Camelia, the Perl 6 bug

IRC log for #mojo, 2011-01-25

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

All times shown according to UTC.

Time Nick Message
01:47 perlrocks Twitter: "Mojolicious:: Routes not make any sense of any specification." (ja) --5_1_5 http://twitter.com/5_1_5/status/29716899608666112
02:06 al joined #mojo
02:11 perlrocks Twitter: "Eventually Mojolicious:: Routes to read." (ja) --5_1_5 http://twitter.com/5_1_5/status/29722950810738688
02:47 ispy_ left #mojo
02:56 perlrocks Twitter: "Mojolicious what went wrong code, what went wrong in my reading." (ja) --5_1_5 http://twitter.com/5_1_5/status/29734210730196992
03:15 perlrocks Twitter: "2 / 19 Hokkaido.pm # 4 yesterday's charsbar Mojolicious:: Lite hands-on, the afternoon's talks with yusukebe There are all kinds! Please tell the speaker the way you come first! http://goo.gl/q5YzW early!" (ja) --hokkaidopm http://twitter.com/hokkaido​pm/status/29739169425137664
03:41 perlrocks Twitter: "Ok, testing code. You don't like me and I don't know you, but I think #mojolicious will make it alright." --obfuscurity http://twitter.com/obfuscuri​ty/status/29745530137485314
03:41 tempire Ack!
03:41 tempire Net::Riak requires Moose.
03:41 tempire >:(
03:42 daviddelikat joined #mojo
04:04 tholen42 joined #mojo
04:06 tl left #mojo
04:31 GabrielVieira2 joined #mojo
04:37 GabrielVieira left #mojo
05:07 ispy_ joined #mojo
05:17 perlrocks Twitter: "I like # Java, Java and Perl's # Mojolicious do is come home to excellence. Five times the volume of different messages that are put in the line of code. It becomes something 作Ritaku." (ja) --flo_jack http://twitter.com/flo_jac​k/status/29769721922461696
05:28 crab can i "inject" some routes into a mojo app during testing?
05:33 Trelane left #mojo
05:34 Trelane joined #mojo
05:44 GabrielVieira2 left #mojo
05:48 xantus[] tempire: yeah, sucks huh
05:48 xantus[] I HATE MOOSE
05:49 xantus[] ls
05:59 crab xantus: why?
06:02 xantus[] its too much
06:12 kaare joined #mojo
06:22 josh left #mojo
06:22 josh joined #mojo
06:58 ysyrota joined #mojo
07:26 Alias left #mojo
07:30 tempire That's a good description if I've ever heard one.
07:30 tempire It's too much.
07:30 tempire One day, I'll probably taste the kool-aid & be sucked in.
07:31 tempire but not today!
07:31 metaperl left #mojo
07:31 * tempire rides his horse into the windmill
07:34 tempire Mojolicious: Come home to excellence. (ja)
07:34 tempire The japanese are a veritable gold mine of slogans
07:35 Christian joined #mojo
07:35 Christian good morning
07:35 purl Lies!
07:37 marcus I've been using moose a bit. It's pretty nice
07:37 marcus the main problem for me is that it is too powerful and encourages architecture astronauts
07:37 marcus people are tempted to do too much with roles and accessors
07:37 marcus making the flow in most programs hard to follow
07:43 koban joined #mojo
07:44 AmeliePoulain joined #mojo
07:47 perlrocks Twitter: "Mojolicious:: Routes Do it continued to read." (ja) --5_1_5 http://twitter.com/5_1_5/status/29807448072130560
07:56 tempire I used moose for a freelance project.
07:56 tempire It turned into an abomination.
07:56 tempire That's not so much mooses fault as it is my amateur usage of it
07:57 tempire I have been trying to find a good use for roles.  They seem so neat.
07:57 tempire But ever since my initial moose debacle, I always find a simpler solution.
08:06 moritz well, finding use cases for roles is the wrong approach
08:06 moritz you should just solve the problem at hand
08:06 moritz sometimes that can neatly be done with roles
08:06 moritz sometimes not
08:08 lee7_ left #mojo
08:10 crab roles are good when they're in the hay
08:17 Kulag left #mojo
08:17 Kulag joined #mojo
08:25 non joined #mojo
08:28 lee7 joined #mojo
08:43 crab hm
08:44 crab no access to debugging logs for when you run tests, right?
