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

IRC log for #mojo, 2014-10-03

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

All times shown according to UTC.

Time Nick Message
00:00 jberger batman: I got home and the last message in the cpan_mojo log is:
00:00 jberger [Fri Sep 26 19:41:24 2014] [error] Connection timed out
00:00 D4RK-PH0ENiX joined #mojo
00:01 jberger though the procecss was still running
00:05 woz joined #mojo
00:10 sri jberger: look into the non-blocking resolver stuff, you're one of the few qualified to help!
00:10 jberger EDATENIGHTWITHWIFE
00:12 fhelmber_ joined #mojo
00:13 jberger going to see Gone Girl
00:13 jberger will report
00:13 jberger and yes I will happily look when I get the chance
00:51 uqurat joined #mojo
01:01 dvinciguerra_ joined #mojo
01:06 woz joined #mojo
02:04 noganex_ joined #mojo
02:08 woz joined #mojo
02:23 sri allright, aside from migrations, i think i have everything i originally wanted working now, maybe time for a first experimental release :)
02:27 firnsy push the button sri ;)
02:56 d4rkie_ joined #mojo
03:07 preaction joined #mojo
03:09 woz joined #mojo
03:10 basic6_ joined #mojo
03:47 Averna joined #mojo
04:10 woz joined #mojo
04:12 fhelmber_ joined #mojo
04:16 davido_ joined #mojo
04:16 jberger ok. go see Gone Girl!
04:17 jberger also my wife is awesome for never taking me to twilight but taking me to movies like that!
04:20 firnsy jberger wife++
04:45 jamesaxl joined #mojo
04:49 cpan_mojo Mojo-Pg 0.01 by Sebastian Riedel - http://metacpan.org/release/SRI/Mojo-Pg-0.01
04:53 sri oops, the bot spoiled it
04:53 sri https://github.com/kraih/mojo-pg
04:53 ryanc joined #mojo
04:53 sri migrations will follow, once i know what exactly i want
05:03 sri i'm certain it's full of bugs
05:04 Adurah joined #mojo
05:06 sri and highly experimental, it has some fun properties
05:07 sri my $db = $pg->db; $db->query(..., sub {...}); $db->query(..., sub {...});
05:07 sri the second query actually ends up on a queue, and will be performed once the first is finished
05:08 sri $pg->db->query(..., sub {...}); $pg->db->query(..., sub {...});
05:08 sri that one would actually use two connections and perform the queries concurrently
05:10 sri there's a dbh and sth cache
05:10 sri both work automatically
05:11 sri with fork protection of course
05:11 woz joined #mojo
05:11 sri looking for stuff to remove again now... minimalism and all :)
05:19 firnsy oooo ... shiny
05:20 KCL_ joined #mojo
05:56 marcus <3 sri <3
06:08 tempire OMGINEEDTHAT
06:09 sujithm joined #mojo
06:09 Adurah We Postgres now?
06:10 tempire This is excellent.
06:10 tempire Exactly what I've been thinking.
06:10 tempire It's what I've been doing in objc
06:11 Adurah Something I'd certainly make use of when I'm called on.
06:12 tempire postgres was the obvious choice
06:12 woz joined #mojo
06:14 sujithm_ joined #mojo
06:14 * firnsy likes the notification exposure
06:36 marcus I want to rewrite iwatchthis in mojo now. using this.
06:42 batman marcus: Haha! I will fix the avatar :)
06:43 marcus batman: <3
06:44 batman jberger: ok. That sucks :/ I will have a look at the code.
06:44 dod joined #mojo
06:46 batman Need another username for the bot. Maybe just "jh"
07:06 uqurat whats the best drawing everything together thing out there?
07:07 uqurat like terminal is liney with programs and can do development
07:07 uqurat something that's just flowing stuff
07:11 rawler joined #mojo
07:14 woz joined #mojo
07:28 mishantil sri: If you're going to do migrations for Mojo::Pg, please consider supporting migration-path downwards as well as up.
07:28 mishantil We wrote some in-house stuff that can go both ways, plus go to a specific version of the schema. Very practical.
