Camelia, the Perl 6 bug

IRC log for #mojo, 2011-02-11

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

All times shown according to UTC.

Time Nick Message
00:17 janus left #mojo
00:51 vel left #mojo
01:09 Alias left #mojo
01:19 tholen_ joined #mojo
01:23 tholen left #mojo
01:28 tempire oh
01:28 tempire my
01:28 tempire goodness
01:28 purl Gracious me!
01:28 tempire github cmd+t support is my christmas
01:29 tempire isitchristmas.com is WRONG today
01:34 crab what is it?
01:34 purl it's it!
01:35 tempire https://github.com/blog/793-​introducing-the-file-finder
01:36 ispy_ joined #mojo
01:40 perlrocks Twitter: "Anyone got a comparison of pros/cons/use-cases for #Mojolicious and #perldancer? Hard to tell which is preferable in a given situation #perl" --MisterHatt http://twitter.com/MisterHa​tt/status/35875892412547072
01:41 MisterHatt it is scary how fast that bot acts sometimes
01:42 sri mojolicious is a catalyst class web framework with a micro framework frontend for prototyping, dancer is a micro framework
01:43 sri mojolicious wants your app to grow over time, dancer wants it to stay small
01:44 MisterHatt that is remarkably concise, and exactly the info I needed actually
01:44 sri :)
01:45 MisterHatt got a small 99% static site that I wanted to be able to adjust functionality easily, wasnt sure what would be better
01:45 MisterHatt not much growth on it tho, about the only thing I can see it needing in the future is a contact form
01:45 crab if i had nothing else to do and wanted to annoy people, it shouldn't be too hard to implement most of dancer as a wrapper over Mojolicious::lite.
01:45 MisterHatt crab: why not just use dancer then? 5 deps isn't too bad
01:45 MisterHatt (until you try to do anything anyway, then it turns into like 50 deps)
01:45 crab misterhatt: how would that annoy people?
01:46 MisterHatt indeed
01:46 sri it's more than 5
01:46 MisterHatt is it?
01:46 purl it's it!
01:46 sri you forget deps of deps
01:46 sri plack has 10 or so
01:47 MisterHatt I run mojo via plack anyway
01:50 MisterHatt office (coffee shop) time now, best job
02:00 tempire sri: Are you going to post your response to misterhatt on the twitters?
02:00 tempire seems like something people need to be reminded of often
02:00 * tempire wonders about the best way to blog about that
02:03 sri tempire: my next post will be about the highlights in 1.1
02:05 sri oh, twitter, not blog
02:06 sri not much of a point answering there i guess, since he's here
02:06 sri doubt i could blog about that topic without starting a new war with someone -.-
02:06 tempire I was thinking for everyone else, but I suppose twitter doesn't usually work like that.  My sense of twitter is skewed due to the bot.
02:10 jamesw joined #mojo
02:11 sherr left #mojo
02:23 sri the bot also deserves its own blog post
02:25 MisterHatt yeah, how DID you make it so quick
02:25 MisterHatt I actually wanted to see how fast it was so put irssi next to my browser and watched it go, was almost instant
02:26 sri https://github.com/kraih/mojoli​cio.us/blob/master/perlrocks.pl # it is built with mojolicious
02:26 MisterHatt I thought as much
02:28 sri whats really cool is that it google translates the messages actually before announcing them here
02:28 crab sri: mind if i submit a patch to the cookbook base rewriting example to handle X-Forwarded-Host too? and a comment in the code about how those headers may all contain multiple values, and you should use the last one?
02:29 sri crab: put it on the wiki
02:29 sri the recipe is not supposed to be more specific
02:29 sri since there are no standards
02:29 crab indeed
02:34 MisterHatt maybe if you're covering headers in the cookbook, have an entry about writing headers for varnish caching?
02:34 MisterHatt I could maybe write it after playing a bit more :S
02:35 sri put it on the wiki
02:35 MisterHatt the request for it or actually writing it?
02:35 sri writing it of course ;p
02:35 MisterHatt orz
02:35 crab heh
02:36 MisterHatt dont think I'm there yet, only just coming back into perl after a few years
02:36 MisterHatt as kulag said, my code is "ugly"
02:36 stephen left #mojo
02:36 sri thats exactly why we have the wiki ;)
02:37 sri you put valuable information into it, people come and clean it up for you
02:37 Kulag MisterHatt: Your code is not "ugly", it is "not built with an understanding of perl or OOP"
02:39 MisterHatt not my fault microcontroller code is procedural for the most part :V
02:39 MisterHatt I'll figure out OOP eventually
02:41 Kulag How about you relearn perl first?