08:46 yko you need to catch log message?
09:06 crab yko: i was thinking it would be helpful
09:07 crab but now i have found the problem through other means
09:07 AmeliePoulain left #mojo
09:07 yko crab: i think catching log messages is better with Test::MockModule
09:15 otaviof joined #mojo
09:16 otaviof left #mojo
09:17 AmeliePoulain joined #mojo
09:19 AmeliePoulain What ID you are using for Perl?
09:19 AmeliePoulain *IDE
09:24 yko AmeliePoulain: vim with exclusive Mojolicious syntax highlighting :p
09:29 yakudzo AmeliePoulain, emacs
09:37 su-bzero is now known as su-bzero[A]
09:43 marcus textmate
09:43 purl well, textmate is an OSX text editor, at http://macromates.com/ or featuring dynamic code folding, "snippets", re-orderable tabs, and smart typing (auto-closing blocks, quoted strings, et cetera.) or the new BBEdit or the editor in all those Rails videos or non-free but quite worth it or notepad for OS X or requires a patch to enter Japanese characters which costs $5
09:43 marcus well, I think that is the problem
09:43 marcus most people see - oh, shiny, roles.
09:43 marcus and overuse it
09:44 marcus causing a lot of action at a distance
09:59 spleenjack joined #mojo
10:02 Alias__ left #mojo
10:03 Alias__ joined #mojo
11:24 perlite left #mojo
11:24 perlite joined #mojo
11:39 vaneska joined #mojo
11:41 su-bzero[A] is now known as su-bzero
11:41 metaperl joined #mojo
11:44 vaneska left #mojo
12:29 jdixon joined #mojo
12:29 jdixon #mojo keeps growing :)
12:30 marcus should graph it
12:33 jdixon I'm running into some weird issues with my tests under Lite.
12:33 jdixon first, it complains that $t masks earlier declaration, for every line $t is used.
12:33 jdixon if I remove "my" then it stops complaining.
12:34 marcus nopaste the test?
12:34 jdixon yup, about to
12:34 jdixon just wanted to give some background :)
12:34 marcus maybe you are using $t in your lite app?
12:34 jdixon hmm
12:34 jdixon nope
12:37 jdixon here's the test:
12:37 jdixon https://github.com/obfuscurity/​undeadly/blob/master/t/index.t
12:37 jdixon result: http://pastie.org/private/qwyucrll4xib2mffdhyja
12:38 jdixon after removing "my":
12:38 jdixon http://pastie.org/private/khw3ifrwth4thyegbdia
12:38 jdixon oh bother
12:38 jdixon fucking semicolon
12:39 jdixon ignore me  :-P
12:39 jdixon jfc
12:39 jdixon I need a "staging" dev irc channel, where I can paste my questions and realize my stupidity first, before unleashing it in the real channel.
12:43 HSlayer|2 joined #mojo
12:48 HSlayer left #mojo
13:27 Akron joined #mojo
14:03 arthas joined #mojo
14:08 kaare left #mojo
14:12 perlrocks Twitter: "Seriously, can testing get any easier than this? http://is.gd/CvZuT0 http://is.gd/WWdX5i #mojolicious" --obfuscurity http://twitter.com/obfuscuri​ty/status/29904464534044672
14:13 perlrocks Twitter: "[Announcement Scheduled 2 / 19 Hokkaido.pm # 4 yesterday's charsbar Mojolicious:: Lite hands-on, the afternoon's talks with yusukebe There are all kinds! Please tell the speaker the way you come first! http://goo.gl/q5YzW early!" (ja) --hokkaidopm http://twitter.com/hokkaido​pm/status/29904629999337473
14:16 perlrocks Twitter: "My HTML5 website for toddlers , designed to work on mobile devices http://www.microangels.in built using #mojolicious #perl" --mkirank http://twitter.com/mkirank​/status/29905330498437120
14:19 marcus cool
14:19 marcus microangels should be added to the wiki
14:19 jdixon when I finish the new OpenBSD Journal I'll add that too. :)
14:24 MisterHatt left #mojo
14:24 MisterHatt joined #mojo
14:50 dekimsey left #mojo
14:51 kaare joined #mojo
14:51 dekimsey joined #mojo
14:52 tempire left #mojo
14:53 tempire joined #mojo
14:54 jdixon left #mojo
14:54 jdixon joined #mojo
14:55 jamesw left #mojo
14:55 jamesw joined #mojo
15:00 AmeliePoulain left #mojo
15:13 fish__ ok, i'm again on my utf8 issue from a few days ago and i setup a small example to trigger this. seems not to be specific to this cms plugin. here is what i did: http://freigeist.org/static/mojo-utf8-test/
15:15 fish__ nothing more than a Mojo::Asset::File... ->slurp and hand it over to the renderer. the file is utf8 encoded (at least i think so..) but the output isn't
15:15 fish__ you can see it running here: http://freigeist.org:3000/
15:23 fish__ ah
15:24 fish__ if i convert this to latin, i get valid utf8
15:25 jdixon left #mojo
15:25 fish__ so i bet something in mojo is trying to convert it from latin1 to utf8, but its already utf8..
