Camelia, the Perl 6 bug

IRC log for #mojo, 2011-07-07

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

All times shown according to UTC.

Time Nick Message
00:18 * crab ponders his next mojo app
00:25 xantus[] I worked on the ev iowatcher last night
00:25 xantus[] crab: rterm? :)
00:33 perlrocks Twitter: "mojo - Mojolicious: Perl Duct Tape For The HTML5 Web! http://bit.ly/gvMUtq" --github_pl http://twitter.com/github_p​l/status/88767433032204289
01:04 xaka left #mojo
01:16 marcus left #mojo
01:19 marcus joined #mojo
01:23 GitHub89 joined #mojo
01:23 GitHub89 [mojo] kraih pushed 1 new commit to master: http://bit.ly/r1a0cl
01:23 GitHub89 [mojo/master] fixed small Hypnotoad bug and made tests more reliable - Sebastian Riedel
01:23 GitHub89 left #mojo
01:25 sri ok, fingers crossed, i think that was it
01:26 xantus[] hey sri is on g+
01:27 tardisx mojolicious circle
01:27 sri yea, and i already don't like it ;p
01:27 xantus[] I need a perl circile and a mojo circle
01:27 sri i mean... it is better than facebook
01:27 sri but twitter is still number one
01:27 tardisx i like the circle concept, but I feel like facebook will retaliate with a similar feature and the status quo will remain
01:27 tardisx on the other hand - it's google
01:28 sri signup form of course had a preselected box for personalized ads...
01:30 vel joined #mojo
01:31 xantus[] I deselected it
01:31 sri me too
01:32 xantus[] I'm afraid my my likeness (pic) and name will be used to promote some ad to a friend.
01:32 sri that would be very google
01:33 xantus[] I'm not sure if it would only happen if I +1 the product/site
01:33 anaio joined #mojo
01:33 sri it will start with "5 of your friends have an ipad...lalala"
01:34 xantus[] hm, does Chris dibona still tinker with perl?
01:37 tempire lulz, sri you must have been listening to twig while adding people to your circles
01:38 sri i selected people randomly
01:38 sri just to try it :)
01:38 sri i've only looked up marcus and xantus
01:43 crab do you have to be invited, or do you just join?
01:43 tempire crab: you had to be invited in during a particular timeframe
01:43 xantus[] msg me your gmail and I'll send you a msg, then it sends you an invite if there are any available
01:44 tempire then they shut it off for er'one
01:44 sri it's a bit weird
01:55 GitHub73 joined #mojo
01:55 GitHub73 [mojo] kraih pushed 1 new commit to master: https://github.com/kraih/mojo/commit/4c​417406fd47f54cf2809a6c7a642a95519a94f9
01:55 GitHub73 [mojo/master] small Hypnotoad tweaks - Sebastian Riedel
01:55 GitHub73 left #mojo
02:01 perlrocks Twitter: "Mojolicious 1:57 by SRI - http://metacpan.org/release/SRI/Mojolicious-1.57/" (ms) --cpan_new http://twitter.com/cpan_ne​w/status/88789668983873536
02:02 xaka joined #mojo
02:06 Alias joined #mojo
02:15 GitHub161 joined #mojo
02:15 GitHub161 [mojo] kraih pushed 1 new commit to master: http://bit.ly/pxKi3h
02:15 GitHub161 [mojo/master] a few more small Hypnotoad tweaks - Sebastian Riedel
02:15 GitHub161 left #mojo
02:27 xantus[] !! A g+ invite widget just appeared on my g+
02:27 xantus[] msg me your email if you want one
02:40 Alias \o
03:02 perlrocks Twitter: "/ Mojolicious reference data in memory to draw the ZIP - http://htn.to/sATKcM Introduction to Perl with sample code" (ja) --memememomo http://twitter.com/memememo​mo/status/88805037534490624
03:36 mattastrophe joined #mojo
03:36 giwi left #mojo
03:46 vel left #mojo
03:48 perlrocks Twitter: "ActiveState acquired Phenona Run Mojolicious / PSGI application: a mini-benchmark Dancer - micro web framework in Perl." (ru) --jcheree3kk http://twitter.com/jcheree3​kk/status/88816736387858433
03:51 kaare joined #mojo
04:04 tempire hooray!