02:41 MisterHatt getting there
02:41 purl getting there is probably the problem. or half the fun
02:41 Kulag lol
02:41 crab or port mojo to assembly language
02:41 MisterHatt I even went to a perlmongers meeting the other day
02:41 MisterHatt where everyone uses catalyst
02:42 MisterHatt was interesting tho
02:46 sri it is the standard atm
02:51 MisterHatt I used it a bit, not much of a fan
02:52 sri well, it is from another time
02:53 MisterHatt indeed
02:53 MisterHatt I wonder if anyone still uses maypole or miril
02:54 sri prolly
02:54 sri perl people are slow with adopting new technologies
02:54 MisterHatt so perl6 comes out, nobody uses it?
02:54 sri you wouldn't believe how many still use CGI.pm
02:55 MisterHatt wouldn't surprise me
02:55 MisterHatt my company has 2 'perl guys'
02:55 MisterHatt one is me and I barely get perl, the other guy is all about CGI.pm
02:55 sri i don't think perl6 will ever reach production quality
02:55 MisterHatt what do you think of perl5i then?
02:56 sri like it
02:57 sri adding new features to perl5 is a much better option than perl6
02:57 sri you just have to look at python3
02:57 MisterHatt does perl5i even count as 'new features'? it's more like additional features from my understanding
02:57 sri that language is ready for years, but has pretty much no adoption at all
02:58 sri perl6 will be many times worse
02:59 sri what's the difference between new and additional?
03:00 MisterHatt turning it on/off? not much lse I guess
03:00 sri you can turn core features on and off too
03:00 sri say() for example
03:01 sri what i would like to see in core is sub signatures for example
03:01 sri they should have been in perl5 for years
03:02 sri i think the biggest problem we have atm is that perl5 and perl6 are mostly two different communities
03:03 sri and especially the perl foundation favors the perl6 community
03:25 crab i think i might still be using CGI.pm somewhere.
03:26 * tempire force-feeds CGI.pm to crab
03:27 perlrocks Twitter: "ORLite the UTF-8 issue, I think I'll work with this change .... https: / / groups.google.com/d/msg/mojolic​ious/21gHCymDKwo/955OwT6J2GMJ" (ja) --5_1_5 http://twitter.com/5_1_5/status/35902805902491648
03:33 tempire I should be able to sub-class Mojo::Client using "use Mojo::Base 'Mojo::Client';", right?
03:34 sri yes
03:34 perlrocks Twitter: "Welcome guest lecturer's announcement @ yusukebe Hokkaido.pm # 2 4 / 19 will be held! The two-part system of morning and afternoon, morning @ charsbar Mojolicious's hands in the afternoon a talk tech. We welcome beginner's! http://goo.gl/q5YzW" (ja) --hokkaidopm http://twitter.com/hokkaido​pm/status/35904568822857728
03:52 sri su-bzero: oh, completely missed your question earlier
03:52 sri su-bzero: well, if you have a proposal that doesn't break the current api too much sure, why not
04:03 tl joined #mojo
04:05 tholen42 left #mojo
04:11 MojoGuest959 joined #mojo
04:11 MojoGuest959 From: http://news.ycombinator.com/x?fnid=Yli2F8rSPr (1 hits)
04:28 perlrocks Twitter: "@ DailySexTips Wrong Darling. I, ME, Very ... am a sexy, lusty, MoJOLICIOUS ANIMAL!" (gl) --ladkibaazbilla http://twitter.com/ladkibaazb​illa/status/35918166483927040
04:37 GeJ servus everyone.
04:40 GeJ is the list of characters in the <Path> section of $Mojo::Message::Request::START_LINE_RE definitive?
04:41 sri GeJ: why do you ask?
04:41 GeJ I mean, what about {, } and | for example? Is there an RFC that forbids them?
04:42 sri we are indeed following RFCs
04:43 sri feel free to recheck for spec compliance though
04:43 sri the most important one is 2616
04:44 GeJ I've been trying to parse several RFCs (2616 included), but I must be too stupid to grok them.
04:44 * GeJ rechecks
04:51 tempire RFCs have their own language.  If you look @ them long enough, your mind will adjust to the madness.
04:51 crab galician?!
04:52 GeJ Looking at http://www.ietf.org/rfc/rfc2396.txt right now. Section 2.4.3. Excluded US-ASCII Characters
04:52 GeJ there's a group of characters called "unwise" containing at least [ and ]
04:53 GeJ these two are listed in $START_LINE_RE but not some of the others characters in this group.