15:38 fish__ well, yes, per default everything gets encoded as utf8. and thats a problem is the content of a variable is already encoded
15:41 Christian cu all
15:42 Christian left #mojo
15:51 koban left #mojo
16:16 bellaire joined #mojo
16:27 Akron left #mojo
17:30 MojoGuest873 joined #mojo
17:30 MojoGuest873 From: http://www.google.com/url?sa=​D&q=http://dev.xantus.org (1 hits)
17:30 MojoGuest873 left #mojo
17:36 ysyrota left #mojo
18:25 mateu joined #mojo
18:59 spleenjack left #mojo
19:14 MojoGuest524 joined #mojo
19:14 MojoGuest524 left #mojo
19:18 MojoGuest434 joined #mojo
19:18 MojoGuest434 From: http://www.google.com/url?sa=t&source=web&​cd=3&ved=0CCUQFjAC&url=http%3A%2F%2Fdev.xa​ntus.org%2F&rct=j&q=irc%20websocket&ei​=qiE_TcWPPIT68Ab16JSBCw&usg=AFQjCNGw6nZv0wO42u​zUnHhWcD0vcKNULg&sig2=_0WeY1BvFo78EVLP1l13KQ (1 hits)
19:18 MojoGuest434 Hello
19:18 MojoGuest434 left #mojo
19:21 arthas left #mojo
19:39 sri evening
19:39 chilts left #mojo
19:40 sri fish__: your problem is not mojolicious
19:40 sri you're just throwing encoded data at mojolicious, which is wrong
19:40 sri *everything* needs to be decoded before you let mojolicious work with it
19:48 tempire left #mojo
19:58 duff left #mojo
20:16 marcus sri: the default generated tests fail now
20:17 marcus because you changed the default content type to include encoding
20:17 marcus maybe the mimetype test should ignore anything after ;
20:19 marcus perl6 is getting ipv6 support now. Wonder how non-blocking sockets are coming along
20:20 GitHub141 joined #mojo
20:20 GitHub141 mojo: master Sebastian Riedel * d9b3a8a (1 files in 1 dirs): fixed generated test - http://bit.ly/gSwBhO
20:20 GitHub141 left #mojo
20:24 sri marcus: fixed!
20:24 marcus sri: I noticed. Well done
20:24 sri lol, ipv6 before non-blocking -.-
20:25 sri anyway, perl6 is at least 10 years away
20:25 sri you just have to look at python 3 (like i tweeted a few days ago)
20:26 sri python 3 was ready 3 years ago... no adoption at all so far
20:26 sri and the perl community is moving so much slower than python
20:28 marcus mm
20:36 sri the bigger projects like django are not even planning the python 3 move yet
20:38 yko sri: can Mojo::DOM behave as JSON at new? i mean parse value
20:38 sri and remember the outrage when i tried to make mojolicious depend on perl 5.10 -.-
20:39 bellaire Has anyone tried using cgi-fcgi with Mojo? A.k.a. the "dumb" way to do fcgi?