04:05 * tempire tries to remember everyone he's supposed to invite
04:13 giwi joined #mojo
04:21 kaare left #mojo
04:58 Alias left #mojo
05:23 xantus[] hey hey
05:23 xantus[] crab: how goes?
05:24 xaka left #mojo
05:24 crab xantus: busy with work stuff now, will get back to reading about escape sequences tonight.
05:25 xantus[] cool
05:25 * xantus[] is working on Mojo::IOWatcher::EV
05:29 crab oh, that's cool. i like EV in principle, even if i don't have much use for it right now (for my mojo app, that is).
05:44 anaio left #mojo
06:08 perlrocks Twitter: "Mojolicious:: Lite will handle routed, if the environment variables will SCRIPT_NAME http://juick.com/1438221" (ru) --UndeRus http://twitter.com/UndeRus​/status/88851877772537857
06:17 koban joined #mojo
06:21 perlrocks Twitter: "@teenCoder I recommend checking out #Mojolicious. It has some pretty good guides, and is easy to start with - http://t.co/8KJw6ea" --marcusramberg http://twitter.com/marcusramb​erg/status/88855069579165697
06:21 marcus I be pimpin'!
06:22 marcus Got google recruiter mail today. Apparently they are impressed with my leet Perl skills.
06:22 alnewkirk|com nice
06:22 giwi left #mojo
06:23 giwi joined #mojo
06:27 crab marcus: i didn't think google had any use for leet perl skills
06:29 marcus "
06:29 marcus I am currently looking for Engineers with hybrid Unix/Linux Systems Engineering/Software Engineering who possess experience in coding in C/C++ or Java and/or scripting skills (Python, Perl, or Shell)."
06:30 marcus "I am part of the Google Staffing team and was wondering if you would be open to exploring engineering opportunities with Google. I did quite a bit of research in Ohloh and I'm impressed with your Perl background and thought your skills could be a fit for our team."
07:06 jamesw interesting
07:11 cosmincx joined #mojo
07:26 giwi left #mojo
07:26 giwi joined #mojo
07:29 arthas joined #mojo
07:36 sugar joined #mojo
07:57 spleenjack joined #mojo
08:10 psimanx1 joined #mojo
08:27 ajgb joined #mojo
08:29 tabbi joined #mojo
08:29 tabbi left #mojo
08:30 xantus[] nice marcus
08:36 eugen joined #mojo
08:54 plu marcus++
09:10 tabbi joined #mojo
09:10 tabbi left #mojo
09:17 eugen left #mojo
09:21 xantus[] EV is almost there
09:25 crab whee.
09:39 xantus[] \o/
09:46 fmerges joined #mojo
10:10 Foxcool joined #mojo
11:23 arthas left #mojo
11:26 arthas joined #mojo
11:36 sh9 joined #mojo
11:54 giwi_ joined #mojo
11:57 giwi__ joined #mojo
11:59 giwi left #mojo
12:01 perlrocks Twitter: "[Mojolicious:: Lite] [Notes] Mojo:: http://d.hatena.ne.jp/meme​memomo/20110707/1310040078 change the configuration of the Template" (ja) --memememomo http://twitter.com/memememo​mo/status/88940659972636672
12:02 giwi_ left #mojo
12:06 anaio joined #mojo
12:12 anaio left #mojo
12:24 sri \o/
12:26 * sri still remembers his encounter with a google recruiter
12:26 elb0w /o\
12:27 sri it was 6 or 7 years ago, i didn't prepare at all and totally screwed up the phone interview :S
12:28 elb0w looks better on my pc at home
12:28 elb0w whoops
12:28 elb0w I hate phone interviews
12:28 elb0w I always bomb them
12:28 elb0w Id be working at goldman right now if I was good at them
12:31 * marty finds it hard to believe that sri could screw up an interview about programming, even on his worst day.
12:36 sri i screwed up soooo bad
12:36 elb0w what did you screw up?
12:37 elb0w Well if you ever want to work in finance in NYC I can get you interviews
12:37 elb0w usually start around 80-100k
12:37 elb0w I've seen higher too
12:38 sri everything, didn't even answer the "syn ack fin" question....
12:39 psimanx1 sri: believe I have found a bug in hypnotoad - where do I report?
