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

IRC log for #mojo, 2015-02-21

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

All times shown according to UTC.

Time Nick Message
00:02 cbendana got another question, can't find a yes or not answer in the docs, does mojo support required query params ? like if para1 is present route X is used ?
00:02 cbendana s/para1/params1
00:06 good_news_everyon joined #mojo
00:06 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/A8G8
00:06 good_news_everyon mojo/master c51d113 Sebastian Riedel: more elegant shadow
00:06 good_news_everyon left #mojo
00:07 sri cbendana: no
00:07 cbendana sri, thanks
00:13 tempire I like the new perldocs
00:13 tempire I don't know about that new search bar
00:14 tempire I miss the bigger corner radius
00:14 sri it's the apple approved radius!
00:16 sri muhahahaha... i just went on an apple page to check... and stumbled upon this abomination https://developer.apple.com/library/ios/navigation/
00:17 tempire A manager must have gotten involved.
00:17 tempire I feel like I'm falling into those search boxes.
00:18 tempire I think the trouble is the darkness of the search input
00:19 Grinnz_ STAY IN THE SHADOWS JACKIE
00:20 * Grinnz_ hopes someone else has played that great game
00:20 sri afraid not
00:23 sri i've played candy crush though
00:28 tempire I never liked candy crush
00:28 tempire But BattleHeart is amazing.
00:44 jberger Grinnz_: wanna work on MCT?
00:50 asarch joined #mojo
00:56 Grinnz on what?
00:56 purl somebody said on was un pronom personnel imbécile qui qualifie celui qui l'emploie
00:56 Grinnz the game was the darkness btw
00:57 Grinnz purl, forget on
00:57 purl Grinnz: I forgot on
01:02 pink_mist if that was based on a comic, I think I read some of the comic
01:03 Grinnz it was
01:06 Grinnz https://www.youtube.com/watch?v=xLB1J2O8Pao
01:16 jberger Grinnz: Mojolicious Conference Toolkit
01:36 relberger jberger: is this your response to ACT ;)
01:38 jberger it is
01:38 jberger we'll see
01:38 relberger aww and I thought we were going to take batman's suggestion
01:38 jberger sjn sent out an email to the conference people, since we had been complaining about ACT saying that we shouldn't be so quick to reimplement it, because there are some complex features it has
01:39 jberger what was that?
01:39 jberger oh at the hackathon?
01:39 relberger yeps
01:39 jberger we might want to have a base to work from
01:39 jberger hackathons go fast
01:41 jberger relberger: you are riche!
01:41 jberger you flipped your naming scheme!
01:41 jberger o/
01:42 relberger yah i saw that email and I agree w him quite a bit.  profiles, payments, calendering, and workflows. its virtually a mission specific cms.
01:42 relberger yah someone using my nick, this is my backup ... sorry wasnt trying to be stealth and all that
01:42 jberger I'm leaving payments mostly out of my prototype
01:43 jberger haha, no worries
01:43 jberger just took me a minute
01:47 relberger when did the mojo reliance on IO::Socket::IP go into effect?  It got me snookered on cygwin
01:47 jberger relberger: https://github.com/kraih/mojo/blob/master/Changes#L36
01:47 jberger not very long ago
01:48 relberger gotcha .. ipv6 test failed .. spent time thinking it was my fault
01:49 relberger but you shoud have seen the mojolicious install fireworks on IO::Socket::IP dep install failure ... it was schizo
01:49 relberger mojo install kept on truckin til the end
01:51 jberger really?!
01:51 jberger and actually we had some issues on installing it too on FreeBSD
01:52 jberger something odd about the hostname setting
01:55 jberger sri: is there anything handy to do update in DBD::Pg
01:58 sri jberger: you linked to an IO::Socket::SSL version bump
01:58 jberger man, I'm really off my game!
01:58 jberger notice that relberger missed it too!
01:58 sri IO::Socket::IP happened here https://github.com/kraih/mojo/commit/5180ae376f75091db92b566fefc26899e55b8681
01:59 sri also, ipv6 tests are still optional
02:00 sri you just have to get IO::Socket::IP installed, which even ships with perl 5.20
02:02 sri and...WAT?