07:33 trone joined #mojo
07:34 Oleg joined #mojo
07:37 neyasov joined #mojo
07:38 batman sri: how would you feel about having a uri instead of dsn+username+password?
07:38 batman not sure how many that does it the same way as heroku, but they use this: HEROKU_POSTGRESQL_RED_URL=postgres://user3123:passkja83kd8@ec2-117-21-174-214.compute-1.amazonaws.com:6212/db982398
07:40 batman dotcloud does the same: postgres://root:pass@7a96f954.dotcloud.com:7780
07:42 rawler joined #mojo
07:42 batman what? you can do PUBSUB with pg??!
07:45 batman <3 Mojo::Pg
07:48 basiliscos joined #mojo
07:50 mishantil batman: We do the same inhouse, only we use pg://
07:51 mishantil Saves some typing. :)
07:51 batman i guess both should be valid
07:55 Vandal joined #mojo
08:06 stryx` joined #mojo
08:13 fhelmber_ joined #mojo
08:15 woz joined #mojo
08:24 sujithm joined #mojo
08:25 denis_boyun joined #mojo
08:30 Lee joined #mojo
08:32 bpmedley joined #mojo
09:16 woz joined #mojo
09:25 marcus mishantil: interesting. We've been discussing a lot and really not seen practical use cases for down migrations.
09:26 fhelmber_ joined #mojo
09:42 neyasov joined #mojo
09:50 batman there is only one way, and that is forward!
09:51 batman i think down migrations should be the same as a new forward rule. i've never understood rollback :/
10:17 woz joined #mojo
10:18 sujithm joined #mojo
10:31 sujithm joined #mojo
10:42 batman jberger: you asleep..? i was wondering how ForkCall avoid children waiting to get reaped.
10:43 batman is there something magical about eval { POSIX::_exit(0) } unless IS_WINDOWS ?
11:18 woz joined #mojo
11:25 sujithm joined #mojo
11:46 rawler joined #mojo
11:46 bpmedley_ joined #mojo
11:57 neyasov_ joined #mojo
11:58 neilhwatson joined #mojo
11:59 d4rkie joined #mojo
12:07 neyasov joined #mojo
12:11 jberger 1 while ++sri
12:12 * jberger 's ocd now compels him to issue ^C
12:14 moritz jberger: no worry, sri will overflow, resulting in a 0 eventually :-)
12:14 jberger hehe
12:14 jberger batman: I waitpid once I know the child is already dead
12:16 jberger hmmmmm but I should issue a kill and waitpid in the "on error" handler too probably eh?
12:16 jberger Does close get emitted if there is an error?
12:17 * jberger uses the source
12:17 batman the source is strong with jberger
12:18 jberger And the answer is yes
12:18 jberger jberger: does that make sense to you?
12:19 batman but i don't get how the "close" event is fired if the child is not wait()ed
12:19 batman i thought the pipe was dangeling until wai/waitpid :/
12:19 woz joined #mojo
12:20 jberger I guess not
12:21 jberger remember that I cribbed this whole thing from ae util and poe wheel
12:21 jberger though I didn't cargo cult mind you
12:21 batman but do you think the fh is closed because of POSIX::_exit() ?
12:22 batman i'm trying to make M::P::CGi a bit simpler, and in that context i have no control over the child
12:22 jberger I ported line by line each when I realized why it was necessary
12:22 jberger batman: probably
12:22 batman ok
12:22 jberger oh
12:22 jberger yes I remember
12:23 jberger that is it
12:23 batman what is it?
12:23 purl it's it!
12:24 jberger _exit
12:24 batman ok. thanks.
12:24 * batman have to figure out something else then...
12:25 jberger is kill with fire, close all handles, do nothing else
12:25 jberger except it kills the parent on Windows
12:25 batman awesome
12:25 batman :)
12:25 batman thanks jberger
12:25 jberger fork emulation by threads :)
12:26 jberger np
12:26 jberger sorry that won't help you
12:26 batman yeah, me to :)
12:26 batman i guess i need to to that Mojo::IOLoop->recurring(0.01 => sub { waitpid $pid }); thing i had :/
12:30 nicomen jberger: where was your simple websocket thing chat again?