12:39 sri psimanx1: here
12:40 sri normally on github, but since we are both here anyway :)
12:40 psimanx1 Ok - when hypnotoad is listening on multiple ports and it loads up a page and the resources for that page - it hangs
12:41 sri can you nopaste your hypnotoad.conf?
12:41 psimanx1 ok 1 sec
12:41 psimanx1 http://nopaste.info/018aa1c044.html
12:42 psimanx1 If I remove the port 4000 element - it works ok.
12:43 sri ok, why do you modify heartbeat?
12:44 psimanx1 In my app - I actually have it a lot higher (3600) as I need to call out to a very long running external app - I have been trying to debug the issue though so lowered it to more reasonable levels.
12:45 psimanx1 In any case - even running default values - the behaviour is the same
12:46 psimanx1 I created a sample app to demonstrate if you want me to zip it up and send it over?
12:46 sri which environment have you tested this in?
12:47 sri best if you nopaste "mojo version"
12:48 psimanx1 http://nopaste.info/ed5ff11565.html
12:48 sri ah epoll
12:48 sri i bet it works if you do MOJO_POLL=1
12:48 psimanx1 that the problem?
12:48 purl it has been said that that the problem is that the dems are turning this into a policitcal battle when it shouldnjt' be
12:49 sri epoll support is relatively untested and experimental
12:49 psimanx1 Yep that works
12:50 psimanx1 So run without epoll for now then?  :)
12:50 sri yes
12:50 psimanx1 kk - many thanks
12:50 sri IO::Epoll is supposed to be compatible to IO::Poll, but it obviously isn't
12:50 dolmen joined #mojo
12:51 psimanx1 Ok - well at least you're aware of the issue now.
12:51 sri i think i've found a kqueue bug, maybe it is the cause for epoll going haywire too, will investigate
12:52 elb0w FUCK YEAH VACATION TUESDDAYYY
12:52 elb0w hellllllllo venice
12:53 elb0w croatia
12:53 elb0w and greek isles
12:53 elb0w woooooooo
12:54 dolmen I upgraded Mojolicious from 1.46 to 1.57 and my application is now broken
12:56 sri dolmen: if you're more specific i'm sure we can help you
12:56 dolmen The clients now gets a 200 OK with "HTTP/1.0 400 Bad Request\r\nContent-Type: text/palin\r\n" as the body
12:57 dolmen both firefox and curl
12:59 elb0w Oh I see the issue
12:59 dolmen on another query, curl gets "302 Found", with "Content-Length: 0" and an empty body
12:59 dolmen obviously curl has no cache
13:00 elb0w purl: pastebin
13:00 purl i heard pastebin was see nopaste. or see shadowpaste
13:00 elb0w purl: what is nopaste?
13:00 purl somebody said nopaste was http://paste.scsys.co.uk/
13:00 xaka joined #mojo
13:00 elb0w purl: thanks so much
13:00 purl no worries elb0w
13:01 elb0w sri: you should make a interface to add things to purl and responses
13:01 elb0w build it up over time
13:02 dolmen there is an interface tu purl
13:02 elb0w rally?
13:02 purl somebody said rally was sweet. So is GT and F1 (especially F1 in cities like Monaco, where they race on regular streets!)
13:02 dolmen elb0w?
13:02 purl elb0w is probably trying to change the dns
13:02 elb0w purl?
13:02 purl yes, elb0w?
13:03 dolmen purl, no elb0w is someone who doesn't know how to teach things to me
13:03 purl bugger all, i dunno, dolmen
13:03 dolmen elb0w?
13:03 purl elb0w is trying to change the dns
13:03 elb0w lol
13:03 elb0w purl: add
13:03 purl somebody said add was Attention Deficit Disorder or mentioned at http://www.mentalhealth.com or at http://www.add.org or spelled in UPPERCASE or planete.qc.ca/sante/elaine/ or self test at http://www.mentalhealth.com/dx/fdx-ch01.html or what was I saying? oooh, shiny! or shiny or a made-up disease cured by giving naughty children a good slap or Advanced Dungeons & Dragons or give me a couple of numbers
13:03 elb0w purl: adhd
13:03 purl i guess adhd is a bugger. :P or ADD crossed with autism or a made-up nothingness invented by doctors who want more money
13:03 sri please don't spam the channel
13:03 dolmen purl, elb0w is also someone who doesn't know how to teach things to me
13:03 purl dolmen: wish i knew
13:04 sri do it in query if you have to
13:04 sri dolmen: you have to provide details for us to be able to debug
13:04 dolmen what kind of details?
