Camelia, the Perl 6 bug

IRC log for #mojo, 2010-07-14

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

All times shown according to UTC.

Time Nick Message
00:12 xantus sri: you around?
00:12 xantus I have upload to couchdb using mojo client working, but its ugly
00:12 xantus https://gist.github.com/76da6a824defd8cb8f08
00:13 xantus I could patch client to check for an asset and auto use that as the content..
00:21 alex joined #mojo
01:23 xantus woot
01:26 xantus bought comet.im
03:07 ashleydev joined #mojo
03:26 janus joined #mojo
03:35 ask joined #mojo
04:05 yakudzo joined #mojo
04:12 alex joined #mojo
04:22 und3f[A] joined #mojo
04:33 alex joined #mojo
04:36 sri xantus: sure, asset check would make sense
04:37 sri xantus: i've readded websocket draft 75 support btw :)
04:37 sri so we now do both
04:45 alex_ joined #mojo
05:21 koban joined #mojo
05:26 GitHub145 joined #mojo
05:26 GitHub145 mojo: master Sebastian Riedel * 44a9909 (17 files in 4 dirs): cleanup - http://bit.ly/9IHe7a
05:26 GitHub145 left #mojo
05:38 ashleydev joined #mojo
05:43 tholen joined #mojo
05:43 ask joined #mojo
05:49 Foxcool joined #mojo
05:53 ashleydev joined #mojo
06:01 yko joined #mojo
06:08 Su-Shee joined #mojo
06:15 Su-Shee_ joined #mojo
06:21 tempire joined #mojo
06:53 koban left #mojo
06:53 koban joined #mojo
07:04 yko Sometimes sri reminds me captain from Romantically Apocalyptic
07:21 ask joined #mojo
07:31 xantus sri: sweet!
07:37 arthas joined #mojo
07:42 xantus ok, it should be easier than this to use mojo client with couch :/
07:49 sri xantus: well, uploading a file from disk is not *that* common
07:50 sri just imagine how hard it would be with lwp ;p
07:52 xantus well, also put updates
07:52 xantus the body content is straight json
07:53 xantus same dance with asset memory and manual decode
07:53 xantus er, encode
07:54 xantus unless I missed something
07:55 sri yko: lol, that captain guy seems very sympathetic!
07:55 crab sri: i tried out that success->dom->search->each thing and it's very nice.
07:56 sri xantus: you can just pass a scalar to get/post/put
07:57 sri $client->put('http://mycouch', '{for => 'teh lulz!'}');
07:58 sri ops, thats not json
07:58 ysyrota joined #mojo
07:58 sri $client->put('http://mycouch/', '{"for": "teh lulz!"}');
07:58 xantus oh yeah
07:58 sri i blame the heat -.-
07:59 xantus cool
08:00 sri i wouldn't mind some magic trick to automatically generate json from a hashref/arrayref, but we already use hashref for headers :(
08:01 sri not sure $client->put_json('http://mycouch', {for => 'teh lulz'}) would make sense
08:02 sri totally depends how common a case becomes and how much extra work it would save
08:03 sri after all we do ->post_form too
08:03 xantus yeah
08:03 xantus post_json?
08:04 xantus :p
08:04 sri post_json, put_json...
08:04 sri someone needs to seriously investigate what would make sense and propose the api with examples
08:05 sri and in the case of couchdb, a specific wrapper around the client might make more sense
08:11 sri i'll be using couchdb for a big project soon, so support for it should become awesome anyway over time :)
08:11 Alias do_json
08:11 Alias send_stuff
08:12 * sri throws a couch after Alias
08:12 * Alias ducks the couch, then is hit by the couch with the duck embedded in it
08:13 Su-Shee smother him with a cushion. ;)
08:13 xantus I have a class that wraps a couchdb doc
08:14 xantus you can set/get/unset/attach_file/has_attach​ment/delete/is_dirty/set_dirty/save
08:15 xantus oh, I actually found that one odd bug
08:16 xantus where client fails and it looks new when dumped
08:16 sri :o
08:17 xantus if you use too many clients
08:17 xantus it bombs on too many open files (epoll)
08:17 xantus i'm using a singleton now
08:18 sri interesting
08:20 crab ->post($url, json => { x => "y" })?
08:20 sri that might be possible
08:22 sri i think right now we allow ->post($url, 'Content-Type' => 'text/lulz') too, but it's undocumented and untested
08:25 xantus so put( $url, file => 'path' ) or put( $url, json => $obj )
08:25 xantus that'd be nice
08:25 sri agreed
08:28 xantus btw, escaping filenames for use with readpipe is tricky
08:28 sri yko: haha, romantically apocalyptic rocks, i love that captain dude
08:29 dotan joined #mojo
08:30 xantus ...I'm making thumbs of videos with ffmpeg and putting them in couch as attachments to the file doc :)
08:31 xantus but the readpipe is for the shell cmd 'file $file'
08:32 xantus there's probably a better way...
08:32 ysyrota joined #mojo
08:40 zloyrusskiy joined #mojo
08:42 zloyrusskiy Hi all, i found a bug in Mojo::DOM, where can i report about it?
08:42 sri zloyrusskiy: right here