12:37 lipizzan joined #mojo
12:38 jberger nicomen: on my gist
12:39 nicomen right right, thanks ;)
12:39 jberger https://gist.github.com/jberger/78d7c1649d7acbab6a94
12:43 nicomen yeah already found it, thanks ;)
12:58 mudler joined #mojo
13:00 mudler joined #mojo
13:01 mudler joined #mojo
13:09 mudler joined #mojo
13:10 mudler joined #mojo
13:18 mudler joined #mojo
13:21 woz joined #mojo
13:22 mudler joined #mojo
13:24 mudler joined #mojo
13:31 sri batman: hahahaha, i actually started with a postgres:// uri
13:31 batman why did you change?
13:32 batman uri++
13:32 sri to support all the configuration variants supported by DBD::Pg
13:33 sri like you know, unix domain sockets
13:33 sri which is picked automatically actually if you don't set a host and the server is the same you compiled DBD::Pg against
13:33 batman can't you do tcp:// and file:/// instead of postgres:// ?
13:34 sri now things get ugly
13:34 batman tcp=pg=postgres
13:34 sri the whole DBD::Pg way is also documented well
13:35 nicomen is it DSN you are looking for?
13:36 batman "pg way" = dsn + \%attribtutes?
13:36 sri DBI way
13:36 batman what about unix:///tmp/socket?attr=foo ?
13:37 batman ok. i see your reasoning. it makes sense. i just hope it didnt ;D
13:38 sri that doesn't say postgres://, it's ugly
13:38 sri btw. https://metacpan.org/pod/DBD::Pg#connect
13:40 sri batman: how do you set a domain socket and database name?
13:40 dp_ joined #mojo
13:40 batman i'm also not sure abut $pg->options() maybe you want to just set defaults that is required for Mojo::Pg to work? most people screw this up :(
13:40 sri your postgres:// uri earlier used /db982398
13:40 batman hm... *ponders*
13:41 sri umm
13:41 batman maybe not set it as the path part, but rather as a qury param?
13:41 sri did you actually look into the module?
13:41 batman unix:///file?db=db982398 ?
13:41 ignacio_ joined #mojo
13:41 batman sorry for changing the subject. i thought the uri discussion was dead :)
13:41 sri https://github.com/kraih/mojo-pg/blob/master/lib/Mojo/Pg.pm#L9
13:41 batman what are we discussing now?
13:42 chansen lol
13:42 batman sri: yeah, i saw the defaults. i like 'em, but i think they should be private and not exposed :)
13:44 batman even so... i guess there are a lot of other options that should be exposed :/
13:45 batman what abotu "pg_enable_utf8" as default on?
13:45 batman sorry.
13:47 batman sri: would you take a pull request that turns and URL into dsn+username+password attributes?
13:47 sri doubt it
13:47 batman ok, hehe
13:48 sri seriously, if it's not pretty it's not going into the module
13:48 batman Mojo::Pg->new('pg://user:pass@server'); # looks good from the outside :)
13:49 sri only if you ignore half of the facts
13:50 sri make it look good for unix domain sockets with username and database name as well as host/port variants
13:52 batman ok, so it's an all-or-nothing pull request then i guess
13:57 sri if it was all or nothing i would just say no
13:57 sri as you can see in the connect docs, there is even more that can be in the DSN
13:58 sri i want the common use cases covered by what we expose as the default way of configuring the database
13:58 batman yeah, it makes sense.
13:58 batman thansk sri
14:00 sri btw. http://www.postgresql.org/docs/9.2/static/libpq-connect.html#LIBPQ-CONNSTRING
14:04 batman is it possible to connect without specifying a database?
14:04 batman does it make sense..?