13:04 sri something that we can replicate
13:04 dolmen I'm running my app with Twiggy
13:05 sri maybe Twiggy is the problem
13:05 dolmen sri: I would like. But I don't know where to start.
13:05 sri then there is absolutely nothing i could do
13:06 elb0w sri: how do you handle multiple inputs, do you pass a comma seperated list and just split it?
13:07 sri purl is not my bot, she's longer on irc.perl.org than me
13:07 elb0w I mean in mojo
13:07 elb0w site.com/process/NYX,MSFT,GOOG
13:07 elb0w is that the common way to do it?
13:08 sri need to fix a hypnotoad bug first
13:25 vel joined #mojo
13:42 sri damn, the bug is right in front of my nose but i can't find it
13:43 sri elb0w: ?a=NYX&a=MSFT&a=GOOG
13:44 sri my @a = $self->param('a');
13:44 elb0w sri: oh, I just went with comma's easier to type
13:45 sri i suppose query params are old school now
13:45 elb0w but I guess I could use Mojo::URL if I do the params
13:45 elb0w decisions decisions
13:45 sri and damn, you're in NY?
13:46 sri i heard it's hell for open source developers
13:46 sri even the stuff you work on in your spare time belongs to your employer
13:47 elb0w yeah
13:47 elb0w well
13:47 elb0w I do free lance on the side
13:47 elb0w depends on the employer I guess
13:51 giwi__ left #mojo
13:51 tempire hooray marcus
13:51 tempire have you ever interviewed with one of the biggins before?
13:51 tempire same thing happened to me with amazon.  I thought I was pretty hot stuff.
13:51 tempire was definitely not prepared for their interview process
13:52 tempire got through 4 interviews, getting progressively worse
13:53 tempire make sure you can implement recursive algorithms over the phone, on the spot, like the back of your hand
13:53 sri those interviews seem to be designed for candidates fresh out of college
13:54 tempire definitely.
13:54 tempire also, know big O concepts if you don't already.  have them on reference, if you don't think about it regularly
13:54 tempire and know your sorting algorithms.
13:55 tempire that one killed me.
13:55 sri yea, they love sorting
13:56 tempire it's a good experience, though.  It illuminated some of the things I was lacking in.  Do the interview process even if you don't want to work for 'em
14:06 giwi joined #mojo
14:07 amf left #mojo
14:15 fmerges left #mojo
14:15 kaare joined #mojo
14:22 perlrocks Twitter: "I'm still running # Mojolicious on Perl ver5.8 and can't get rid of it for a while." (ja) --jamadam http://twitter.com/jamadam​/status/88976242568736769
14:26 amoore joined #mojo
14:29 sugar left #mojo
14:31 sugar joined #mojo
14:37 perlrocks Twitter: "Again, my favorite # Komodo Edit, and it is not self-indulgence with # Mojolicious it work:)" (ru) --MpaK666 http://twitter.com/MpaK666​/status/88979960303321089
14:38 * sri doesn't get it
14:39 sri what would be different with two listen sockets?
14:39 sri one listen socket works flawless, but with two kqueue bails out with bad filedescriptor right when i'm adding the first
14:41 sri and IO::Poll works flawless for both
14:43 * sri is tempted to kick out epoll and kqueue support
14:45 ispy_ joined #mojo
14:47 sri and it olny happens after a fork
14:47 spleenjack1 joined #mojo
14:48 sri so two listen sockets, after fork, when added to kqueue result in a bad filedescriptor error
14:48 sh9 left #mojo
14:48 sri remove any of the 3 conditions ant it works
14:53 spleenjack left #mojo
14:55 dekimsey joined #mojo
14:58 perlrocks Twitter: "I was a great person to be credited Mojolicious" (ja) --jamadam http://twitter.com/jamadam​/status/88985280190754816
14:59 koban left #mojo
15:06 marcus tempire: I told google I wasn't interested.
