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

IRC log for #mojo, 2015-01-09

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

All times shown according to UTC.

Time Nick Message
00:04 good_news_everyon joined #mojo
00:04 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/M_dOTA
00:04 good_news_everyon mojo/master 2628b2a Sebastian Riedel: improved Mojo::IOLoop::Client not to start timers before the event loop is running
00:04 good_news_everyon left #mojo
00:10 rem_lex|pivo joined #mojo
00:11 dotan1 joined #mojo
00:11 lb_ joined #mojo
00:11 bpmedley_ joined #mojo
00:11 firnsy_ joined #mojo
00:12 graf joined #mojo
00:12 Snelius joined #mojo
00:12 Gedge_ joined #mojo
00:12 HtbaaPi joined #mojo
00:12 o3u joined #mojo
00:12 Andreas2 joined #mojo
00:12 jmmills joined #mojo
00:12 psandin joined #mojo
00:12 ashleydev joined #mojo
00:13 * sri wonders what's going on
00:14 Insane joined #mojo
00:14 Kundun joined #mojo
00:14 Kharec joined #mojo
00:14 romel joined #mojo
00:14 Trelane joined #mojo
00:14 Rallias joined #mojo
00:14 Grinnz_ O_O
00:14 marty joined #mojo
00:14 skittles_ joined #mojo
00:14 yusukebe joined #mojo
00:14 absolut_todd joined #mojo
00:14 elsanchez joined #mojo
00:14 herl joined #mojo
00:14 d4rkie joined #mojo
00:14 linagee joined #mojo
00:14 mgrimes joined #mojo
00:14 Fistful_of_Coins joined #mojo
00:14 Grinnz_ i think something died
00:14 omega joined #mojo
00:14 jnbek joined #mojo
00:14 jkramer joined #mojo
00:15 mmp joined #mojo
00:15 davido___ joined #mojo
00:15 Foxcool joined #mojo
00:15 Nemix__ joined #mojo
00:15 jb360 joined #mojo
00:15 jonquest joined #mojo
00:15 Dandre joined #mojo
00:16 cpan_mojo joined #mojo
00:16 xtetsuji joined #mojo
00:16 doby joined #mojo
00:18 jeff joined #mojo
00:18 rofl_ joined #mojo
00:18 klapperl joined #mojo
00:18 panzana` joined #mojo
00:18 dabudabu joined #mojo
00:18 cfedde joined #mojo
00:18 murre joined #mojo
00:18 michael joined #mojo
00:19 memowe joined #mojo
00:19 stryx` joined #mojo
00:19 disco joined #mojo
00:19 btyler joined #mojo
00:19 hernan604 joined #mojo
00:24 sivoais joined #mojo
00:27 marcus joined #mojo
00:42 good_news_everyon joined #mojo
00:42 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/XUPrtw
00:42 good_news_everyon mojo/master 8aefa99 Sebastian Riedel: improved Mojo::Reactor::EV to update the current time before starting a timer if EV is not running already
00:42 good_news_everyon left #mojo
00:43 asarch joined #mojo
00:54 mtj- joined #mojo
01:02 Grinnz found the problem; the tests need to set $irc->server("127.0.0.1:$port") not $irc->server("localhost:$port") now
01:03 Grinnz that should still work on old versions before you changed the localhost workaround thing right sri?
01:06 Grinnz now why localhost doesn't work... no idea
01:15 Nemix__ maybe this: https://groups.google.com/forum/#!topic/mojolicious/ZUg2yaZLNmU
01:16 Grinnz yes thats what i was referring to... but connecting to localhost should still work shouldn't it?
01:18 good_news_everyon joined #mojo
01:18 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/5PmGng
01:18 good_news_everyon mojo/master 47b67d8 Sebastian Riedel: use the current time consistently for timers
01:18 good_news_everyon left #mojo
01:18 sri oh well, it's a little slower but consistent
01:22 Nemix__ Grinnz:  sri is on som I am going to withhold my commentary.  it would probably only serve to expose my ignorance.
01:23 Grinnz i can connect to localhost with ojo..
