it allows stuff like "null" to be a valid JSON document
16:39 tianon ah
16:39 sri so our undef return value in case of failure is broken
16:39 tianon basically trying to legitimize the way most JSON parsers handle that stuff
16:39 sri we either stick with the old definition of JSON, or make serious breaking changes
16:40 tianon well, you are running out of version numbers, so 5.0 is right around the corner, right? >:)
16:40 sri nope
16:40 sri months away
16:41 tianon well, this is still just proposed, right?
16:41 sri no
16:42 sri it obsoletes RFC 4627
16:42 sri and it's in the standards track
16:42 tianon gotcha
16:42 sri if the spec breaks, i don't think we have much of a choice but to follow with breaking changes
16:43 Coke can you have a toggle and support both versions of the spec?
16:43 sri no
16:43 sri the api is unsupportable
16:43 sri we specifically depend on an undef return value, which is now a valid JSON document
16:44 hesco I get: "Mojolicious::secret is DEPRECATED in favor of Mojolicious::secrets", but is it really?  https://gist.github.com/hesco/db0dbfde0b9f845e2326
16:45 tianon hesco: yes?  why would Mojo give you a warning if it isn't?
16:45 tianon hesco: see also https://github.com/kraih/mojo/commit/57e5129436bf3d717a13e092dd972217938e29b5
16:46 hesco tianon: don't know but how would you then explain the subsequent error reported in the gist paste?  Apparently I do not understand how to apply this advice.
16:46 tianon $app->secrets(['new_passw0rd', 'old_passw0rd', 'very_old_passw0rd']);
16:46 hesco thanks looking now
16:46 Coke sri: is Mojo::Util used outside of mojo, or is it intended solely for internal utility functions?
16:46 tianon this way you can add a new encryption password without invalidating all the previous sessions
16:51 hesco ah, get it.  It wants an arrayref I guess.  That made the error go away in one of the two times it got thrown
16:52 tianon sri: maybe this would be a good opportunity to switch to thrown exceptions in Mojo::JSON, as you've discussed doing before (as long as we've got a big breaking change ahead of us, might as well go all the way, right?)
16:53 davido_ I think that discussion is closed, as of a week or so ago.  Besides, if you want thrown exceptions in Mojo::JSON, you can use the functions interface.
16:54 davido_ (I'm assuming you're using a recent Mojolicious version)
16:56 tianon davido_: this is specifically in the context of http://tools.ietf.org/html/rfc7159
16:56 tianon davido_: where now "undef" has to be a valid return because "null" is a valid JSON object
16:58 davido_ I see what you're saying.  Currently (based on rfc7159) when using the object interface, if the returned json object is 'undef' (null), I guess one should check $j->error() to see if there is one.  In the functions interface, exceptions should just work.
17:03 sri the OO interface works if you check ->error... but j() doesn't
17:03 sri Coke: designed for internal used, but get used outside... no, i would not port stuff perl6 already has
17:04 davido_ I've lost track... doesn't 'j' already throw now?
17:04 Coke sri: I mean, obviously the port would be trivial... but good, that's what I was thinking.
17:04 sri nope
17:04 davido_ ah, return undef if it fails.
17:04 sri we made no breaking changes in Mojo::JSON yet
17:05 sri i guess the OO interface can stay the way it is for backwards compatibility... BOM and charset stuff will prolly not be missed by anyone... but j() is kind of a problem
17:05 sri there's also a few assumptions about JSON only being a hashref or arrayref in other parts of mojolicious :S
17:06 sri those parts tend to use j() currently
17:06 sri json event for WebSockets... json_* tests
17:11 davido_ Of course j is an easy change, except for what it would break.
17:12 sri in itself... the things that depend on it are not so easy to change
17:12 sri suddely all json data types are valid JSON text
17:14 davido_ Code will start feeling a lot like Java.
17:19 sri stuff like $c->req->json is also broken
17:19 sri since it used to return undef on failure
17:20 sri we can't really do non-fatal anymore
17:20 davido_ hm... one little change to the spec, and it throws a bucket full of monkey wrenches into the mechanism.
17:21 sri my $value; eval { $value = $c->req->json; 1 } or ...handle exception...;
17:21 sri things would get ugly
17:22 davido_ Yeah, that's what I meant by "feeling a lot like Java"
17:22 dotandimet joined #mojo
17:23 sri for backwards compatibility we could leave stuff the way it is for now... but the valid JSON text "null" would be unsupported
17:24 davido_ So Mojo::JSON would no longer be a full RFC-compliant implementation.
17:24 sri it would, just one API not
17:24 davido_ oh... yes, of course.
17:24 sri or rather... it would... but the exception/"null" case would be ambiguous
17:25 davido_ It's too bad that was deemed a useful revision. :)
17:35 klapperl joined #mojo
17:43 sri oops
17:43 sri j() is totally broken anyway
17:43 davido_ lol
17:43 sri since it can't encode anything else than arrayrefs and hashrefs
17:44 sri it has to be at the very least be deprecated
17:44 davido_ Yeah, there's no sane way to fix that.