02:02 sri jberger: that one was for you ;p
02:03 jberger hahahhaa
02:03 jberger yeah, I'm telling you I can't focus on everything, learning Pg (and SQL for the most part) is TOUGH
02:04 * sri nods
02:04 sri Minion::Backend::Pg still pretty much sums up my current style
02:05 jberger I'm trying to write a "update the columns specified in this hash" query
02:05 jberger trying to see if I can do it without generating the sql
02:05 preaction nope :p
02:07 sri not so fast
02:09 sri at the very least you should be able to use a conditional to reassign the current value if the one passed to the placeholder is null
02:09 relberger sri: the v6 fail was in IO::Socket::IP v6only test, not mojo ... https://rt.cpan.org/Public/Bug/Display.html?id=102257
02:10 relberger but the mojo install just kept going even tho IO::Socket::IP failed to install
02:10 sri you're using 5.14 on windows? Oo
02:10 relberger seemed like virtually every test failed
02:10 sri that's crazy
02:10 relberger man it's cygwin latest system perl gimme a break :p
02:11 relberger I can't get anything 5.16+ to compile
02:11 Adurah Eh, 5.18 for cygwin here.
02:11 relberger ??
02:11 sri i wouldn't dare touching anything before 5.18 on windows
02:11 relberger 6bit?
02:11 relberger *64
02:12 relberger Adurah: I know 32 is ok but I have to run 64
02:12 Adurah HAVE to?
02:12 Adurah But, naw... this is the i686 one.
02:12 relberger have to == yes.  work related.
02:13 Adurah Clickin' those version numbers in the cygwin installer not helping?
02:13 relberger perhaps not because I install everything
02:14 relberger perhaps I need to look into those details
02:14 relberger :p
02:14 Adurah Click on the version number in setup.exe until it becomes 5.18, hopefully.
02:16 relberger in the middle of a dist upgrade so I will check after.  thanks for the tip.  curious tho I can't even brew a 5.18
02:17 Adurah Wonder if I can 5.20 it currently...
02:21 riche joined #mojo
02:30 good_news_everyon joined #mojo
02:30 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/A89Y
02:30 good_news_everyon mojo/master 8c92714 Sebastian Riedel: give graceful shutdown tests a little more time
02:30 good_news_everyon left #mojo
02:45 klapperl joined #mojo
02:45 riche Adurah: ok I am looking at the cygwin installer, and there is no option to change version of perl, have no idea how you are achieving that
02:46 Adurah It's the icon to the right of the version number, the little... reload icon.
02:46 Adurah Keep clicking it until the number changes.
02:47 Adurah By default it says Keep.
02:47 riche yah it's only cycling keep-reinstall-5.14-uninstall
02:47 riche anyways
02:47 Adurah Oh well.
02:47 riche guys sorry for being off topic I'm done with it now
03:01 jontaylor joined #mojo
03:23 good_news_everyon joined #mojo
03:23 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/A8bO
03:23 good_news_everyon mojo/master f4a5f89 Sebastian Riedel: explain how Mojo::Loader works
03:23 good_news_everyon left #mojo
03:26 good_news_everyon joined #mojo
03:26 good_news_everyon [mojo] kraih tagged v5.81 at 75f7408: http://git.io/A8br
03:26 good_news_everyon left #mojo
03:28 jberger sri: puts a lot of emphasis on the __DATA__ handling though doesn't it?
03:29 jberger I would at least add "also" as in "It also allows"
03:30 cpan_mojo Mojolicious 5.81 by Sebastian Riedel - http://metacpan.org/release/SRI/Mojolicious-5.81
03:32 good_news_everyon joined #mojo
03:32 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/A8NE
03:32 good_news_everyon mojo/master dddc904 Sebastian Riedel: bump version
03:32 good_news_everyon left #mojo
03:35 cpan_mojo Mojo-Pg 1.10 by Sebastian Riedel - http://metacpan.org/release/SRI/Mojo-Pg-1.10
03:37 cpan_mojo Minion 1.08 by Sebastian Riedel - http://metacpan.org/release/SRI/Minion-1.08 (depends on Mojolicious)
03:57 noganex joined #mojo
04:07 tempire WEBSOCKETS!!