01:26 Grinnz all Mojo::IRC is doing is passing "localhost" to the "address" arg for Mojo::IOLoop->client
01:29 zivester joined #mojo
01:34 good_news_everyon joined #mojo
01:34 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/q44F5A
01:34 good_news_everyon mojo/master afd9f37 Sebastian Riedel: next_tick can be a little faster
01:34 good_news_everyon left #mojo
02:39 klapperl_ joined #mojo
03:26 hshong joined #mojo
03:29 Insane joined #mojo
03:32 noganex joined #mojo
04:16 good_news_everyon joined #mojo
04:16 good_news_everyon [mojo] kraih deleted pluggable_resolver at 49acb93: http://git.io/k_OJTg
04:16 good_news_everyon left #mojo
04:25 good_news_everyon joined #mojo
04:25 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/2PlHSA
04:25 good_news_everyon mojo/master cfa3988 Sebastian Riedel: mention version command more often
04:25 good_news_everyon left #mojo
04:30 sri hmm, that does not look as good as i had hoped :/ http://mojolicio.us/perldoc/Mojo/IOLoop#DESCRIPTION
04:55 good_news_everyon joined #mojo
04:55 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/rXE4vA
04:55 good_news_everyon mojo/master c9cd59d Sebastian Riedel: mention version command only in FAQ
04:55 good_news_everyon left #mojo
05:05 irq joined #mojo
05:27 sh4 joined #mojo
05:37 good_news_everyon joined #mojo
05:37 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/g9NO-w
05:37 good_news_everyon mojo/master a78a582 Sebastian Riedel: do not mention version command in FAQ answer
05:37 good_news_everyon left #mojo
05:38 sri commands without arguments just don't look nice in examples
05:59 rem_lex|pivo joined #mojo
06:21 rem_lex joined #mojo
06:40 scrappy joined #mojo
06:40 scrappy Does Mojolicious not support the following: find('a.class:not(.otherclass)') ?
06:41 scrappy It's definitely not working as expected :/
06:41 scrappy tmux
06:46 scrappy exit
06:53 basiliscos joined #mojo
07:13 Eddy joined #mojo
07:37 salparadise man so simple yet so powerful http://mojolicio.us/perldoc/ojo  thanks for making my life easier sri
07:50 basiliscos joined #mojo
08:04 eseyman joined #mojo
08:07 dotandimet joined #mojo
08:09 Yelu joined #mojo
08:09 dod joined #mojo
08:11 basiliscos joined #mojo
08:17 Vandal joined #mojo
08:36 trone joined #mojo
08:39 Yelu joined #mojo
08:57 irq joined #mojo
09:09 batman Grinnz: thanks! could you try to install mojo-irc on 5.10.1?
09:09 batman cpanm https://github.com/jhthorsen/mojo-irc/archive/master.tar.gz
09:10 jkramer left #mojo
09:11 batman Grinnz: oh! sorry. didn't see your pr on github :/
09:12 shorty_mu joined #mojo
09:28 denny joined #mojo
09:47 batman Grinnz: i merged almost all your pull requests. let me know if you can test again before i make a new release.
09:48 batman Grinnz++ # making my life easier
10:15 mtths joined #mojo
10:39 amon joined #mojo
10:50 batman not related to mojo, but... anyone know of a module that does this: https://gist.github.com/jhthorsen/ab99593db5bcb77c3dbd ?
10:50 batman mst: got ideas? ^
10:57 mtths joined #mojo
11:01 eseyman joined #mojo
11:07 dotandimet joined #mojo
11:17 denis_boyun joined #mojo
11:18 nicomen batman: by "does this", you mean: add a hash to a list of hashes, except if a hash with the same key names exists, in that case, replace/update the given hash?
11:19 nicomen (although I don't understand the last @keys parameter in that code
11:19 batman i mean a cpan module that has already implemented the logic i described in the gist
11:20 batman the @keys tell the function which keys make up the unique part to figure out what to compare
11:20 nicomen I tried to confirm I understood your description
11:20 nicomen ah
11:20 batman it doesn't add anything. it returns update or add $data
11:21 nicomen ah right
11:22 nicomen http://search.cpan.org/~bolav/Hash-Diff-0.005/lib/Hash/Diff.pm <-- maybe this, but use the @keys as a hash slice after getting the diff back?