14:07 batman looks like they have taken care of the socket part: postgresql:///mydb?host=/path/to/socket
14:08 batman "Note, however, that the slash is a reserved character in the hierarchical part of the URI. So, to specify a non-standard Unix-domain socket directory, either omit the host specification in the URI and specify the host as a parameter, or percent-encode the path in the host component of the URI: postgresql:///dbname?host=/var/lib/postgresql"
14:08 batman or even postgresql://%2Fvar%2Flib%2Fpostgresql/dbname
14:08 batman haha! #ugly
14:11 dp_ joined #mojo
14:16 genio did you just.... hashtag something?
14:16 * genio feels all ewwwww inside for having said "hashtag"
14:17 chansen batman: you should read RFC3986, especially the section about authority
14:18 batman genio: yeah. my brain has support for hashtags ;)
14:18 batman chansen: about what?
14:18 chansen batman: you probably need to real the whole RFC ;)
14:19 genio sri++   # Mojo::Pg   and signatures in the example
14:19 batman sounds like you know something i don't know. so why not just do the simple solution: tell me.
14:19 batman chansen: ^
14:20 genio who knew that    use experimental 'signatures';     could make me so happy.
14:21 batman hehe
14:21 woz joined #mojo
14:23 jberger_ those aren't hashtags those are comments
14:24 dabudabu And here I thought that perl programmers were the greatest Twitter fans ever
14:24 cfedde perl fans are a pretty diverse group.
14:25 genio I have yet to post a single twitter thingy
14:25 chansen batman: You are trying to design a URI scheme without knowing the basics of a URI. If you read the spec you will gain knowledge about URI's.
14:25 batman chansen: http://www.postgresql.org/docs/9.2/static/libpq-connect.html#LIBPQ-CONNSTRING <--- i got it from here
14:25 sri hahaha... i was just testing the %2ffoo%2fbar unix domain socket path stuff... and kept thinking "but what about windows paths?" -.-
14:26 genio sri needs coffee
14:27 batman sri++
14:34 chansen postgresql:///path/to/unix/socket
14:34 chansen postgresql:///c:/path/to/windows/named/pipe
14:35 sri that's actually wrong
14:35 sri postgresql://%2fc:%2fcpath%2fcto%2fcwindows%2fcnamed%2fcpipe
14:36 dod joined #mojo
14:39 chansen sri: wrong for the new scheme in Mojo or wrong for the URI in libpq?
14:40 sri both are the same
14:41 chansen No, the only sensible way to indicate a local resource is to have a empty authority component
14:41 chansen .. if you are using the same scheme for both local and remote resources
14:42 sri you might want to bring that up on #postgresql ;p
14:44 chansen I'm talking about URI's, not PostgreSQL's implementation. Insane implementations can be found everywhere, if I would bring up every insane implementation I have seen I would be busy typing emails instead of chatting in this channel
14:45 D4RK-PH0ENiX joined #mojo
14:48 batman isn't that what you're doing..?
14:50 chansen batman: what am I doing?
14:50 batman bring up insane implementation you have seen
14:50 * batman have to stop.
14:51 * batman gets another beer and cheers up
14:52 chansen batman: you probably had to many already ;oP
14:53 batman i don't feel all *cheery*
14:53 batman what's that supposed to mean anyway, chansen?
14:56 sri allright, stop it you two
14:56 chansen batman: I suggest you read the first sentence in my message before you ask "isn't that what you're doing..?"
14:56 batman i don't get it.
14:57 basiliscos joined #mojo
14:57 batman you don't have to convince me. i trust sri.
14:59 chansen Why would I try to convince you? You haven't even read the spec about URI's?
15:00 Oleg left #mojo
15:06 D4RK-PH0ENiX joined #mojo
15:07 TitanOfOld joined #mojo
15:09 sri done https://github.com/kraih/mojo-pg/commit/1b4d1a72c21a2bdccfd51699e49945e91ec06aaf
15:15 woz joined #mojo
15:17 TitanOfOld sri, ooh, nice! Mojo understands the PostgreSQL protocol.
15:18 batman sri: awesomelicious :)
15:19 sri and for those that want old school DBI config there's the attributes
15:19 sri Mojo::Pg->new->dsn('dbi:Pg:dbname=test')->username('sri');
15:20 batman sri: what about: sub new { @_ == 2 1 ? shift->SUPER::new->from_string(@_) : shift->SUPER::new(@_) } ?
