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

IRC log for #mojo, 2014-02-08

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

All times shown according to UTC.

Time Nick Message
00:25 s1037989_ left #mojo
00:27 s1037989_ joined #mojo
00:27 d4rkie joined #mojo
00:27 s1037989_ What version of Perl is required for mojo?
00:27 sri says right on the frontpage
00:28 s1037989_ K, that's what I thought.
00:28 s1037989_ The version of Perl that I'm on needed to have Socket upgraded it seems.  Does that sound right?
00:29 sri what version is that?
00:29 purl version that is current but maximally forward compatible
00:29 s1037989_ 5.10.1
00:29 sri how did you notice?
00:29 s1037989_ mojo version showed IO::Socket::IP not installed but cpanm kept saying it was latest version 0.27
00:30 s1037989_ $ perl -e 'use IO::Socket::IP; say $IO::Socket::IP' Socket version 1.97 required--this is only version 1.82
00:30 sri ah, that's only for ipv6 support
00:30 s1037989_ Then I updated Socket to 2.013 and now it all works.
00:31 d4rkie joined #mojo
00:34 s1037989_ Is that something that should be documented?  That the optional IO::Socket::IP requires (apparently) Socket 1.97?
00:35 sri no
00:35 s1037989_ k!
00:35 s1037989_ Just thought I'd ask!  :)
00:46 s1037989 joined #mojo
00:54 good_news_everyone joined #mojo
00:54 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/pSwhKA
00:54 good_news_everyone mojo/master 117ac45 Sebastian Riedel: documentation tweaks
00:54 good_news_everyone left #mojo
01:35 ashleydev joined #mojo
01:37 snap_ joined #mojo
01:44 asarch joined #mojo
02:07 hesco using Test::Mojo, I have a simple $get_ok('/route')->status_is(200)->content_like() test in my test script.  which was working fine until I changed the route to that target.  In a browser, the page loads expected content.  but the test script reports "got: '500'; expected: '200'".  As the browser shows me different than the test, I'm trying to sort out where to start looking for the discrepancy.
02:08 hesco Any pointers would be appreciated.
02:12 beyondcreed joined #mojo
02:14 good_news_everyone joined #mojo
02:14 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/Q3OKtg
02:14 good_news_everyone mojo/master 54d13a7 Sebastian Riedel: improved get_line performance
02:14 good_news_everyone left #mojo
02:16 sri that's actually a hotspot, gets me a few hundred rps in a micro benchmark ;p
02:19 klapperl_ joined #mojo
02:23 hesco thanks.
02:27 ka2u joined #mojo
02:37 s1037989 joined #mojo
02:40 good_news_everyone joined #mojo
02:40 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/JJo-Ug
02:40 good_news_everyone mojo/master 3107397 Sebastian Riedel: deprecated Mojo::Util::get_line
02:40 good_news_everyone left #mojo
02:41 sri mavericks is a terrible terrible operating system... hangs all the time
02:41 d4rkie_ joined #mojo
03:25 sri oh, didn't know you could do "use feature :all"
03:25 sri maybe that would be nice to have in Mojo::Base
03:26 sri or maybe not
03:27 sri too much risk of breakage when you upgrade perl
03:36 buu Damn the breakages and full speed ahead
03:48 s1037989 joined #mojo
03:50 ka2u joined #mojo
04:02 good_news_everyone joined #mojo
04:02 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/EHA6-g
04:02 good_news_everyone mojo/master a629a38 Sebastian Riedel: small optimization
04:02 good_news_everyone left #mojo
04:05 Stefan joined #mojo
04:06 jb360 joined #mojo
04:10 jack joined #mojo
04:19 btyler joined #mojo
04:24 s1037989 joined #mojo
04:29 irq joined #mojo
04:50 ka2u joined #mojo
05:33 jberger joined #mojo
05:45 irq_ joined #mojo
05:50 sri haha, the number one hotspot in mojolicious is this line now https://github.com/kraih/mojo/blob/master/lib/Mojo/Headers.pm#L22
05:55 davido_ joined #mojo
06:02 tempire_ joined #mojo
06:08 davido_ So... I believe Types::Serialiser doesn't overload stringification of ::true and ::false, so the return values are equal to numeric 1 and 0, not string 1 and 0, as the t/Mojo/JSON.t expects.  Is this an issue I should address, or simply document?