04:07 purl WEBSOCKETS!!!
04:07 tempire (on ios)
04:08 tempire Is there an index of ACT features?
04:08 tempire Maybe we only need one or two
04:09 tempire 1) Register and pay (use eventbrite) 2) Submit a talk. 3) Edit a talk. 4) Admins approve talk. 5) Talk editing is limited
04:09 tempire So that's 5
04:13 Grinnz WEBSOCKETS!!!!!
04:13 purl WEBSOCKETS!!!
04:13 Grinnz also, wow redis is easy to setup, i might use that for this project
04:13 Grinnz ive never used nosql before :s
05:09 sri interesting, there are a still mojolicious tags that still result in 20k rps with examples/microhttpd.pl
05:09 sri like v0.999950
05:10 sri https://github.com/kraih/mojo/compare/v0.999941...v0.999950
05:10 sri there it went from 15k to 20k
05:10 sri (it hurts a bit to look at all the shitty code ;p)
05:23 Oleg joined #mojo
05:24 s1037989 First paragraph of the README: "Mojolicious is a new attempt at implementing this idea using state of the art technology."  Is it fair to say that Mojolicious is no longer and well past being "an attempt"?  Such wording might mislead peoples' perceptions...?
05:26 sri biggest drop in performance was here https://github.com/kraih/mojo/compare/v1.48...v1.49
05:28 sri yea, not surprising in this commit ;p https://github.com/kraih/mojo/commit/4c342c609af3d77d26bac24e75023a1b79583e3a
05:31 sri i've learned something interesting though
05:31 sri the write before checking if handle is writable thing was not responsible for the performance
05:38 good_news_everyon joined #mojo
05:38 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/A4t9
05:38 good_news_everyon mojo/master a7e03cd Sebastian Riedel: Mojolicious is not an attempt anymore
05:38 good_news_everyon left #mojo
05:38 sri s1037989: better? :)
05:39 s1037989 :)  I'd also say it's not new!!  Yeah??
05:40 s1037989 But yes, I dig it.
05:43 s1037989 What happened to the commit bot that its name is "good_news_every*on*"?
05:44 n2plato joined #mojo
05:55 avenj bytes are expensive
05:56 Adurah The name of the bot is a joke about Japanese people being unable to pronounce the word... I guess.
05:56 alnewkirk joined #mojo
06:00 sri starting to wonder if the performance difference is simply the cost of abstraction
06:03 good_news_everyon joined #mojo
06:03 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/A4m7
06:03 good_news_everyon mojo/master e6360a8 Sebastian Riedel: mention which files are used
06:03 good_news_everyon left #mojo
06:07 sri although response time got *a lot* better
06:07 sri longest with v1.0 was 33ms, and only 7ms with 5.81
06:12 s1037989 What measurement is yielding this poor performance that you're referring to?
06:13 sri i'm profiling the event loop with examples/microhttpd.pl
06:21 kaare joined #mojo
06:51 dotandimet joined #mojo
06:57 irq joined #mojo
06:57 marmez left #mojo
07:20 Anon021 joined #mojo
07:23 dotandimet joined #mojo
07:32 csson joined #mojo
07:46 Grinnz batman, fyi Mojo::Redis2 tests fail on redis older than 2.6 because the tcp-keepalive option didn't exist: https://metacpan.org/source/JHTHORSEN/Mojo-Redis2-0.17/lib/Mojo/Redis2/Server.pm#L112
07:47 Grinnz i dont have a version to test with, it happened on a friend's system
08:13 dotandimet joined #mojo
08:20 Vandal joined #mojo
08:31 trone joined #mojo
09:05 melo joined #mojo
09:42 amon joined #mojo
09:46 jontaylor joined #mojo
09:55 berov joined #mojo
10:03 sh4 joined #mojo
10:27 punter joined #mojo
10:28 punter batman: could you answer this, please? http://stackoverflow.com/questions/28639178/cant-get-mojoredis2-to-subscribe
11:16 trone joined #mojo
11:39 Oleg joined #mojo
12:13 trone joined #mojo
12:20 dotandimet joined #mojo
12:38 punter joined #mojo
12:38 punter batman, you there?