15:06 sri interesting, i have found the bug... and it is a really weird one
15:06 marcus tempire: I already run a company with 9 employees :p
15:07 sri wow, you've grown
15:07 marcus well, 7 right now. 9 from august
15:08 marcus we've just hired another designer, and http://judofyr.net/
15:08 sri cool
15:10 dekimsey left #mojo
15:12 tempire marcus: You should tell them you'd be interested in a buy out
15:13 tempire starting bid: 10m
15:13 marcus euro :p
15:13 tempire boosh
15:13 marcus == 14.24 mln USD
15:16 marcus http://answers.google.com/answ​ers/threadview/id/254787.html wow that shit is complicated
15:19 GitHub95 joined #mojo
15:19 GitHub95 [mojo] kraih pushed 1 new commit to master: http://bit.ly/o8uBzp
15:19 GitHub95 [mojo/master] fixed fork bug in Mojo::IOLoop - Sebastian Riedel
15:19 GitHub95 left #mojo
15:19 sri psimanx1: epoll might work again
15:20 sri i suspect it has the same fork problem as kqueue
15:20 psimanx1 cool - I'll give it a try and let you know
15:22 abra left #mojo
15:22 sri good, i'll release later today if it works for you
15:23 spleenjack1 left #mojo
15:24 pbaker joined #mojo
15:25 spleenjack joined #mojo
15:25 pbaker hello all, just started working with Mojo - quick question, I'm prototyping an app with Mojolicious::Lite and was wondering if there was a way to use before_dispatch to run a helper method.  IE in my case to check in each route that the Content-Type is set to application/json - could also be used to globally check that the user has logged in, etc
15:26 pbaker of course I could put the helper method in each route I have, would just think it would be cleaner to be able to globally set it to run for each route
15:26 sri yes you can call helpers on $self
15:27 sri no you should not be using a hook for that
15:27 sri see "under"
15:27 sri in the lite tutorial
15:28 pbaker taking a peak now - thanks
15:28 psimanx1 sri: yes - looks good - working now.
15:28 sri http://mojolicio.us/perldoc/Mojolicious/Lite#Under
15:28 sri psimanx1: \0/
15:28 psimanx1 good job!  :)
15:28 pbaker doh - I dont know how I missed that
15:28 * pbaker facepalms
15:29 sri pbaker: you're not the first
15:29 pbaker well this is exactly what I need - many thanks
15:29 sri i think hooks need a warning, that they are mostly just used to extend mojolicious itself, not for application logic
15:30 pbaker maybe, or I could have just read the documentation more carefully :)
15:30 sri pbaker: how did you stumble over hooks anyway?
15:30 pbaker just from the table of contents in the document
15:30 pbaker looking in Mojolicious
15:31 sri you're like the third person to draw the same conclusion, so i think the documentation needs to be tweaked ;)
15:31 pbaker not Mojolicious::Lite
15:31 sri ah
15:31 pbaker I was searching around and around, and couldnt find anything, so I went back to Mojolicious since there was some Lite stuff sprinkled into the document
15:31 pbaker Under just didnt pop at me
15:32 pbaker so I saw the dispatch methods and thought that might be it
15:32 sri http://mojolicio.us/perldoc
15:32 sri that should really be the first page everyone looks at
15:33 pbaker indeed
15:37 xaka left #mojo
15:44 tabbi joined #mojo
15:44 tabbi left #mojo
15:50 GitHub36 joined #mojo
15:50 GitHub36 [mojo] kraih pushed 1 new commit to master: http://bit.ly/qTysaB
15:50 GitHub36 [mojo/master] Improved Hypnotoad tests and Mojolicious documentation - Sebastian Riedel
15:50 GitHub36 left #mojo
15:52 sri http://mojolicio.us/perldoc/Mo​jolicious#Want_To_Know_More3F # hope that's a bit mroe obvious
15:54 xaka joined #mojo
15:56 sri hmm
15:56 sri i don't want to upload a new release :(
15:56 sri http://www.cpantesters.org​/distro/M/Mojolicious.html # so much green
15:58 cosmincx left #mojo
15:59 kaare left #mojo
16:01 xaka green is so boring
16:03 sri boring is good
16:04 xaka boring is kind of end of life
16:06 marcus supergreen!
16:07 marcus sri: poke!
16:22 * sri pokes marcus back, harder
16:25 crab mann gegen mann!
16:30 sri !