11:23 batman nicomen: why aren't you using metacpan?
11:24 batman nicomen: that module does not compare against a list of possible matches
11:25 nicomen yeah you are right
11:25 nicomen batman: because google ;-/
11:29 batman hehe
11:30 batman nicomen: https://metacpan.org/about/faq#cani_automatically_redirectlinkspointingatsearch.cpan.orgtometacpan.org
11:34 nicomen heh
12:09 batman seen this sri? https://pbs.twimg.com/tweet_video/B6rzTYfCYAAMixU.mp4 :)
12:12 batman hm... i should have linked to the tweet :/
12:56 neilhwatson joined #mojo
13:07 dp_ joined #mojo
13:18 human39 joined #mojo
13:34 elsanchez joined #mojo
13:37 Insane joined #mojo
13:47 basiliscos joined #mojo
13:53 sh4 joined #mojo
14:14 asarch joined #mojo
14:18 jberger batman: I'm not sure what you are trying to do, but if you are trying to compare data structures in a fuzzy look at Test::Deep
14:19 jberger in a fuzzy way
14:19 batman thanks. my function already works, but i'm not very good at this, so i was hoping this was already on cpan
14:20 jberger ok, I just wasn't sure what problem you were trying to solve
14:20 batman i wonder how to describe it... guess i need to add some documentation ;)
14:23 zivester joined #mojo
14:35 ryozi joined #mojo
14:41 marty joined #mojo
15:14 Grinnz batman, i tested setting the server connections back to localhost and t/register.t hangs again, i think the timer isn't set correctly
15:14 Grinnz otherwise, it works ok
15:14 batman oops! thanks :)
15:15 batman Grinnz: fixed now. (and pushed to master)
15:18 batman Grinnz: i greatly appreciate if you could try to make it fail again
15:18 Grinnz batman, does not hang now, including the tls.t
15:19 batman thank you!
15:19 Grinnz however now i'm wondering why tls.t is failing...
15:19 batman crap. 0.15 is already on way to cpan, hehe
15:19 Grinnz seems no name resolution is working on this machine?
15:20 Grinnz but it works when i try with ojo
15:20 cpan_mojo Mojo-IRC 0.15 by Jan Henning Thorsen - http://metacpan.org/release/JHTHORSEN/Mojo-IRC-0.15
15:20 batman i don't know... good thing it's skipped by default :P
15:21 batman weird... when did cpan_mojo come alive again?
15:21 shorty_mu left #mojo
15:21 Grinnz i think this is a problem with either my machine or mojo client... not mojo irc
15:22 Grinnz ill have to test more
15:23 batman it's fine by me. i'm happy we sorted out the localhost issue
15:23 batman hope cpantesters agree...
15:46 * genio wants to kill EL7 and systemd
15:46 jberger batman: maybe that means your host is un-banned again?
15:47 * jberger missed cpan_mojo
15:47 batman ah!
15:47 batman jberger++
15:47 jberger also might mean you want to disconnect demo from it, just in case
15:47 jberger don't want to get re-banned
15:47 batman jberger: going to move cpan_mojo to dogfood
15:48 jberger dogfood?
15:48 batman yeah. another server, just for invites
15:48 jberger ah ok
15:49 batman oh. it's already running from my personal server.
15:49 batman not going to move it then.
15:50 batman personal = not personal, but a completely different server
15:52 batman but you're right: not banned anymore
15:52 * batman adds iptables rule to demo.convos.by
15:53 genio actually, my ire should be in the direction of nginx.  :/
15:53 nicomen 16:48 < sandsmark> <PovAddict> "djangocon europe has been rescheduled because there are literally no free hotels in the entirety of Cardiff for the last weekend of it, because One Direction suddenly went "let's  have two shows in cardiff" and several thousand hotel rooms disappeared just like that"
15:53 batman genio: ire?