17:45 beyondcreed joined #mojo
17:46 davido_ It was fun while it lasted.
17:46 sri that makes ojo::j impossible too
17:47 sri jberger, tempire, marcus, crab: i could really use some help here
17:48 davido_ It would be really lame to do: j(encode => ...) and j(decode => ...).  But I can't think of any way to reliably differentiate now that everything is fair game.
17:56 sri the whole point was conveninece
17:57 Coke sri: whee. Think I found a bug in rakudo.
17:57 sri whee
18:03 davido_ It must be a feature.
18:08 hrupp joined #mojo
18:09 mattastrophe joined #mojo
18:10 mire__ joined #mojo
18:10 meshl joined #mojo
18:20 batman sri: i take convenience. meaning, when do you need to serialize anything else than array and hashes?
18:20 batman i really like j(). i use it all the time
18:20 sri that's bullshit
18:20 batman could be, but not to me :/
18:20 sri spec compliance goes first
18:21 sri valid JSON is now more than arrays and objects, we have to deal with that reality
18:21 batman so you will deprecated j() then?
18:21 sri yes
18:21 batman that sucks big time :(
18:21 sri complain to the IETF
18:22 batman why can't you have a "helper" with limitations? so the methods would be complient, but j() would support a subset?
18:22 batman methods = $j->encode + decode
18:23 sri because it sucks to explain in docs
18:23 sri people already have trouble grasping what is valid and what is not
18:24 batman i would very much like to help you with the docs if that is what it takes to keep j() around
18:24 sri you're welcome to propose a fool proof explanation, if it gets support in the community i'll consider it
18:24 batman right. should i just make a PR for that?
18:25 batman i will look into it later this week. right now i'm too tired to come up with anything useful :(
18:27 sri i want this fixed now
18:33 good_news_everyone joined #mojo
18:33 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/o1XbSw
18:33 good_news_everyone mojo/master f9d4985 Sebastian Riedel: fixed RFC 7159 support in Mojo::JSON
18:33 good_news_everyone left #mojo
18:36 sri i've not made big breaking changes yet, but a lot of stuff is ambiguous now
18:55 mudler joined #mojo
18:56 mudler joined #mojo
18:57 mudler joined #mojo
19:03 dotandimet joined #mojo
19:06 sujithm joined #mojo
19:07 * malikai searches for any usage of j()
19:11 bluescreen_ joined #mojo
19:12 denisboyun joined #mojo
19:20 good_news_everyone joined #mojo
19:20 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/tEGcxw
19:20 good_news_everyone mojo/master 0254ad1 Sebastian Riedel: fixed RFC 7159 compliance bugs in Mojo::Transaction::WebSocket and Test::Mojo
19:20 good_news_everyone left #mojo
19:21 sri that should actually address most problems... aside from the exception/"null" ambiguity
19:21 dotandimet joined #mojo
19:21 sri j() is not deprecated yet... dunno what to do with it
19:31 davido_ Deprecate, cause it to throw a warning, remove soon.
19:49 punter joined #mojo
19:49 sri i guess deprecating has the most support now
19:54 mire__ joined #mojo
19:56 beyondcreed joined #mojo
20:02 good_news_everyone joined #mojo
20:02 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/J_4VdQ
20:02 good_news_everyone mojo/master a0db1ba Sebastian Riedel: use values instead of data structures in descriptions
20:02 good_news_everyone left #mojo
20:06 bluescreen joined #mojo
20:18 good_news_everyone joined #mojo
20:18 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/VInG3g
20:18 good_news_everyone mojo/master 2063544 Sebastian Riedel: improved trim performance a little (closes #599)
20:18 good_news_everyone left #mojo
20:27 malikai is there a public need for an ldap module for mojo?
20:28 malikai i've needed one and hacked up a way to do it using net::ldap.. might be able to make it a plugin
20:29 jberger_ JSI: hash creation evaluates the headers call in list context
20:29 jberger_ Therefore when there are no headers by that name you get an empty list
20:29 malikai by ldap i mean strictly bind() :)
20:30 malikai something like mojo::auth::ldap
20:31 jberger_ JSI: you might want to do `->headers(...) || undef` or else evaluate in scalar context
20:37 sri hmm... browsers already support "JSON.parse("null");"
20:39 sri jberger: what's your opinion on j()
20:39 lukep joined #mojo
20:40 sri the current description https://github.com/kraih/mojo/blob/master/lib/Mojo/JSON.pm#L365
20:41 good_news_everyone joined #mojo
20:41 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/ZNa9vA
20:41 good_news_everyone mojo/master ca08ed7 Sebastian Riedel: documentation tweak
20:41 good_news_everyone left #mojo
20:42 davido_ Did "C<Hash>" get broken in that commit?
20:42 good_news_everyone joined #mojo
20:42 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/-FD4BA
20:42 good_news_everyone mojo/master 4464ac3 Sebastian Riedel: fixed typo
20:42 good_news_everyone left #mojo
20:43 davido_ :)
20:43 sri what the hell is going on
20:43 purl A complete clusterfuck.