04:53 GeJ (I confess I'm more interested in the '|' bastard)
04:54 crab sri: could you explain the url handling bit in form_for, please? my @url = (shift); push @url, shift if ref $_[0] eq 'HASH'
04:55 GeJ I'm working on a Mojolicious-powered web service to act (for now) as a bug-for-bug replacement for something that our client wants to get rid of.
04:55 sri route name and hash
04:56 crab oh, it's the form_for login => {foo => 'bar'} => (method => 'post') thing
04:56 GeJ Sadly, the 'query' part of the URL (after the '?') contains a list of '|'-delimited products.
04:56 crab sri: but what does url_for(login => {foo => 'bar'}) mean?
04:57 sri placeholders of course
04:57 GeJ and it took me some time to understand why Mojo was returning '400 Bad request' untill I found $Mojo::Message::Request::START_LINE_RE .
04:57 crab oh, it can take controller, action etc as a hash OR a hashref?
04:57 MojoGuest758 joined #mojo
04:57 MojoGuest758 From: http://news.ycombinator.com/item?id=1277067 (218 hits)
04:57 MojoGuest758 left #mojo
04:58 sri GeJ: oh, that might not be valid http
04:59 GeJ Given that their current version runs on IIS with aspx, I'd argue that it's certainly MS-sanctioned HTTP.
04:59 GeJ ;)
05:00 crab GeJ: | is in the "unwise" list, iirc
05:01 crab i.e. should not occur unescaped in URLs
05:01 GeJ crab: true, as are '[' and ']' (which are acceptable characters in the $START_LINE_RE).
05:03 GeJ I'd understand that adding '|' in the list is not desired ootb. My question would be : what if I wanted to make Mojo accept bad HTTP?
05:04 GeJ what options do I have?
05:05 crab that seems a bit suspicious
05:06 sri monkey patching the parser i guess
05:06 GeJ would it be possible to have 2 regexes and a configuration switch? or maybe make $START_LINE_RE 'our' scoped so that irresponsible people like me can mess with it?
05:06 sri it's really not meant to accept invalid http
05:08 GeJ I know, and I totally understand. It's just that I'm facing a problem and I'm looking at all the available options.
05:08 GeJ but I'll probably go with monkey patching then.
05:08 GeJ thanks.
05:08 crab well, that RE will accept invalid http
05:09 GeJ I'll let you know how the thing runs when it's online. I might even try the toad.
05:09 sri i'm not entirely opposed on being a little more liberal, as long as you can convince me that there are no risks
05:10 sri s/on/to/
05:11 sri that can only be achieved with lots and lots of unit tests though
05:15 crab well, what is the purpose of the checking of request-uri strictly in START_LINE_RE?
05:16 crab it isn't protecting mojo::url->parse from invalid input really
05:16 crab but at the same time it's not at all clear to me that just adding unwise characters to it would be an improvement
05:17 sri how is it not protecting Mojo::URL?
05:17 crab well, most obviously in that it doesn't check that % occurs as part of a valid escape sequence
05:18 sri that's the job of Mojo::URL
05:18 sri the parser makes sure you don't throw binary garbage at Mojo::URL
05:20 crab ok, then extending the list with tests doesn't sound too bad
05:22 crab i've always parsed start-line by biting off the two ends first, then dying if there are spaces or non-ascii in the remainder, otherwise handing it off to the url parser
05:22 crab damn, it sounds gory if i say biting off.
05:25 crab can form_for ever not be used with begin/end?
05:27 sri <form /> seriously?
05:28 crab heh
05:29 perlrocks Twitter: "Mojolicious:: Guides:: Rendering even finished reading. http://bit.ly/hROHaI / / ep but I also think it looks simple and immediate solution, json or can be rendered, Dumper Arekore found a pretty handy looking like a helper." (ja) --tsukamoto http://twitter.com/tsukamot​o/status/35933300782534656
05:41 kaare joined #mojo
05:43 perlrocks Twitter: "After translation with translation memory https: / / sites.google.com / site / tsukamoto / doc / perldoc / mojolicious tried to upload. This is the first time but try it, or translate the new version may be more convenient here." (ja) --tsukamoto http://twitter.com/tsukamot​o/status/35936976444067840
05:54 crab damn, if i telnet to my server i have to type bloody fast to beat mojo's request timeout
05:55 sherr joined #mojo
05:58 MojoGuest867 joined #mojo
05:58 MojoGuest867 From: http://www.google.com/search?q=websocket+irc (1 hits)
05:58 MojoGuest867 left #mojo
05:59 MojoGuest11 joined #mojo
05:59 MojoGuest11 From: http://www.google.com/search?q=websocket+irc (2 hits)
05:59 MojoGuest11 left #mojo
06:23 jwang left #mojo
06:31 rshhh left #mojo
06:32 rshhh joined #mojo
06:34 koban joined #mojo
07:31 AmeliePoulain left #mojo
07:35 su-bzero[A] is now known as su-bzero`
07:40 Christian joined #mojo
07:40 Christian morning
07:40 sri moin
07:52 Christian have mojo something to encode/decode html entities?