06:09 davido_ My inclination is that it's not going to be a problem in the wild, but should be documented.
06:20 crab follow your inclination, then
06:33 jberger sri that line only gets called one per monkey patch right?
06:33 jberger Hard to believe that's such a hotspot
06:41 sri jberger: *after* the monkey patching
06:42 jberger Oh the closures themselves. Of course
06:42 * jberger places head firmly on desk
06:43 jberger It's too late for geeking
06:44 * jberger reminds sri about sleep
06:49 * sri googles "sleep"
06:50 sri oh, btw. my profiling workflow for the http stack... perl -Ilib -d:NYTProf examples/fast.pl daemon -l http://*:8080 -m production; wrk -c 100 -d 10s http://127.0.0.1:8080/; nytprofhtml; open nytprof/index.html
06:50 d4rkie joined #mojo
06:51 sri s/fast.pl/hello.pl/ for basic mojolicious stack profiling
06:51 davido_ Crossed a dozen tests off the TO_DO list.  About a dozen to go.
06:52 davido_ sleep time.
06:53 d4rkie joined #mojo
06:57 gnephiak joined #mojo
07:06 Vandal joined #mojo
07:31 laouji joined #mojo
07:41 s1037989 joined #mojo
07:42 laouji joined #mojo
07:42 abra joined #mojo
09:04 firnsy joined #mojo
09:09 voegelas joined #mojo
09:10 meshl joined #mojo
09:38 punter joined #mojo
09:59 irq_ joined #mojo
10:04 denisboyun joined #mojo
10:09 firnsy joined #mojo
10:09 firnsy joined #mojo
10:12 basiliscos joined #mojo
10:14 cleverfox joined #mojo
10:53 jack joined #mojo
10:55 laouji joined #mojo
10:56 jack joined #mojo
10:57 laouji joined #mojo
10:59 sh4 joined #mojo
11:17 laouji joined #mojo
11:28 dvinciguerra joined #mojo
12:05 hummeleBop joined #mojo
12:30 bowtie_ joined #mojo
12:39 d4rkie joined #mojo
12:46 dod joined #mojo
12:57 d4rkie joined #mojo
13:06 dod joined #mojo
13:08 mire__ joined #mojo
13:31 cleverfox joined #mojo
13:33 mattastrophe joined #mojo
13:44 cleverfox joined #mojo
13:54 laouji joined #mojo
14:03 cleverfox joined #mojo
14:20 cleverfox joined #mojo
14:23 jberger I'm so glad perl isn't this opinionated
14:23 jberger http://neopythonic.blogspot.com/2009/04/tail-recursion-elimination.html?m=1
14:24 jberger ^^ ick
14:56 asarch joined #mojo
14:57 jack joined #mojo
15:25 laouji joined #mojo
15:36 mire__ joined #mojo
15:38 zivester joined #mojo
15:42 laouji joined #mojo
15:53 laouji joined #mojo
15:53 s1037989 joined #mojo
16:09 alnewkirk joined #mojo
16:28 s1037989 joined #mojo
17:13 mudler joined #mojo
17:24 laouji joined #mojo
17:39 ryozi joined #mojo
18:00 punter joined #mojo
18:06 beyondcreed joined #mojo
18:25 cpan_mojo Mojolicious-Plugin-CSRFProtect 0.16 by Viktor Tuskyi - http://metacpan.org/release/KOORCHIK/Mojolicious-Plugin-CSRFProtect-0.16
18:26 ka2u joined #mojo
18:41 jberger_ joined #mojo
18:48 rem_lex| joined #mojo
19:03 mattastrophe joined #mojo
19:19 jberger joined #mojo
19:20 jberger_ joined #mojo
19:26 laouji joined #mojo
19:29 jberger joined #mojo
19:30 radixo joined #mojo
19:30 jberger_ joined #mojo
20:31 ncls joined #mojo
20:32 ncls mst: hi ! I Was watching Scotland-England rugby play earlier, and I wondered : which team do you support ? (assuming you like rugby)
20:38 mst ncls: nominally england when I remember to watch it, but mostly I just enjoy watching rugby
20:39 ncls mst: nice victory then
20:40 ncls I also like to watch for the game rather than for a team, Ireland - Wales was a good play too
20:40 ncls think Ireland will win the tournament
20:42 ncls will / could / should
20:43 punter joined #mojo
20:44 jberger joined #mojo
21:04 jberger_ joined #mojo
21:05 skullbocks joined #mojo
21:06 mixMonkey joined #mojo
21:09 sri hmm, is there a good reason why type checks in Mojolicious::Types should be case sensitive?