15:54 genio "anger"
15:55 nicomen at least they knew mojoconf is normally held in may in oslo, so they waited till june 19th before they come here
16:12 sri http://isperlalive.com/ # good to know!
16:24 batman haha :)
16:28 Grinnz_ genio: what did nginx do? :)
16:28 Insane joined #mojo
16:30 nicomen sri: I like that you can subscribe to it: http://isperlalive.com/rss.xml
16:31 sri :D
16:31 Grinnz_ haha
16:31 nicomen first published in 2008 it seems
16:32 genio Grinnz_: They disagree with me slightly on a philosophical point with SSL keys being encrypted with a passphrase
16:32 Grinnz_ ah... yeah a lot of things don't like that
16:33 Grinnz_ it's kind of impossible to work with them without putting that passphrase in a plaintext config somewhere though, so it doesn't really help
16:34 sri don't most certs get stolen from memory anyway? (hello heartbleed!)
16:34 Grinnz_ heh
16:35 Insane hey all, did you ever had any frontend work with phone number inputs?)
16:36 Grinnz_ sure? it's just a string
16:37 Insane from bakend side - yes, it's just a string, but from frontend side it might be more
16:37 Grinnz_ meaning?
16:37 purl i guess meaning is in the individual, whether a definition of a word, connotation of a symbol, or impression of a painting.
16:37 * sri pats purl
16:37 * purl stabs
16:37 sri :(
16:41 nicomen Insane: ?
16:41 * genio pushes purl down the stairs
16:41 * purl attaches herself to genio using suction and a ham straw
16:42 nicomen Insane: <input type='tel' pattern='[\+]\d{2}[\(]\d{2}[\)]\d{4}[\-]\d{4}' title='Phone Number (Format: +99(99)9999-9999)'>
16:43 Eddy joined #mojo
16:43 Insane it will force user to enter +()-, I wish to have something to do is for user
16:43 Insane so user just have to print numbers
16:43 nicomen true
16:43 Grinnz_ Insane: that's up to the javascript
16:44 nicomen I guess the most common thing is to have a pull-down for country code, then free-form for the phone number
16:44 nicomen unless you know every possible format for every country (you won't I'm quite sure), I think having a text input is just fine
16:45 jberger that "isperlalive" is awesome!
16:46 jberger would maybe be better if the feed had regular updates
16:46 jberger but I get the point :-)
16:51 punter joined #mojo
17:09 Insane wow... here is what comes to mojo as http post (as ngrep reports): json={"name":"Test user","id":"72","mobile":["+7 (123) 456-78-90"]}, and here is what comes into $c->req->param('json'): { id 72, mobile [ " 7 (123) 456-78-90" ], name "Test user" }... where is my plus character!?
17:11 Grinnz_ Insane: how are you posting it?
17:11 Insane javascript does http post query
17:12 Grinnz_ javascript should encode it as needed...
17:15 Grinnz_ hopefully that is not verbatim what you get, as that's not actually valid JSON
17:18 Insane {"name":"Test user","id":"72","mobile":["+7 (123) 456-78-90"]} is a valid json
17:19 Grinnz_ right, what you got from the param is not
17:19 Insane javascript creates json object, writes fields and post to mojo
17:19 Grinnz_ or what you wrote there anyway
17:23 Insane here is the code and its log output: http://pastebin.com/cTDm5Nt1
17:23 buu Insane: So.. properly url encode the plus sign
17:25 Grinnz_ javascript should be doing this
17:25 Grinnz_ also why are you using from_json? should be decode_json
17:29 edenc joined #mojo
17:50 * sri gets more and more sceptical of native http/2 support in mojolicious
17:53 punter If nginx supports HTTP/2, and translates queries into HTTP/1 to Mojolicious, isn't that enough support?