07:58 koban Christian: Mojo::Util
07:59 Christian very nice
08:00 AmeliePoulain joined #mojo
08:00 ysyrota joined #mojo
08:06 otaviof left #mojo
08:16 crab git log --oneline --graph rocks
08:32 marcus o/' GIVE ME ALL YOUR LOVING, ALL YOUR HUGS AND KISSES TOO o/'
08:36 yko Christian: see Mojo::ByteStream - http://mojolicio.us/perldoc?Mojo/ByteStream
08:37 yko also take look at it's alternative constructor in the bottom of synopsis
08:37 * yko waves
08:38 yko crab: you may like 'tig' then
08:53 sri damn, i think sparrow is losing my mail :S
08:58 al that's what you get for trusting a pirate with your mail
08:59 sri -.-
09:04 otaviof joined #mojo
09:06 koban left #mojo
09:06 koban joined #mojo
09:41 tholen_ left #mojo
10:11 marcus seen andyg
10:11 purl andyg was last seen on #catalyst-dev 1 years, 98 days, 15 hours, 29 minutes and 34 seconds ago, saying: it never got ported to M::I?  [Nov  4 18:42:21 2009]
10:12 marcus :+/
10:20 jwang joined #mojo
10:21 marcus hai jwang. :)
10:23 jwang marcus: hey how goes?
10:24 jwang :)
10:31 MisterHatt hurr
10:32 marcus jwang: good. trying to get music playing from my squeezeserver to my lamp imac.
10:32 marcus jwang: you??
10:32 crab hurr är det?
10:32 marcus oops, double ??
10:34 jwang marcus: nice. finally getting around to correctly handling a CD by removing the offending content (vs. taking the entire site down ;)
10:34 jwang marcus: CD = cease & desist
10:34 MisterHatt crab: cool swedish there
10:35 jwang marcus: also just ordered some snowboarding stuff on eBay
10:36 crab jwang: what are people CD-ing you about?
10:37 jwang crab: I made a glossary aggregation site a while back. apparently one group sells their glossary for $$ and you're not allowed to re-purpose it unless you pay them a fee
10:37 jwang crab: the site was never that popular so I just took it down at the time. though it did get quoted in a government report
10:38 jwang crab: the funny thing, is that I'm now a member of the organization that sent it to me
10:38 crab heh
10:39 jwang marcus: how do you like squeezeserver?
10:41 spleenjack joined #mojo
10:45 marcus jwang: I like it a lot
10:48 crab what is a lamp imac?
10:48 jwang marcus: nice, I was thinking of installing it a while back but didn't get around to it
10:49 memememomo_ is now known as memememomo
10:51 memememomo is now known as memememomo_
11:51 perlrocks Twitter: "[Perl] [websocket] [Mojolicious] / Mojolicious:: Lite and WebSocket I made using a simple real-time chat - Getting Started with Perl code samples http://htn.to/2sqre7" (ja) --aklaswad http://twitter.com/aklaswa​d/status/36029595459256321
12:39 MisterHatt making me wanna play with IRC bots and mojo
12:40 yko irc bot via websockets?
12:41 MisterHatt mojo::client
12:41 purl mojo::client is too good. and very flexible
12:41 MisterHatt yeah
12:41 MisterHatt was reading what sri did with perlrocks
12:42 MisterHatt pretty simple and interesting
12:43 yko id shouldn't use Client for irc, only for some twitter crawling
12:44 yko i was thinking about bind one of existing irc bot libraries to ioloop
12:45 yko days i was using POE irc i found many plugins already written
12:47 MisterHatt https://github.com/kraih/mojoli​cio.us/blob/master/perlrocks.pl seems to use client and ioloop only
12:47 yko it uses client for fetch stream from twitter
12:47 yko and translate it via google translate
12:47 purl babelfish cannot translate from en to en.  Try translating through English.
12:48 MisterHatt yko: looks like it uses client to handle IRC as well though
12:48 MisterHatt unless I dont get websocket at all
12:49 yko MisterHatt, no. perlrocks reuses client's ioloop to be async
12:50 MisterHatt still client and not websocket then
12:50 MisterHatt unless I REALLY dont get websocket (very possible)
12:50 yko there's no websockets
12:51 yko you need websockets only if you are building something over HTTP
12:51 yko but you actually easy can build irc bot chained to your web application using IOLoop. Just like it's done in perlrocks
12:51 MisterHatt hmm
12:53 yko as i said before, i'd preffer to reuse one of existing IRC libraries. Just need to find if POE has ->tick function and chain it to IOLoop->on_tick of your application
12:53 yko but perlrocks is rally awesome minimalistic irc bot!