15:20 TitanOfOld sri, why DBD::Pg instead of DBIx::Connector?
15:20 batman oops
15:20 batman sri: what about: sub new { @_ == 2 ? shift->SUPER::new->from_string(@_) : shift->SUPER::new(@_) } ?
15:21 batman TitanOfOld: why would you want DBIx::Connector?
15:21 batman TitanOfOld: it use DBI. it's blocking. it doesn't help you.
15:21 batman TitanOfOld: Mojo::Pg does ping(), so you're home free
15:23 TitanOfOld Yes, but does it reconnect automatically?
15:24 batman you never connect. it handles it for you. https://metacpan.org/source/SRI/Mojo-Pg-0.01/lib/Mojo/Pg.pm#L29
15:30 thowe joined #mojo
15:32 chansen perhaps he wonders how errors propagate from the server to the client?
15:33 sri oh, i'm almost certain there are bugs in the connection management, please send patches if you're better at it than me
15:34 sri especially the fork stuff is funky, atm. i can only clean up inactive handles in the queue, not active ones that are in use at the time of the fork
15:38 woz joined #mojo
15:40 TitanOfOld Well, there's the errors, too, but that's kind of a seperate issue.
15:40 TitanOfOld I'm not seeing how the errors percolate up.
15:40 sri although, i kinda doubt DBIx::Connector would be of much help
15:41 TitanOfOld sri, I'm only mentionig DBIx::Connector for its connection management. I kind of hate how I have to send queries to Pg through it.
15:42 TitanOfOld Good looking code quickly looks ugly when using it.
15:42 sri TitanOfOld: are you just mentioning it because you've learned that it's best practice to use it, or because you're referring to a specific feature Mojo::Pg is missing?
15:42 basic6 joined #mojo
15:43 sri Mojo::Pg has to manage a poll of handles, so i'm not sure how DBIx::Connector relates
15:43 sri s/poll/pool/
15:44 TitanOfOld sri, well, it's looking like Mojo::Pg pings every time a query is made.
15:44 TitanOfOld And I'm having trouble seeing if it reconnects, or if i have to manage the reconnect.
15:46 TitanOfOld And I'm mentioning it because i think it's a specific feature that Mojo::Pg is missing. Not because I recently learned about it. (^_^)
15:50 sri i don't see it
15:50 sri when exactly does DBIx::Connector reconnect and Mojo::Pg doesn't?
15:58 OlegG joined #mojo
16:00 OlegG btw, explanation about threaded perl library VS non threaded perl: http://stackoverflow.com/questions/13587325/threads-vs-pthread-in-perl
16:03 OlegG and looks like this is not perl specific, i made example with plain C which segfaults at least on my Debian 7: https://github.com/olegwtf/sandbox/tree/master/threaded-so-for-non-threaded-app
16:04 OlegG but if i'll make LD_PRELOAD=/lib/i386-linux-gnu/i686/cmov/libpthread.so.0 ./main segfault goes away :)
16:05 OlegG sri: interesting to know is this C example segfaults on OS X
16:10 ua_ joined #mojo
16:10 chansen OlegG: are you targeting Windows as a deployment OS?
16:12 OlegG deployment for what?
16:14 OlegG windows is bad for any deployment, but somebody uses it :)
16:17 chansen OlegG: true! Please forgive my question, I now recall the issue with your DNS module.
16:18 woz joined #mojo
16:18 OlegG hmm, convos.by ate my "Oleg" nick. Now i can't use it with other irc client. It says: "choose other"
16:19 TitanOfOld sri, https://metacpan.org/pod/DBIx::Connector#Connection-Modes
16:19 sri TitanOfOld: so you want run(), txn() and svp() methods in Mojo::Pg?
16:20 TitanOfOld Specifically fixup.
16:20 TitanOfOld Not so much the methods.
16:21 sri TitanOfOld: please be more specific
16:21 purl If you'd explained that ahead of time, you wouldn't have received an inappropriate answer.