17:53 sri not on its own
17:53 sri you also want to benefit from server push
17:54 sri *but* there could be nginx extensions for that, like a cache learning the relationships between resources, and then pushing them automatically
17:54 DaTa joined #mojo
17:54 batman sounds like they spent too much time on this http2 thing, and now they just have to ship it, even when they know it's no good :(
17:55 sri in the end http/2 might really end up as a protocol only used by reverse proxies and browsers
17:55 sri as an optimization
17:55 sri and we just keep using http/1 for everything else
17:56 sri i don't see Mojo::UserAgent benefit much from http/2 for example
17:56 sri since you don't need to load more resources from the same host while you're parsing/rendering the html
17:57 sri only reason i see would be http/2 only hosts
17:58 sri one thing is certain, http/1 is never going away
17:59 sri you'll never see http/2 being used for applicances for example (like network printers, wireless routers, washing machines...)
17:59 sri who would update the certs regularly?
18:11 ribasushi batman: this is what happened with ipv6 fwiw
18:11 ribasushi "got to ship, crap or not"
18:11 batman ribasushi: yeah, i've been thinking the same thing.
18:11 batman ipv6 is pretty outdated, before adopted :(
18:12 ribasushi it's being adopted due to address space necessity, not because of any other (anti)feature it brought to the table
18:12 ribasushi basically if you think http2 is bad, you haven't looked in depth at ipv6
18:14 ribasushi one could put on the tinfoil-realist hat and argue that ipv6 was designed to be terrible to facilitate the upgrade treadmill (you can't do it all in userland software, context switching is too slow for core routing, hence lots of $$$ for hardware)
18:14 ribasushi for http2 - I am not sure what the "economic reasoning" is
18:15 mst "nobody can get consensus so let's just rubber stamp SPDY with tweaks"
18:15 sri this
18:17 ribasushi mst: there is a lot more crack in http2 compared to spdy though...
18:17 sri they've changed SPDY just enough so everybody can feel like they've contributed something
18:20 sri and SPDY keeps getting quick updates
18:20 sri gonna be fun once someone writes a big blog post about how SPDY/7 or whatever is so much more efficient than http/2 at everything
18:21 mst sri: maybe they should rename it to ALOT
18:21 Grinnz_ what is SPDY?
18:21 purl SPDY is not used that way
18:21 Grinnz_ thanks, purl.
18:21 sri http://en.wikipedia.org/wiki/SPDY
18:22 Grinnz_ ah
18:23 Grinnz_ TLS is required? thats gonna be fun
18:24 cfedde It's about time.
18:28 disputin joined #mojo
18:38 cpan_mojo Swagger2 0.22 by Jan Henning Thorsen - http://metacpan.org/release/JHTHORSEN/Swagger2-0.22 (depends on Mojolicious)
18:39 Grinnz_ Swagger just makes me think of the pokemon move
18:39 Grinnz_ is that weird
18:41 batman :)
18:41 batman i guess i'm the guy with the funny module names...
18:41 batman (i've deleted mypp from cpan)
18:42 Grinnz_ jberger: following from the @CARP_NOT stuff you were looking into before.. i'm trying to make someone else's module not shit all over my logs when it calls cluck
18:42 Grinnz_ but not having much luck...
18:43 dod joined #mojo
18:44 Grinnz_ even if i set @TheirModule::CARP_NOT = qw/Mojo::Base/ i still see the whole mojo stack
18:46 genio Actually, I think my philosophical problem lies with systemd and not so much on nginx.
18:58 disputin joined #mojo
19:10 disputin joined #mojo
19:32 Grinnz_ this is rather frustrating... @CARP_NOT does not seem to be working at all
19:32 Grinnz_ i set it in the offending module itself and it still spits it all out
20:07 jberger Grinnz_: where does their cluck emit from?
20:12 Grinnz_ functions in that module
20:16 Eke- joined #mojo
20:19 Insane joined #mojo
20:23 denis_boyun joined #mojo
20:24 neilhwatson joined #mojo
20:29 Grinnz_ hmm... i guess if i want to change the log path in a running app i have to create a new Mojo::Log, the log->handle is already set from the path the first time it's used
20:43 batman Grinnz_: i do delete $log->{handle};
20:44 batman i really would like $log->path() would delete it for me though.