12:54 yko MisterHatt: http://mojolicio.us/perldoc?Mojo/IOLoop#on_tick - just like it's documented
12:54 jamesw loop_do_timeslice i think
12:54 yko jamesw: thanks, looking at it
12:54 DaTa MisterHatt: https://data.rbfh.de/p/2011-01-13TwZedE7J this is a patch that adds a lite app into perlrocks for receiving github notifications ;)
12:54 MisterHatt ah I see
12:55 MisterHatt thanks for the clarification
12:55 MisterHatt ok, that is just plain cool
12:56 MisterHatt DaTa: I can see that being VERY useful for work
12:58 yko so actually you acn set up POE IRC bot (years ago i did one, it was perfect)
12:58 yko and instead of doing $poe_kernel->run(); just $ioloop->on_tick(sub { $poe_el->loop_do_timeslice()})
12:59 yko i'm not sure if it would realy works, but it should
12:59 yko s/poe_el/poe_kernel/
13:01 yko i was afraid that Mojo::Message and Mojo::Content stores headers in asset. shame on me
13:07 yko sri++ # for Puscifer
13:08 al left #mojo
13:10 perlrocks Twitter: "http://koorchik.blogspot.com/ - Victor of Tours wrote about the # perl, # mojolicious, # epic, # gpl, # subversion and other konstware" (ru) --shoorick77 http://twitter.com/shoorick​77/status/36049471590768640
13:20 crab "Victor of Tours" sounds like some mediaeval knight
13:20 crab what is puscifer?
13:24 crab oh. it's translating turskyi as "of tours". haha.
13:25 crab might even be right
13:26 esskar2 joined #mojo
13:27 al joined #mojo
13:28 esskar left #mojo
13:29 yko crab: James Keenan  (Tool) side effect
13:35 bellaire joined #mojo
13:36 jamesw left #mojo
13:55 jamesw joined #mojo
14:06 kaare left #mojo
14:14 ispy_ left #mojo
14:24 AmeliePoulain I have Pages controller/model which get page content from the database (nested sets) by 'href' field. Routing - 'route('/:controller/:acti​on/:id')->to('pages#show', id => index)'. When I use 1st level of pages (e.g. site.com/articles) it's easy to find a page by 'href' == 'articles' but I don't know how to parse many levels of pages. I have idea to get a part of the first level and find it in db then get the next part and find it in the 1'st level
14:24 AmeliePoulain childrens and do it for the next "levels" but I think it's no good.
14:30 su-bzero_ joined #mojo
14:35 su-bzero left #mojo
14:40 cstamas left #mojo
14:41 ispy_ joined #mojo
14:45 cstamas joined #mojo
15:06 AmeliePoulain left #mojo
15:14 perlrocks Twitter: "Mojolicious:: Lite and HTML:: FillInForm:: Lite fill in a form - scrawled in http://bit.ly/gzSJdT Tatsumi" (ja) --wPerlnews http://twitter.com/wPerlnew​s/status/36080529673560064
15:14 ato left #mojo
15:21 espent hey, how can i access my dbic model from mojolicious. tried this: http://bit.ly/fki5zm but it doesnt work that well
15:32 dekimsey joined #mojo
15:41 yko should Mojo::URL after ->to_abs loose userinfo?
15:46 Christian nice weekend to you all
15:51 yko bb Christian
15:51 yko oh.. got it now. ->to_abs gets authority from base url which overrides any existing userinfo
15:51 yko not sure if it's supposed to work that way
15:52 Christian left #mojo
15:54 koban left #mojo
16:01 otaviof left #mojo
16:11 rshhh left #mojo
16:12 rshhh joined #mojo
16:14 xantus joined #mojo
16:15 xantus is there an easy way to limit mojo client's connections to a host
16:16 xantus ...a site said I had 9000 simlu connections
16:17 kaare joined #mojo
16:25 su-bzero` is now known as su-bzero[A]
16:27 su-bzero_ left #mojo
16:28 su-bzero joined #mojo
16:36 xantus bbl
16:36 xantus left #mojo
16:41 ato joined #mojo
16:43 su-bzero_ joined #mojo
16:44 MojoGuest959 left #mojo
16:47 su-bzero left #mojo
17:01 otaviof joined #mojo
17:03 rupertthehacker joined #mojo
17:08 sri damn, nokia had 1500 people work on symbian
17:22 yko oO
17:22 yko they don't work anymore on it?