21:10 sri as in my $mime_type = app->types->type('PNG')
21:11 sri i guess not
21:22 good_news_everyone joined #mojo
21:22 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/0o_ytw
21:22 good_news_everyone mojo/master 246b9d3 Sebastian Riedel: fixed case sensitivity bug in Mojolicious::Types
21:22 good_news_everyone left #mojo
21:27 laouji joined #mojo
21:34 jberger joined #mojo
21:46 jberger_ joined #mojo
21:50 maxhq joined #mojo
21:53 jberger_ mudler, why this? https://github.com/jberger/Toastr/pull/1/files#diff-407646638b6bfebe229329697d4501c5R9
21:53 jberger_ that's supposed to highlight line 9
21:54 jberger_ doesn't really look like it for me tho
21:54 russum joined #mojo
22:16 good_news_everyone joined #mojo
22:16 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/PoK2zQ
22:16 good_news_everyone mojo/master 44708bc Sebastian Riedel: small optimization
22:16 good_news_everyone left #mojo
22:17 s1037989 What's the best way to "install" an app on a Linux box so it starts on system boot?  I want it to run by hypnotoad but I'm using perlbrew and not the system Perl.
22:18 jberger_ s1037989, if you get a good answer to that, I will be interested too
22:18 jberger_ hypnotoad + perlbrew on a non-root user is my typical setup too
22:18 jberger_ (with nginx for reverse proxy)
22:19 s1037989 jberger_: What do you think about doing it the way convos is built with carton?  Then you point your upstart script or whatever to /path/to/carton exec /path/to/hypnotoad /path/to/app
22:20 jberger_ I haven't looked at that yet
22:20 jberger_ sounds ok
22:20 jberger_ how do non-root upstart jobs work
22:20 jberger_ ?
22:21 * jberger_ is no sysadmin
22:21 jberger_ :-)
22:25 s1037989 I don't know how non-root upstart jobs work, either!
22:25 s1037989 su user?  :D
22:26 s1037989 Perhaps the program needs to get launched as root then you just set your uid and gid to non-root?  Drop your root priviledges...  Wouldn't that be sufficient?
22:27 s1037989 But then I'd wonder how you'd call a specific perlbrew from a non-root user from upstart.
22:28 jnbek^dt joined #mojo
22:30 s1037989 jberger_: this looks promising: http://muraiki.com/2014/01/20/deploying-catalyst-with-nginx-and-upstart/
22:31 s1037989 If I make something of it and finetune it for Mojo, I'll post it to the wiki.
22:33 jberger_ s1037989, looks very promising
22:33 jberger_ good find!
22:34 purl good find is, like, hard to man
22:53 russum joined #mojo
22:54 priodev joined #mojo
23:01 good_news_everyone joined #mojo
23:01 good_news_everyone [mojo] kraih tagged v4.78 at 3b4c134: http://git.io/m3povg
23:01 good_news_everyone left #mojo
23:02 cpan_mojo Mojolicious 4.78 by Sebastian Riedel - http://metacpan.org/release/SRI/Mojolicious-4.78
23:04 jberger_ \o/
23:06 sri \o\
23:06 sri /o/
23:09 punter joined #mojo
23:11 Akron joined #mojo
23:15 s1037989 sri: at your pace, your gonna his that 5.0 mark real soon!!!  I assume you're reserving 5.0 for major modifications.  What will you do when you reach 4.99?
23:16 sri nope, this time there are no big breaking changes planned for 5.0, it will just be released when we run out of numbers ;)
23:16 jberger_ the only one I know of (besides the announced deprecations) is the change in behavior of Mojo::JSON::j
23:17 s1037989 Bummer!  :D
23:17 * sri was about to link to the open issues https://github.com/kraih/mojo/issues?page=1&state=open
23:17 jberger_ mojo is getting stable, sri++
23:17 sri s1037989: what would you like to see changed? ;p
23:17 s1037989 Haha!  :D  You're asking the wrong person!