16:22 sri TitanOfOld: are you sure you really know how DBIx::Connector works?
16:22 sri so far i've been assuming i got something wrong
16:22 sri but i'm having doubts now
16:22 basiliscos joined #mojo
16:23 chansen OlegG: Have you used IO::AIO? It aslo uses threads behind the implementation (without require that per is compiled with thread support), I have used it on several posix platforms without problems
16:23 sri as far as i understand it, once you call $connector->dbh you're responsibe for the handle, there is no magic to hold your hand
16:24 OlegG chansen: Yes, we using it in our current project. But we have threaded perl :) Hmm, i should check its source code. Thanks
16:24 TitanOfOld sri, yeah, that's what I'm getting at. If the SQL fails to execute, it checks the connection (ping) at that point.
16:25 sri TitanOfOld: where does it do that?
16:25 TitanOfOld sri, but I could be reading the code wrong. Is Mojo::Pg pinging the DB with every query?
16:26 preaction joined #mojo
16:27 TitanOfOld sri, "in the event the block throws an exception, if DBIx::Connector finds that the database handle is no longer connected, it will reconnect to the database and re-execute the block."
16:27 sri ok, you've had my attention for quite some time now, but i'm afraid i have to assume you got something wrong now
16:27 Adura joined #mojo
16:27 sri please make a test case if there's really something wrong
16:27 TitanOfOld sri, you are almost certainly correct.
16:32 sri there are actually pg_* functions you can use to trigger a connection close and stuff, pretty nice http://www.postgresql.org/docs/current/static/functions-admin.html#FUNCTIONS-ADMIN-SIGNAL-TABLE
16:33 OlegG chansen: Ha, that's how it: https://metacpan.org/source/MLEHMANN/IO-AIO-4.31/Makefile.PL#L45
16:37 sri OlegG: that example works flawless on os x btw.
16:38 OlegG sri: cool, looks like OS X has good libc implementation :)
16:38 Kripton joined #mojo
16:38 sri hahahaha... i think you're like the first to ever say that
16:39 chansen OlegG: Glad you found your issue =)
16:41 sujithm joined #mojo
16:43 chansen I have deployed IO::AIO on two different sites using Solaris and Linux without any problems, and myself prefer to develop on darwin (also use there without any issues, albeit only for development)
16:45 chansen OlegG: btw, a big part of the OS X (darwin) libc is based on FreeBSD ;)
16:46 chansen <http://opensource.apple.com/source/Libc/Libc-997.90.3/>
16:47 OlegG I see. Often they have same bugs
16:49 Akron joined #mojo
16:50 OlegG The big surprise was when i realised that posix unnamed semaphores are not implemented in os x
16:51 OlegG So, more IFDEFS came in https://github.com/olegwtf/p5-Net-DNS-Native/blob/master/Native.xs#L24
16:53 OlegG this is how i start hating os x :)
16:54 cpan_mojo Mojo-Pg 0.02 by Sebastian Riedel - http://metacpan.org/release/SRI/Mojo-Pg-0.02
16:54 kogurr joined #mojo
16:59 ryozi joined #mojo
17:00 woz joined #mojo
17:13 jamesaxl joined #mojo
17:19 chansen OlegG: OS X, have had it's share of surprises, even though it's been (and still is a) qualified as a "UNIX" certified OS according to Open Group
17:20 KCL joined #mojo
17:21 disputin joined #mojo
17:30 neyasov joined #mojo
18:02 disputin joined #mojo
18:07 dod joined #mojo
18:10 woz joined #mojo
19:02 jamesaxl joined #mojo
19:11 woz joined #mojo
19:13 irq joined #mojo
19:17 irq sri: Thanks for Mojo::Pg. But seems it misses possibility of having "dbi:Pg:service=$service"  as a $dsn
19:17 fhelmber_ joined #mojo
19:17 sri irq: http://mojolicio.us/perldoc/Mojo/Pg#dsn
19:19 irq so I am not sure what should be argumen to new() ?