17:24 yko actually i don't like nokia last ~5 years. Last good model was 7610... i still have one working somewhere
17:26 chansen sri: where did you get that number?
17:39 a|newkirk[assoc] joined #mojo
17:42 mattastrophe left #mojo
17:46 vel joined #mojo
17:47 metaperl1 joined #mojo
17:48 marty left #mojo
17:48 marty joined #mojo
17:50 metaperl left #mojo
17:51 sri chansen: http://www.hs.fi/talous/artikkeli/1135263739290
18:04 marcus Nokia actually uses a much larger percentage of its revenue on R&D than apple does.
18:06 chansen I still doubt that's a correct number (I know a former nokia employee that worked on Symbian OS and the number of people when he worked there wasn't near that figure)
18:10 sri i guess their only problem is cultural, having the hardware people make all decisions just doesn't work anymore
18:10 sri at least when you're trying to build your own os
18:11 tholen joined #mojo
18:19 chansen I doubt a partnership with MS will save them ;P
18:22 tholen left #mojo
18:30 marcus Probably.
18:31 sri yea
18:32 marcus But ms actually puts some strict demands on the hardware.
18:32 marcus We have a windows phone 7 for testing purposes. It is not bad.
18:32 marcus A bit immature.
18:34 chansen How does is compare to Android or iPhone?
18:34 tholen joined #mojo
18:35 chansen do they have a appstore?
18:36 ysyrota left #mojo
18:37 marcus Yes.
18:37 marcus I think it is about equiv to iphone 1. Gen
18:38 marcus No multitasking, no copy paste, no norwegian keyboard :)
18:38 marcus But pretty nice ui. Lots of good games because of xbox arcade.
18:44 chansen ok
18:57 rupertthehacker left #mojo
19:04 ispy_ left #mojo
19:06 perlrocks Twitter: "Wow! #mojolicious couldn't make it any easier to build a proxy server.... 5 lines of code and I'm getting what I want. Awesome!" --jamiepopkin http://twitter.com/jamiepopk​in/status/36138913466814465
19:09 mmcleric left #mojo
19:10 mmcleric joined #mojo
19:12 al left #mojo
19:18 mmcleric left #mojo
19:18 mmcleric joined #mojo
19:24 sri left #mojo
19:24 sri joined #mojo
19:24 mtths left #mojo
19:25 mtths joined #mojo
19:32 stephen joined #mojo
19:33 stephen left #mojo
19:38 stephen joined #mojo
19:45 mmcleric left #mojo
19:45 spleenjack left #mojo
19:45 mmcleric joined #mojo
19:57 GitHub131 joined #mojo
19:57 GitHub131 mojo: master Sebastian Riedel * b31a892 (1 files in 1 dirs): added more faq answers - http://bit.ly/eho43N
19:57 GitHub131 left #mojo
19:58 sri marcus, tempire: sounds good?
20:05 chansen I don't agree (and I know a lot of other CPAN developers that will be offended) with "Components specifically optimized for usability and development speed are not as common on CPAN as you might think."
20:07 sri feel free to suggest a better alternative to say it
20:07 JonChisciotte joined #mojo
20:07 JonChisciotte left #mojo
20:08 sri but i do think usability is not very common on cpan
20:09 chansen What do you mean with usability?
20:09 sri you disagreed with me, what do you think i mean? :)
20:10 chansen I know what it mean, it's you saying that CPAN modules isn't optimized for it
20:14 sri i don't see this discussion ending well
20:14 bellaire maybe you could include both sides?  if you optimized for development speed and usability, what did you sacrifice to get it?
20:15 chansen sri: if a module is documented to implement X and implements X, isn't that usability
20:15 sri no it is not
20:15 purl i already had it that way, sri.
20:15 sri but i'd rather not continue this discussion
20:15 sri there is no way we could agree on something
20:17 sri bellaire: the tradeoff is always flexibility
20:17 chansen at-least you could elaborate on what your definition of usability is
20:17 sri http://en.wikipedia.org/wiki/Usability
20:18 sri a focus on providing a great user experience
20:19 bellaire sri: just thought maybe by saying you sacrificed some flexibility for the things you optimized for, might make it seem balanced.  course people will probably find something to be offended about regardless...