20:43 sri Atom breaks my edits somehow
20:44 davido_ Perfect, because that's I<exactly> what you want an I<editor> to do. ;)
20:44 good_news_everyone joined #mojo
20:44 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/lONyeg
20:44 good_news_everyone mojo/master 1c8029b Sebastian Riedel: fixed another typo
20:44 good_news_everyone left #mojo
20:44 malicious well
20:44 malicious time for a new car
20:45 malicious http://blogimages.thescore.com/nhl/files/2013/06/Shut-up-and-take-my-money.jpg
20:45 jberger_ Tbh I haven't had time to read the whole backlog
20:45 sri what i imagine tempire would drive http://24.media.tumblr.com/24bff850c0a6a2cf8b92ae96132356b5/tumblr_mvz3qo55yE1sn3q9eo1_1280.jpg
20:45 jberger_ But it looks like they changed the standard on us
20:45 malicious I'd be all over that.
20:45 malicious As long as it has carplay
20:46 davido_ All it needs is air-cond and a side awning.
20:47 sri but seriously... think about what to do with j() https://github.com/kraih/mojo/blob/master/lib/Mojo/JSON.pm#L365
20:47 davido_ Bonus points if it's a home window AC jury rigged to the back window.
20:48 sri problems are a) return value may be undef even for valid JSON, and b) you can't encode anything else than array or hash
20:50 sri this change also affects $c->req->json kinda... since it was supposed to return undef for exceptions too
20:51 jberger_ Well we were going to make it die anyway right?
20:51 jberger_ So the problem is b
20:51 malikai if you're going for that vintage, you might as well go with style.. http://www.airliners.net/photo/Untitled-(The-Flying/Bede-BD-5J/2078551/L/
20:52 sri we have the functional api for that now, j() was supposed to stay the same
20:52 jberger_ Then maybe j should just stay as is
20:53 jberger_ With noted caveats tuned for common use
20:56 good_news_everyone joined #mojo
20:56 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/xPAarA
20:56 good_news_everyone mojo/master 8d2fe50 Sebastian Riedel: use encode_json instead of j in examples
20:56 good_news_everyone left #mojo
20:57 * sri shrugs
20:58 jberger_ Was there any advance warning of this change?
20:58 sri what kind of advance warning?
20:59 sri it's an RFC... there's prolly been drafts for like years
20:59 davido_ The browser vendors seemed to know of it.
20:59 sri but you would have to follow the specific working group
20:59 jberger_ Ick
21:00 sri apparently the change is meant to bless browser implementations
21:11 gryphon joined #mojo
21:12 skullbocks joined #mojo
21:21 meshl joined #mojo
21:23 aggrolite joined #mojo
21:23 * malicious just discovered cpanm's backpan support
21:45 mattastrophe joined #mojo
22:15 good_news_everyone joined #mojo
22:15 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/rwWAqw
22:15 good_news_everyone mojo/master 349482b Sebastian Riedel: explain what an undef return value means
22:15 good_news_everyone left #mojo
22:15 sri ok, that should cover all cases... doesn't sound as bad as i expected
22:15 sri http://mojolicio.us/perldoc/Mojo/JSON#j
22:19 good_news_everyone joined #mojo
22:19 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/67_Kmw
22:19 good_news_everyone mojo/master 3f01d42 Sebastian Riedel: documentation tweak
22:19 good_news_everyone left #mojo
22:23 davido_ Yeah, that's the best possible "keep the feature" documentation.
22:24 sri hahaha... mlehmann is about as pissed as i am about the rfc
22:24 sri http://www.rfc-editor.org/errata_search.php?rfc=7158
22:24 sri and he has a point
22:24 sri json protocols are all broken now
22:26 davido_ Wow, he does have a point.
22:27 sri doesn't matter though, browsers trump everything
22:28 davido_ Well, and even people who don't plan to produce JSON using the new "features" has to be concerned with consuming json or risk being handed something that is both valid and not understood by their implementation.
22:29 good_news_everyone joined #mojo
22:29 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/1to42Q
22:29 good_news_everyone mojo/master ad3722b Sebastian Riedel: link to error attribute
22:29 good_news_everyone left #mojo
22:45 sri yea, i guess editor games will become a thing with Atom :D https://github.com/mojombo/asteroids
22:49 skullbocks joined #mojo
22:49 sri btw. on the plus side... Mojo::JSON requires less lines of code now
22:57 good_news_everyone joined #mojo
22:57 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/fwDeJw
22:57 good_news_everyone mojo/master bd94b59 Sebastian Riedel: bring back test for surrogate pair
22:57 good_news_everyone left #mojo
23:03 marty joined #mojo
23:24 sri the new RFC is actually stronger than the old one, it went from INFORMATIONAL to PROPOSED STANDARD
23:36 sri the little screwup was funny too... RFC 7159 has the exact same content as RFC 7158... except for the date... which said 2013 -.-

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