16:32 dolmen left #mojo
16:33 marcus mann gegen dwarf
16:33 marcus mann gegen zwerg
16:35 sri http://www.youtube.com/watch?v=qYNeT2nzEgA
16:36 marcus joffrey is hardly a man
16:41 amoore left #mojo
16:56 abra joined #mojo
16:56 abra left #mojo
16:56 abra joined #mojo
17:11 spleenjack left #mojo
17:35 arpadszasz joined #mojo
17:48 pbaker left #mojo
17:55 GitHub33 joined #mojo
17:55 GitHub33 [mojo] kraih pushed 1 new commit to master: http://bit.ly/rmsHyM
17:55 GitHub33 [mojo/master] added tests for Mojo::IOWatcher - Sebastian Riedel
17:55 GitHub33 left #mojo
17:57 perlrocks Twitter: "#mojolicious has just reached another milestone with 6000 unit tests. :) #perl" --kraih http://twitter.com/kraih/status/89030249072951296
17:58 sri \o/
17:58 sri and about 1300 LoC more in t/ than in lib/
18:02 elb0w there must be a better way to this, $self->render({json => JSON::XS->new->utf8->decode($basketData)});
18:03 sri why are you decosind the json data?
18:03 elb0w the json string is pulled from redis
18:03 sri s/s/d/
18:03 sri json is json
18:04 sri just pass it through
18:04 elb0w when I try to do json => $jsonString it escapes all quotes
18:05 sri http://mojolicio.us/perldoc/Moj​olicious/Controller#render_data
18:06 elb0w very nice
18:06 elb0w ty
18:06 elb0w sri you have a very impressive product
18:06 elb0w :)
18:06 sri MY PRECIOUS!
18:07 elb0w funny though, benchmark is pretty much the same
18:07 elb0w lol
18:09 sri well, json encode/decode is not very expensive
18:10 elb0w especially with xs
18:10 sri even a pure perl implementation can usually encode a few thousand non trivial documents per second
18:10 elb0w Mojo doesnt use XS if its present though huh
18:10 elb0w I dont really care, just curious
18:11 sri nope, it wouldn't make a difference for our use cases
18:19 sri if i use xs automatically it generally has to be for a feature that would be impossible without
18:19 sri like epoll, kqueue, tls, ipv6...
18:20 D4RK-PH0ENiX joined #mojo
18:23 d4rkie left #mojo
18:23 xantus[] I worked on EV last night
18:23 xantus[] I ran into an issue where it would only accept the first socket
18:25 sri hah
18:25 sri that might be fixed
18:26 sri if you're refering to a hypnotoad test with two listen sockets
18:27 sri problem was that epoll/kqueue got instantiated too early
18:27 sri and then the fork messed them up
18:27 sri now iowatcher instantiation should be delayed until after the fork
18:28 xantus[] nah, normal daemon
18:28 sri oh
18:28 xantus[] timers, everything worked
18:29 sri weird
18:34 giwi left #mojo
18:36 giwi joined #mojo
18:42 xaka sri: too early?
18:42 shonorio joined #mojo
18:50 arthas left #mojo
18:51 glick joined #mojo
18:51 GitHub78 joined #mojo
18:51 GitHub78 [mojo] kraih pushed 1 new commit to master: http://bit.ly/nqNb92
18:51 GitHub78 [mojo/master] more Mojo::IOWatcher tests - Sebastian Riedel
18:51 GitHub78 left #mojo
18:52 sri xaka: kqueue does not survive a fork
18:52 sri and i guess it's the same for epoll
19:01 xantus[] it looks like ev could
19:01 xantus[] from the docs
19:05 elb0w there a easy way to turn json into csv?
19:05 GitHub186 joined #mojo
19:05 GitHub186 [mojo] kraih pushed 1 new commit to master: http://bit.ly/nMYUsz
19:05 GitHub186 [mojo/master] even more Mojo::IOWatcher tests - Sebastian Riedel
19:05 GitHub186 left #mojo
19:05 preflex left #mojo
19:05 xantus[] yeah, sprinkle some perl around it
19:05 elb0w lol
19:06 d4rkie joined #mojo
19:07 xantus[] use JSON qw(decode_json); my $obj = decode_json($json); ... loop and print in csv format ...