20:19 sri right, people will find something else ;p
20:20 GitHub8 joined #mojo
20:20 GitHub8 mojo: master Sebastian Riedel * 81e7dfe (1 files in 1 dirs): mention tradeoff - http://bit.ly/hhz0Px
20:20 GitHub8 left #mojo
20:21 sri bellaire: not a bad idea though
20:21 sri also goes well with the next sentence
20:22 ispy_ joined #mojo
20:22 ispy_ left #mojo
20:23 yko i'm thinking about configurable dafault package to display at perldoc
20:24 yko $self->plugin('pod_renderer' => {package => __PACKAGE__}) if $self->mode eq 'development';
20:24 sri yko: sure
20:24 yko ^--  that could be nice if more than one developer is working on project for example
20:25 sri {start => 'Catalyst::Manual'}
20:30 bellaire So, if I want to send a file using $self->render(data => $blob, format => $known_type), it'll get chunked/streamed automagically?
20:30 sri nope
20:30 bellaire damn ;)
20:30 sri $blob is in memory already
20:30 sri you would use a Mojo::Asset::File
20:31 yko render_static does chunked?
20:32 sri not chunked, but streams in chunks
20:33 yko I prefer just to store files in public and serve them via static if needed. if they are not top secret for sure.
20:33 sri $self->res->content->asset(​Mojo::Asset::File->new(path => '/foo.txt')); $self->rendered;
20:33 sri this is pretty much what the static dispatcher does
20:33 yko wow.
20:33 yko sri++
20:33 yko thanks for that hint, should be very useful
20:34 bellaire mine are top secret ;) thanks
20:34 sri feel free to turn it into a recipe :)
20:34 yko i also found that having own test library per application is a good stuf
20:35 yko jsut something that loads Test::Mojo and imports Test::More
20:35 yko with such stuff you can implement addictional tests in Test::Mojo
20:36 yko like login_ok is a must - saves much time on testing private areas
20:36 sri anything we could reuse in core?
20:39 perlrocks Twitter: "Answering two more frequently asked questions about #mojolicious today. :) http://t.co/4l1dZ7Z http://t.co/aYhG7j6 #perl" --kraih http://twitter.com/kraih/status/36162452282413056
20:40 yko unsure. i embedd such library im my default skeleton at the moment
20:40 yko looks like https://gist.github.com/212aa131d04bf79b9875
20:41 yko it fixes $0 for json_config and other stuff, set up application instance and allows add new methods to Test::Mojo
20:41 sri the $0 problem with json_config can be annoying
20:43 yko yeah, i was building applications with predefined file => __PACKAGE__.conf for a while
20:43 yko but then dropped it and moved everything in test library
20:43 perlrocks Twitter: "Mojolicious is the OpenBSD of web development frameworks. http://bit.ly/gt29zh #sane #defaults #clean #abstraction" --obfuscurity http://twitter.com/obfuscuri​ty/status/36163441764868096
20:44 yko just use Test::MyApp; my $t = Test::MyApp->new; and you have ready to go test suite
20:44 yko 50% of test files beginning with $t->login_ok; which actually includes 2-3 or even 4 tests
20:45 yko test becames clean, much uinderstable and i'm not so lazy to make them now :D
20:46 sri you are using the latest github version right?
20:47 yko 1.1 at least
20:47 sri and you did not do that thing you pasted to workaround a config file bug?
20:47 yko maybe 1 or two days old
20:47 sri ah ok
20:48 yko sorry, did not get your last sentence
20:48 sri i don't understand the import hack
20:49 sri $ENV{MOJO_EXE} = 'script/my_app'; for example
20:50 yko sri: json_config autodetects config file name via MOJO_EXE: https://github.com/kraih/mojo/blob/master​/lib/Mojolicious/Plugin/JsonConfig.pm#L25
20:51 yko so i just load plugin like $self->plugin('json_config' => {ext => 'conf'}
20:51 yko well, i just like conf more than json for configs :)
20:51 sri yes, but is MOJO_EXE really neccessary?
20:52 yko if you load aplpication from test file t/01.basic.t it tries to load 01.basic.json, doesn't it?
20:52 sri you didn't try?
20:52 yko maybe that from older version, i will check now
20:53 yko (and upgrade mojolicious also)
20:55 sri ok, can't be detected
20:55 sri just tried
20:56 yko sub startup { .. $self->plugin('json_config'); .. }; prove -lrv => Config file "/tmp/json/basic.json" missing, maybe you need to create it?
20:56 sri just ./script/my_app test works
20:56 yko that's why such a hack
20:56 yko ohmmm
20:57 yko ye, works.