20:45 mst package My::Mojo::Log; use Moo; extends 'Mojo::Log'; after path => sub { delete $_[0]->{handle} if @_ > 1 };
20:52 Grinnz_ batman: ah! that works, thanks :)
20:52 Insane joined #mojo
20:52 mst this is where the trigger/clearer pairing in M* comes in handy
20:53 mst has handle => (is => 'lazy', builder => sub { ... }, clearer => '_clear_handle');
20:53 Grinnz_ it would be nice to be default behavior, i don't see any reason someone would set path and not want to reset the handle
20:53 mst has path => (is => 'rw', trigger => sub { shift->_clear_handle });
20:53 mst a Mojo implementation would, I guess, just be doing that the old-fashioned way
20:58 denis_boyun_ joined #mojo
21:00 jberger Grinnz_: that's not what I mean, I'm asking where does it say "at <<file>> <<line>>"
21:01 Grinnz_ where the cluck happens
21:03 Grinnz_ anyway i gave up on that.. i'm just going to dump all of this module's logging into another file
21:06 basiliscos joined #mojo
21:09 jberger Grinnz_: that doesn't make sense
21:09 purl I think you mean "I don't understand"
21:10 jberger it shouldn't cluck from the location that the "cluck" occurs should it?
21:10 jberger it should occur at least from the caller
21:10 jberger nm
21:11 mst purl++
21:11 Grinnz_ cluck is the whole stack
21:12 jberger oh, well then why do you expect that CARP_NOT is going to do anything :-)
21:12 Grinnz_ because it says it affects cluck :P
21:12 Grinnz_ docs--
21:12 mst I think CARP_NOT is stripping initial frames
21:12 mst not skipping anything later
21:13 Grinnz_ hmm, that could be
21:13 jberger the only thing that CARP_NOT is going to do is make the location of the emit different, its still going to cluck
21:13 Grinnz_ in which case it's useless for this case, because the initial frames are what i want to see
21:13 akhasanov joined #mojo
21:18 trone joined #mojo
21:19 dotandimet joined #mojo
21:36 denis_boyun joined #mojo
21:42 noganex joined #mojo
21:45 sugar joined #mojo
21:51 Grinnz_ joined #mojo
21:54 denis_boyun_ joined #mojo
21:58 noganex joined #mojo
22:01 denis_boyun joined #mojo
22:40 good_news_everyon joined #mojo
22:40 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/0BYj-w
22:40 good_news_everyon mojo/master 67b2377 Sebastian Riedel: EV does not generate circular references
22:40 good_news_everyon left #mojo
22:44 cpan_mojo Mojolicious-Plugin-FormFieldsFromJSON 0.24 by Renee Baecker - http://metacpan.org/release/RENEEB/Mojolicious-Plugin-FormFieldsFromJSON-0.24
22:47 Grinnz_ sri: i was going to suggest something like _id the other day, i can use that for my POE reactor as well
22:48 sri it's a private api, you can't use it
22:49 sri i guarantee that it will break
22:50 good_news_everyon joined #mojo
22:50 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/U-2J6w
22:50 good_news_everyon mojo/master 7d5f180 Sebastian Riedel: a little more consistency
22:50 good_news_everyon left #mojo
22:50 Grinnz_ right, well... what i'm doing now is worse :P
22:51 Grinnz_ i guess i will just copy that function
22:54 sri you'd think making id generation a little faster would make a difference, but in the greater scheme of things those syscalls are just meh
22:58 Grinnz_ oh also i meant to ask, in Reactor::EV you have "$after ||= 0.0001 if $recurring;" in _timer; do you remember the reason for that? wondering whether i need it for POE
22:58 Grinnz_ Reactor::Poll doesn't do that
23:00 sri EV quirk
23:01 Grinnz_ i'm re-setting the timer manually in POE, as POE doesn't have a "recurring timer" function, so i'm guessing i wouldn't need it
23:09 trone_ joined #mojo
23:25 Grinnz_ joined #mojo
23:28 kmx joined #mojo
23:55 Grinnz_ aha, i don't need it, i was doing something else stupidly :P

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