19:07 xantus[] ;)
19:07 preflex joined #mojo
19:07 elb0w yeah i know, wanted to see if theres a hash to csv method somewhere
19:08 elb0w I guess that would have limited functionality huh
19:08 xantus[] theres some csv modules on cpan
19:09 D4RK-PH0ENiX left #mojo
19:15 sri some testing of t/mojo/iowatcher.t would be appreciated
19:15 sri i want to release soon
19:16 sugar left #mojo
19:17 glick what's the best way, using Mojo::Lite, to download a binary file?
19:18 glick I'm using $self->ua->get($url)->res->body;
19:18 xaka sri: do u ask us run tests on a diff platforms or ?
19:18 glick but it's giving me a different md5sum than LWP::Simple or wget
19:19 crab sri: passes on linux here, with TEST_EPOLL too
19:19 sri crab: great, thanks
19:19 sri xaka: yes
19:20 GitHub76 joined #mojo
19:20 GitHub76 [mojo] kraih pushed 1 new commit to master: http://bit.ly/nR6o5D
19:20 GitHub76 [mojo/master] made tests a little more portable - Sebastian Riedel
19:20 GitHub76 left #mojo
19:21 sri glick: there are quite a few recipes in the cookbook
19:23 elb0w odd why wouldnt render text => use \n to format the line
19:23 elb0w only when I view source
19:24 elb0w do I need to set doctype?
19:24 sri http://mojolicio.us/perldoc/Moj​olicious/Controller#render_text
19:25 elb0w haha
19:25 elb0w o man
19:33 preflex left #mojo
19:36 preflex joined #mojo
19:37 elb0w does hypno run faster than morbo or just more stable?
19:40 D4RK-PH0ENiX joined #mojo
19:45 elb0w I just built a REST that walks through all my companies executions, creates baskets, does some minor analytics and gives both JSON and CSV format outputs in 4 hrs
19:45 elb0w I want to hump your leg sri
19:46 d4rkie left #mojo
19:47 xaka elb0w: morbo for development, hypno for production
19:50 GitHub181 joined #mojo
19:50 GitHub181 [mojo] kraih pushed 1 new commit to master: http://bit.ly/qAIi2d
19:50 GitHub181 [mojo/master] fixed small Hypnotoad issue and made tests more reliable - Sebastian Riedel
19:50 GitHub181 left #mojo
19:50 xaka do sessions work during Mojo::Test process?
19:52 d4rkie joined #mojo
19:52 xantus[] left #mojo
19:55 D4RK-PH0ENiX left #mojo
19:56 glick sri: found the problem, I hit the MOJO_MAX_MESSAGE_SIZE limit, I though res->body was trying to parse or encode the binary data
19:56 glick <-- n00b ;)
19:56 perlrocks Twitter: "Mojolicious 1:58 by SRI - http://metacpan.org/release/SRI/Mojolicious-1.58/" (ms) --cpan_new http://twitter.com/cpan_ne​w/status/89060241098027008
19:59 sri Firefox 6 is now beta, one step closer to reliable websockets \o/
20:14 xantus[] joined #mojo
20:50 xaka sri: is there any way to do "relaxed" json_content_is test? i know values of fields A and B, but field C is always random so cant pass it's value to test
20:55 sri no
20:56 sri we compare with is_deeply
20:56 xaka so looks i have to use is/ok combination
20:57 xaka btw, what the preferred way: render_json or render(json =>) ? :)
20:57 sri enitrely up to you
20:57 sri personally i don't have a favorite
21:04 xaka documentation says that helper will be avaiable as a method of app object, but when i run tests and try to do "$t->app->some_helper" - it fails "Can't locate object method ..."
21:05 xaka from controller side everything is ok
21:06 d4rkie left #mojo
21:07 d4rkie joined #mojo
21:11 xaka code bug or docs bug? :) actually "sub helper" doesn't modify application object, only adds a helper to renderer
21:20 xaka oh, u have AUTOLOAD, my bad, but it's about my analize problem. the primary issue still with us
21:22 Foxcool left #mojo
21:30 xaka oh, omg, i fixed it by creating an instance of my app. Every test which i saw is beginning with "my $t = Test::Mojo->new(app => 'MojoliciousTest');" so $t->app always returns a string, not an object.