20:59 yko damn, TextMate snippets format looks terrible :p
20:59 sri yay for xml
21:00 yko next mojo.vim version will contain stippets for snipmate.vim (for TagHelepers also
21:00 sri neat
21:00 yko *snippets
21:02 sri think i hate the config file detection code :/
21:02 sri so many ways to break it
21:02 yko oh, wanna ask: does it naturally that link_to 'foo' => 'bar' consider foo (link to!)  as text description and 'bar' as url?
21:03 sri the text foo links to bar
21:04 yko i thought link_to "to where is this link?" => "some description"
21:04 yko nevermind
21:04 a|newkirk[assoc] left #mojo
21:04 yko just thinking how long tag helpers will be experimental module - people (me to) afraid to use them
21:05 sri the more feedback i get the shorter it will be ;)
21:05 yko but tag heleprs are so damn neat! it's so tempting jsut to take and use them
21:05 yko hmm. i should send at least one feedback a day
21:05 yko and two on weekends :)
21:07 sri ;p
21:07 sri i guess link_to makes sense both ways
21:07 yko btw, you can have feedback right now: i'm using them often (in spite of fear :)  and i think that tag heleprs are awesome
21:07 sri seemed more natural this way i guess
21:08 yko stuff i havent used yet is select_field
21:08 yko most complicated tag helepr i guess :)
21:08 sri yea, it was even broken until recently -.-
21:09 yko but it's only because i'm not using select fields at all
21:09 sri actually i'm already so afraid to make changes to tag helpers i might as well make them stable
21:10 yko sri++
21:10 sri then again, javascript and stylesheet just recently changed again
21:10 yko the only think i'd change is link_to 'href' => 'desc' but ignore me - youll have thousands of angry users here if you change it :)
21:11 sri link_to foo => {some => 'thing'} => 'Foo' is awkward too
21:12 sri vs. link_to Foo => 'foo' => {some => 'thing'}
21:12 yko ohm... i was thinking to add correct syntax highliting inside of javascript and stylesheed tag helpers. thank you for reminding me it
21:13 yko sri, actually yes, it have sense. like link_to Thing => some tech specs
21:14 yko I just thinking from other (tech) side. that's why this duscussion
21:14 yko like link_to spec => your human-readable stuff :)
21:15 yko anyway, it's time to say that mojolicious is awesome
21:15 sri :)
21:25 aiglenoir left #mojo
21:26 aiglenoir joined #mojo
21:27 aiglenoir left #mojo
21:31 aiglenoir joined #mojo
21:31 kaare left #mojo
21:32 aiglenoir left #mojo
21:36 aiglenoir joined #mojo
21:38 al joined #mojo
21:38 aiglenoir left #mojo
21:41 aiglenoir joined #mojo
21:53 sri yko: how good is your vim highlighting?
21:53 bellaire left #mojo
21:56 sri do you have templates in the data section working?
21:57 sri hmm, looks like it
21:57 purl No it doesn't, shut your hole!
22:00 tholen left #mojo
22:00 al left #mojo
22:04 aiglenoir left #mojo
22:04 aiglenoir joined #mojo
22:04 aiglenoir left #mojo
22:09 aiglenoir joined #mojo
22:11 aiglenoir left #mojo
22:14 aiglenoir joined #mojo
22:19 dekimsey left #mojo
22:21 al joined #mojo
22:29 vel left #mojo
22:33 MojoGuest252 joined #mojo
22:33 MojoGuest252 From: http://www.networkedblogs.com/33Knp (63 hits)
22:41 sri yko: hmm, data section highlighting seems not to work for me :(
22:43 yko it's disabled by default
22:43 yko do :let mojo_highlight_data = 1
22:43 yko end then :e! or so
22:44 marcus sri: switching to vim?
22:45 sri yko++
22:45 marcus Hating mail on my ipad now that I have Sparrow on the mac
22:45 sri just playing a bit with it
22:45 yko good night
22:46 sri nn
22:46 sri and please add css/js highlighting :D
22:46 sri managing addons is really nice with pathogen
22:46 sri i just can't get used to buffer management
22:47 sri my mind works better with tabs
22:48 robn joined #mojo
22:52 robn left #mojo
22:54 marcus of all the things I've lost, I miss my mind the most.
22:56 omega and for such a long time it's been missing? :)
23:09 spleenjack joined #mojo
23:20 sri marcus: i'm a little angry with sparrow, it seems to have lost some mails :/
23:23 MojoGuest252 left #mojo
23:29 sri and of course the vim site is broken when i want to look up some stuff -.-
23:29 mmcleric left #mojo
23:29 aiglenoir left #mojo
23:30 aiglenoir joined #mojo
23:35 daviddelikat joined #mojo
23:37 aiglenoir left #mojo
23:39 mmcleric joined #mojo
23:50 aiglenoir joined #mojo

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