23:17 * jberger_ is most excited about the codename
23:18 sri most changes these days are so minor they can be made with simple deprecations
23:18 * jberger_ hopes for FACE SCREAMING IN FEAR
23:18 * sri thinks we'll do 5.0 around mojoconf
23:18 jberger_ oh, not around, AT!
23:19 jberger_ sri: when mojo makes a big  bump in perl version, say for signatures, we should call that one PILE OF POO
23:19 Akron jberger: I know some services use djb's daemontools to allow adding services by non-root users. Works well with hypnotoad (you just have to run it in foreground for some reasons). http://cr.yp.to/daemontools.html
23:19 Adura I use a crontab @reboot
23:20 s1037989 Akron: any examples you can give?
23:20 sri unicode 7.0 has some fun new characters http://babelstone.blogspot.de/2013/10/whats-new-in-unicode-70.html
23:21 sri REVERSED HAND WITH MIDDLE FINGER EXTENDED
23:21 jberger_ Adura, that's not a bad idea, would be nice to be able to use the service command for hot-reload, but, I would trade that for ease of use
23:21 jberger_ sri: hahahahaha
23:21 purl LOLCON 5 reached.
23:21 jberger_ does Perl support unicode 7.0 yet?
23:22 sri i bet 5.20 will
23:22 jberger_ I know it keeps pretty close to the forefront
23:22 sri besides Mojo::JSON::j i was also thinking about breaking Mojo::Log::format... maybe turn it into a callback attribute... app->log->format(sub {...});
23:23 sri HTTP/2 will require bigger breaking changes, but that's a 6.0/7.0 feature
23:24 jberger_ MAN IN BUSINESS SUIT LEVITATING
23:24 sri that's a good candidate
23:25 Akron s1037989: Uberspace https://github.com/kraih/mojo/wiki/Uberspace
23:25 Adura The @reboot screep needed a ". /home/USER/.perlbrew/init" line for perlbrew, though.
23:25 Adura *script
23:26 s1037989 Akron: Great!  Thanks!
23:27 * sri wonders if choosing the next codename should be a mojoconf event, possibly with a vote :)
23:27 s1037989 sri: how did you set up mojolicio.us to auto start?
23:27 sri that could go horribly wrong though... *cough* pile of poo *cough*
23:28 laouji joined #mojo
23:29 jberger_ sri: we could have a list of allowed (or disallowed) candidates
23:30 * s1037989 wonders if he's even allowed to attend at all!!  :D
23:30 Akron jberger_: candidates for voters or name nominees?
23:32 * jberger_ bans s1037989 from the mojoconf refreshments table :-P
23:32 s1037989 Haha!  Aw....  :(
23:33 sri s1037989: i didn't, it just uses git hooks to start/restart hypnotoad
23:33 sri and runs inside a sandbox, so i don't care too much
23:33 * jberger_ reinstates s1037989 pending a formal hearing
23:33 s1037989 sri: Well that's neat!  But what happens if the OS restarts?  How does the server launch?  Do you have to manually launch it?
23:34 Akron sri: So mojolicio.us is always on a current version - restarting on every commit or on every version tag?
23:34 sri yes i do, like i said it's a sandbox i test stuff in all the time
23:34 sri Akron: yea
23:35 Akron That's nice. Didn't know that.
23:35 s1037989 sri: Dunno if this is something you could do, but it'd be interesting for you to discuss at mojoconf some of the projects you're working on.  Looking at your github, it looks like you spend all your time creating Mojo for all of us and you don't even use it yourself!  :D
23:36 sri it's hypnotoad, so updating all the time and zero downtime restarting doesn't cost anything :)
23:38 Akron Yeah. On the job I tend to commit broken tests sometimes - so there I wouldn't be able to be bleading edge all the time. But I like this concept.
23:39 s1037989_ joined #mojo
23:39 sri i suppose deploying automatically from travis-ci would be fun, to catch test failures
23:40 btyler joined #mojo
23:46 russum joined #mojo
23:46 LordVorp joined #mojo
23:56 jberger_ s1037989, sri is very private about his other projects ;-)
23:56 jberger_ he wouldn't show us Mango for a long time
23:59 sri i also don't ever talk about consulting clients, unless they've specifically given permission

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