21:30 xaka i thought Test::Mojo->new automatically creates an object, but it isn't
21:54 vel left #mojo
22:09 perlrocks Twitter: "Mojocast #1: Fun with clouds! - Create your first Mojolicious web app http://t.co/VTU137C" --skylos http://twitter.com/skylos/status/89093720032542720
22:13 trone_ joined #mojo
22:13 psimanx1 left #mojo
22:18 trone left #mojo
22:18 GitHub5 joined #mojo
22:18 GitHub5 [mojo] kraih pushed 1 new commit to master: http://bit.ly/qCRKJp
22:18 GitHub5 [mojo/master] cleanup - Sebastian Riedel
22:18 GitHub5 left #mojo
23:01 sri tempire: btw. you could use the command line "get" with css selectors in a screencast, bet that would look really awesome
23:02 sri "Lets see if our title came out right."..."./myapp.pl get /foobar title"..."yes, looks great."
23:03 sri it wows people every single time
23:05 xaka sri: what about my research above? it was really hidden issue that $t->app returns not an object but string as it was passed to Test::Mojo contructor
23:05 sri xaka: working as intended
23:06 xaka serious?!
23:06 purl i guess serious is "This is sewius!" http://www.youtube.com/watch?v=KwxwUiDbRCk
23:06 xaka man, i dont know, your policy...
23:06 sri what policy?
23:06 purl Questions $2, Answers $6, Right Answers $10, Questions about policy $20, Web questions $300.  KLB service available at an extra charge.
23:07 sri you've not proposed any carefully researched solution
23:07 xaka how you make the deal. doc says "helper will be available via app object". doc says "$t->app return application object". i tried to use it all and i got in stuck because i didn't expect that actual return value is string
23:08 sri as far as i'm concerned it's allright
23:08 xaka and now you says "as intended"
23:08 ispy_ left #mojo
23:08 sri WAIT!
23:08 sri where does it say application object?
23:09 sri it should say something like applications, no more no less
23:09 sri s/s//
23:09 xaka Application to be tested.
23:10 sri then you lied
23:10 xaka really?
23:10 sri "doc says "$t->app return application object"
23:11 sri ok, here's the deal, if you want something to change you have to propose a solution
23:11 sri a carefully researched solution that doesn't break existing code
23:12 marcus smells like a documentation patch to me
23:12 sri at this point i usually say "i'm not interested in problems, only solutions"
23:12 sri :)
23:13 marcus at this point I usually say "Fuck it, I'm off to bed"
23:13 marcus ;)
23:13 sri yes you do ;p
23:14 marcus fuck it, I'm off to bed. Nite guys =)
23:14 sri nn
23:15 xaka in this case i believe that if u passed string it must be converted to an application object in any case. I was thinking about object (and other guys will do the same, 99%) because one doc part has "my $t = Test::Mojo->new(app => 'MyApp');", other doc part has "$t      = $t->app(MyApp->new);" and so on. I was thinking that i can pass string or object and the result always be an object
23:15 sri xaka: you're missing the point
23:15 sri i'm not arguing against that
23:15 sri i want you to find a solution
23:16 sri $t->app always being an instance sounds useful, but there are many many side effects that need to be checked and tested
23:17 sri only once that has been done a decision can be made
23:17 xaka you know framework better than me so i can only provide a solution, not a checks or tests because i dont know all internals
23:17 xaka i can do it, not a big deal, but i dont know what to test
23:17 sri then you're not providing a solution
23:18 sri neither do i
23:18 sri i would have to read up on code for hours too
23:19 xaka but reading the code is not a checking, you won't belive me if i say "yeay, code is code, let's do it"
23:19 xaka *code is clean
23:19 sri if you say you've done the checks, provided enough unit tests with the patch i will of course believe you
23:22 sri anyway, why would you want to call helpers that way?
23:24 xaka i have 2 kind of client in my web app: real client and fake client. I would like to add some data to fake client and retrieve it via GET request to check that web app works fine. Client is a helper because i use it in every controller. IN dev mode i use fake client, in prod mode i use real.
23:25 sri btw. don't just duplicate functionality from Mojo::Server
23:25 sri find a really clean solution
23:28 xaka hm, actually mojo::server has a clean solution or even more - the once solution. It's your Test::Mojo and you made the only way to initiate test helper, via Test::Mojo->new(app => ...). You already have 6000 tests which use this way. You have a lot of users which use this way. There is only one solution - reuse mojo::server logic.
23:30 sri sounds like you're on the right track :)
23:31 xaka i prefer superbike tracks
23:31 xaka or sportbikes, ah, who cares

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