12:50 jberger the batman is away for a few days
12:51 punter ok
12:51 punter would anyone here happen to know to use Mojo::Redis2 as well?
12:52 punter my program doesn't work: http://stackoverflow.com/questions/28639178/cant-get-mojoredis2-to-subscribe
12:52 reneeb joined #mojo
12:54 punter ah, I had to store the return value of $redis->publish in a durable variable - now it works
13:00 berov1 joined #mojo
13:07 kaare joined #mojo
13:23 reneeb joined #mojo
13:23 cpan_mojo Mojolicious-Plugin-WebAPI 0.01 by Renee Baecker - http://metacpan.org/release/RENEEB/Mojolicious-Plugin-WebAPI-0.01
13:30 irq_ joined #mojo
14:02 riche well, ppl are already signing up for mojoconf via eventbrite
14:20 dod joined #mojo
14:21 stryx` joined #mojo
14:25 dod joined #mojo
15:05 cpan_mojo Mojo-Log-More 1.00 by ODC - http://metacpan.org/release/ODC/Mojo-Log-More-1.00
15:10 asarch joined #mojo
15:45 marmez joined #mojo
16:23 jberger punter: now that you have something working, and now that I am more awake
16:23 jberger can you try storing the redis instance in a more durable variable
16:25 jberger it seems that subscribe returns the instance (which is a little unusual from a mojo-style standpoint, but ok)
16:25 jberger so you are in fact storing the redis instance
16:28 jberger added an answer and comment to that effect
16:30 mst jberger: probably to enable
16:30 mst my $thing = Thing->new->subscribe(...);
16:33 jberger agreed, and yet typically we have had things like subscribe return the callback, so that you can do this
16:34 jberger my $cb = $thing->on(event => sub { ... });
16:34 jberger then
16:34 jberger $other->on(finish => sub { $thing->unsubscribe(event => $cb) });
16:34 jberger neither is right or wrong
16:35 jberger (thus my "but ok")
16:39 mst also, assuming it extends Mojo::Base, you already have ->tap
16:39 mst my $thing = Thing->new->tap(subscribe => ...);
16:40 jberger indeed, though few people seem to use tap (in my experience)
16:40 jberger I'm starting to use it in my routers
16:41 jberger mst: https://gist.github.com/jberger/c69599f61c67b6149af8
16:41 mst it's a nice way to contain mutable state
16:41 * sri has been wondering if returning the callback was the wrong decision
16:42 riche joined #mojo
16:43 sri it was done for ->once() at the time
16:43 sri https://github.com/kraih/mojo/blob/master/lib/Mojo/EventEmitter.pm#L36
16:43 mst jberger: that's really quite cute
16:44 pink_mist s/at the/upon a/ # just kidding :P
16:44 mst pink_mist++
16:45 jberger hehe
17:01 sri interesting, i totally forgot why the once() internals look a little awkward
17:02 sri it was for app->plugins->once(before_dispatch => sub {...}) and the like
17:02 * sri wonders if that's actually a feature
17:19 good_news_everyon joined #mojo
17:19 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/ARPJ
17:19 good_news_everyon mojo/master 978b14f Sebastian Riedel: more tests for one-time events
17:19 good_news_everyon left #mojo
17:19 * sri shrugs
17:25 good_news_everyon joined #mojo
17:25 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/ARX9
17:25 good_news_everyon mojo/master e59dad4 Sebastian Riedel: better description for Mojo::Loader
17:25 good_news_everyon left #mojo
17:35 asarch joined #mojo
17:36 sri suppose this might make a good gsoc project too https://github.com/kraih/mojo/pull/737#issuecomment-75382373
18:00 cpan_mojo Mojo-Log-More 1.01 by ODC - http://metacpan.org/release/ODC/Mojo-Log-More-1.01
18:05 phillipadsmith Wasn't there a `deploy heroku` command at one point?