19:20 berov joined #mojo
19:26 sri irq: http://mojolicio.us/perldoc/Mojo/Pg#new
19:26 * berov Sometimes needs to dump perl objects in templates as JavaScript objects but Mojo::JSON::encode_json encodes to UTF8 and as a result garbage
19:28 * berov wishes to write a helper     $app->helper(ref2js =>sub { Mojo::JSON::_encode_value($_[1]) }); but sri will say THIS IS A PRIVATE METHOD
19:29 sri berov: THIS IS A PRIVATE METHOD
19:29 berov sry: ?any recommendation?
19:30 berov for converting perl objects to js objects easily?
19:30 berov in templates
19:30 sri you'll laugh, every now and then i wanted to encode to JSON without the UTF-8 encoding too, but i couldn't think of a good function name
19:31 irq sri: so needed to use a dummy $pg = Mojo::Pg->new("postgresql://foo/bar") to be able then to $pg->dsn("dbi:Pg::service=test"), is that right?
19:31 sri irq: WTF?
19:32 berov sri: and I wanted to propose  to have a new method in Mojo::JSON for such purposes :)
19:33 sri berov: so, i'm open to the idea of adding a new public function, if it's consistent (there needs to be a way to decode non-UTF-8 encoded JSON too) and you can come up with a good name
19:34 * berov starts thinking hard....
19:34 sri irq: is that really what you got from the docs i linked to?
19:35 irq sri: from metacpan, actually, isn't that the same?
19:35 sri berov: next problem, it also needs a good description for the docs... "non-UTF-8 encoded JSON" is shitty
19:36 berov aha,ok will suggest something later
19:36 berov but "non-UTF-8 encoded JSON"  is so clear :D
19:42 zgarnog joined #mojo
19:42 basic6_ joined #mojo
19:46 * berov starts experimenting
20:13 sri irq: maybe you shouldn't be using Mojo::Pg, it's highly experimental
20:13 woz joined #mojo
20:14 marty joined #mojo
20:14 irq not using, just playing with what I have been waiting for
20:16 irq there is https://metacpan.org/pod/Mojolicious::Plugin::PgAsync, but buggy
20:19 disputin joined #mojo
20:22 basiliscos joined #mojo
20:47 neyasov joined #mojo
20:48 kogurr joined #mojo
20:54 sri berov: best i can think of is Mojo::JSON::encode_json_text and relaxing Mojo::JSON::decode_json to allow wide characters
20:54 sri berov: if you want to make a formal proposal you should try getting one or two core devs on your side
20:59 Akron joined #mojo
21:02 genio ugh... DBD::mysql over SSL is soooooooooo slow
21:09 basiliscos joined #mojo
21:09 basiliscos joined #mojo
21:12 basiliscos joined #mojo
21:14 woz joined #mojo
21:33 berov sri,jberger,marcus,batman I was thinking about decode_javascript and encode_javascript
21:35 berov amd in the documentation mentioning what they can be used for with an example. we will need also eventually helpers
21:36 berov and*
21:43 berov but I get some stranje error when using in my app encode_javascript |"encode_javascript" is not exported by the Mojo::JSON module|
21:43 berov strange*
21:44 berov and I have ot in @EXPORT_OK
21:44 berov s/o/i/
21:45 berov when using on the command line with > mojo eval it is fine?!?!
21:54 berov and also use Mojo::JSON qw(encode_javascript);
22:15 woz joined #mojo
22:44 kogurr joined #mojo
23:17 woz joined #mojo
23:18 fhelmber_ joined #mojo
23:25 cpan_mojo Minion 0.38 by Sebastian Riedel - http://metacpan.org/release/SRI/Minion-0.38 (depends on Mojolicious)
23:31 * sri has killed Minion::Backend::Mango
23:32 sri turned out to be a really good thing some of you convinced me to do pluggable backends, or Minion as a whole would be dead now :)
23:53 firnsy \o/
23:59 * berov RFC sri,  jberger, marcus, tempire, batman ...: Relaxed json encoding and decoding: http://git.io/UA_J2g

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