20:40 sri no idea what that is
20:40 sri yko: not sure
20:41 yko ok
20:41 sri yko: actually Mojo::JSON doesn't do that either
20:42 bellaire It's a little C utility for people whose system administrators haven't gotten around to installing FCGI at apache.  It creates a socket and starts the FCGI server on the first requests.  The server persists and subsequent requests go from apache through cgi-fcgi and its socket to the server process.
20:42 yko sri: i mix it with package _Bool. grep is not always cool :)
20:43 sri for quick hacks you can always use ojo and x()
20:43 yko sri++
20:44 sri https://github.com/kraih/mojo/blo​b/master/lib/Mojo/Message.pm#L227 # there we already use the current ->new behavior
20:45 yko oh, yes. you're right
20:47 sri http://aws.amazon.com/ses/ # interesting
20:48 bellaire How about under SELinux? :)
20:48 sri quite cheap compared to the competition
20:48 sri bellaire: you don't make much sense
20:48 bellaire Yeah, I know. ...
20:48 sri more context needed i guess
20:49 bellaire We're running our stuff on RHEL5 with SELinux, and we have to do lots of stuff in userspace that you'd normally do at Apache.  Just makes it hard to troubleshoot when we're doing things so... unusually.
20:50 sri not really a mojolicious question then
20:50 bellaire Hence my earlier weird question about FCGI.  We don't have an apache handler for it, we have to use a CGI binary (cgi-fcgi, which comes from the FastCGI people) as a wrapper to get us the "same" functionality.
20:50 bellaire Probably not.  But we've got other comparable apps that run, and Mojo doesn't.
20:50 sri i guess you can do that using some psgi binding
20:51 bellaire That's a fantastic idea.  This is probably not a Mojo problem, but if I knew more about Mojo it would help me find the real problem and go beat on our sysadmins.
20:51 sri core mojolicious only does basic cgi and basic fastcgi
20:52 bellaire So there are some differences between Mojo's FCGI support and, say, FCGI.pm from CPAN.
20:52 sri yes, ours is pure perl
20:52 sri FCGI.pm is a messy xs module
20:52 bellaire Yeah, we noticed. ;)
20:52 sri just looking at it will make you feel very dirty :)
20:53 bellaire But yeah, since our situation is so unusual, we'll just have to keep investigating.  Just wanted to see if anyone else had a craptacular stack like us. :D
20:54 sri couldn't you like switch to pure http?
20:54 bellaire You mean have Mojo serve the requests itself?
20:54 sri if you're allowing an external fastcgi proc manager anyway it's not much different than running a backend http server
20:54 sri apache reverse proxy + hypnotoad
20:55 bellaire Also a good idea.  We can probably use that as a threat to get them to turn on real FCGI.
20:55 sri hehe
20:56 sri nginx + hypnotoad is very nice to work with
20:56 sri zero downtime software upgrades out of the box and stuff
20:56 bellaire Sounds nice.  I'm trying to contain my jealousy.
20:58 sri software upgrade for me these days is mostly just "git pull origin master; sudo kill -s 'USR2' `cat hypnotoad.pid`"
21:00 bellaire Software upgrade for us is make a convincing case to sysadmins, wait several weeks, ask about it again, have them say something about security and make a specific request to us, which we implement, only to miss the academic term breaks meaning we'll have to wait another 12 weeks.  Assuming they finish their security audit of the proposed change with a satisfactory result.
21:00 bellaire So, a little more slow than yours.
21:00 sri haha, reminds me of my telco days
21:03 fhelmber_ left #mojo
21:13 pid joined #mojo
21:18 perlite left #mojo
21:20 perlite joined #mojo
21:44 metaperl left #mojo
21:45 metaperl joined #mojo
21:51 jamesw left #mojo
21:53 bellaire left #mojo
22:12 jamesw joined #mojo
22:15 duff joined #mojo
22:40 snearch joined #mojo
22:44 kaare left #mojo
22:44 snearch left #mojo
22:56 polvo left #mojo
23:01 sri http://theoatmeal.com/comics/state_web_winter # brilliant
23:01 tempire joined #mojo
23:22 ispy_ left #mojo
23:32 pid left #mojo
23:36 ispy_ joined #mojo

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