18:19 gryphon joined #mojo
18:35 Oleg phillipadsmith: https://metacpan.org/pod/Mojolicious::Command::deploy::heroku
18:35 phillipadsmith Oleg: Yep, found it, not working for me… going to move on for now, but thanks. :)
18:36 Oleg see also this http://stackoverflow.com/questions/27771510/why-is-my-use-of-hypnotoad-crashing-on-heroku
18:36 Oleg looks like Mojolicious::Command::deploy::heroku doesn't work well
18:37 sri oh yea, that thread still makes me angry
18:38 sri brian d foy lost a lot of brownie points with me for that
18:43 sri specifically for this http://stackoverflow.com/questions/27771510/why-is-my-use-of-hypnotoad-crashing-on-heroku#comment44222059_27898742
18:44 sri but i guess stackoverflow is just not for me
18:51 sri ooh, Atom now supports multiple root folders in the sidebar
18:51 sri that's neat
19:00 tempire I didn't know people were still using that.
19:01 tempire I guess they would be.
19:01 tempire I should really be better about maintaining it.
19:02 tempire Should still work, though.
19:29 riche joined #mojo
19:35 zivester joined #mojo
19:37 riche sri: thanks for laughing at my 5.14 yesterday, now I am figuring out why POSIX addset() isn't passing tests :p
19:37 sri :)
19:39 riche I don't even know where to ask why it's failing, POSIX.pm and its xs is a wild beast.  but it's probably cygwin's fault
19:43 riche but there's a pattern.  adding a signal say 15 adds another signal at +32 ... consistently
20:33 Nemix joined #mojo
20:37 damaya joined #mojo
20:37 damaya Hi everyone. I want to do something, but not sure if there are docs on this.
20:38 damaya Basically, I want morbo to run as a service (systemd, whatever). I'd also like hypnotoad to run as a service.
20:40 damaya the thing is,
20:40 purl the thing is, is in spanish or la cosa es
20:41 damaya wtf
20:41 damaya joined #mojo
20:41 preaction purl: hush
20:41 purl o/` I can't say what I want to, even if I'm just kidding o/`
20:41 preaction purl: botsnack
20:41 purl :)
20:42 damaya :)
20:50 n2plato joined #mojo
21:26 disputin joined #mojo
21:33 damaya exit
21:52 good_news_everyon joined #mojo
21:52 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/AET3
21:52 good_news_everyon mojo/master f021448 Sebastian Riedel: has_subscribers can be a little more efficient
21:52 good_news_everyon left #mojo
21:53 berov joined #mojo
22:13 good_news_everyon joined #mojo
22:13 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/AEmH
22:13 good_news_everyon mojo/master 1acf76d Sebastian Riedel: just use a one-liner
22:13 good_news_everyon left #mojo
22:28 punter joined #mojo
22:31 jberger new good phrase: "screw multitasking, start omnitasking"
22:52 good_news_everyon joined #mojo
22:52 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/AElD
22:52 good_news_everyon mojo/master 0294962 Sebastian Riedel: improve Mojo::Reactor::EV performance slightly
22:52 good_news_everyon left #mojo
22:57 sri it's fun when a 200 rps improvement can be described with "slightly"
22:59 pink_mist well, that depends ... if it's 1000 + 200, that's quite a lot ... if it's 100000000000000000000 + 200, then it is indeed a very slight improvement :P
23:00 sri it's 15000 + 200
23:00 pink_mist ah, that is a bit more than slight then, in my humble opinion =)
23:01 sri and only for an artifical test case ;p
23:01 pink_mist sri++
23:01 sri in a real app it's maybe 0.5 rps :)
23:03 sri the top15 subs now http://i.imgur.com/GZoCX5J.png
23:04 sri Mojo::EventEmitter::emit is a bit curious
23:04 sri not much left to remove there
23:06 sri interesting that exclusive time is so high there
23:07 sri and looking into it http://i.imgur.com/Csvlgur.png
23:08 sri 1.35s on return $self
23:16 bobkare joined #mojo
23:27 buu joined #mojo
23:30 cpan_mojo Mojolicious-Plugin-WebAPI 0.02 by Renee Baecker - http://metacpan.org/release/RENEEB/Mojolicious-Plugin-WebAPI-0.02
23:32 buu_ joined #mojo
23:58 sri top15 with the poll reactor are more interesting http://i.imgur.com/Fm413Uh.png

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