08:43 zloyrusskiy good, it using selectors for searching elements
08:43 zloyrusskiy but it's a case-sensitive
08:44 sri elements are not case sensitive
08:44 zloyrusskiy and in documentation to CSS3: All Selectors syntax is case-insensitive within the ASCII range (i.e.    [a-z] and [A-Z] are equivalent), except for parts that are not under  the    control of Selectors. The case sensitivity of document language  element    names, attribute names, and attribute values in selectors depends on  the    document language. For example, in HTML, element names are    case-insensitive, but in XML, they are case-sensitive.
08:44 zloyrusskiy i talking about tags
08:44 zloyrusskiy look at first example at http://clubs.ya.ru/4611686018427388089/​replies.xml?item_no=1520&ncrnd=4935
08:45 zloyrusskiy try to execute it with http://localhost:3000/title/?url=http://google.com
08:45 garfield [ Вопрос про Mojolicious - Perl - я.ру ]
08:45 und3f joined #mojo
08:46 zloyrusskiy google.com returns result with tags in upcase and this example doesn't work
08:47 sri zloyrusskiy: which version are you using?
08:47 zloyrusskiy with sites with lower-case tags this example works right
08:48 zloyrusskiy i've tested with Mojolicious 0.999926
08:48 sri ok, github version has a test case for uppercase tags
08:49 sri so i suspect its already fixed
08:49 zloyrusskiy ok, thanks
08:49 zloyrusskiy P.S. your Mojolicious module is Awesome!!!
08:50 sri thanks :)
08:51 sri http://github.com/kraih/mojo/​blob/master/t/mojo/dom.t#L158 # here's the test case btw
08:51 garfield [ t/mojo/dom.t at master from kraih's mojo - GitHub ]
08:51 garfield http://xrl.us/bhsrad
08:51 sri going all out with html1
08:56 xantus http://jsdo.it/
08:56 garfield [ jsdo.it - share JavaScript, HTML5 and CSS ]
08:56 xantus (avoid using the facebook login) its broke
08:56 sri heh
08:59 xantus see if you can join me http://jsdo.it/xantus/jbsa/edit
08:59 garfield [ jsdo.it ]
09:00 sri nothing to click on
09:01 xantus hmm
09:09 yko sri: that last strip and '...blastwave' - that's how i imagine you
09:09 sri :D
09:13 zloyrusskiy left #mojo
09:28 spleenjack joined #mojo
09:34 spleenjack joined #mojo
09:42 yko joined #mojo
10:00 Su-Shee I would do a serializer via "format" and the renderer?
10:01 sri what kind of serializer?
10:01 Su-Shee json, csv..
10:01 sri json is built in
10:02 sri ->render(handler => 'csv'...)
10:02 sri thats how you would do it if you add a csv renderer
10:02 yko :)
10:02 sri format is really just about the content type
10:02 Su-Shee ah.
10:03 Su-Shee so it would be handler=> 'csv', format => 'text' essentially?
10:03 sri if thats what you want sure :)
10:04 sri you can also have your handler set a content type itself
10:04 Su-Shee hm, I'll have to see what looks better in code later that other devels see it right away..
10:04 sri but i wouldn't go that far generally, ->render(data => MySerializer::doit($data));
10:05 Su-Shee I need like text, csv, pdf, json, vcard, xml, html and probably 4 others I forgot right now. ;)
10:05 sri it only really makes sense if that format is your everyday business
10:05 sri then it might make sense :)
10:06 Su-Shee we're keeping our system extremely open for "want to stuff something in?" and "let's get some data out" ;)
10:06 sri i'm sure there are quite a few examples on github already
10:07 sri in general handlers are very free to do whatever they want, you just tell the renderer which one to invoke with the handler stash value
10:07 Su-Shee food, then docs again. ;)
10:08 sri and they in turn have full access to the stash
10:08 Su-Shee I'll have to do it with one example later, then I'll probably get it.
10:21 vti pastelicious has an json,html,raw example
10:29 dotan joined #mojo
10:39 GitHub55 joined #mojo
10:39 GitHub55 mojo: master Sebastian Riedel * 8473569 (3 files in 3 dirs): fixed Mojo::DOM to support escaped selectors - http://bit.ly/duIip6
10:39 GitHub55 left #mojo
10:39 sri <div id="♥"> anyone?
10:39 sri with html5 those will happen :)
10:40 sri and we can now handle them
10:41 sri the selector is simply #\2665
10:45 vti sri++
10:45 dotan sri: the single benefit I can think of for relaxing ids in html5 is that now you can write: <div id="<%= $my_numeric_database_id_suck_it_html_4 %>"> and it validates...
10:46 dotan but maybe I just have no poetry left in my soul...
10:47 yakudzo yes what benefits of that ?
10:49 dotan numeric ids without prefixes are sorta useful, characters you need to escape in your CSS - not so much.
10:50 yko dotan: usualy i just use <div dbid='<%= $item->col('id') %>'>... :D
10:51 koban joined #mojo
10:51 dotan :o !
10:51 yko that's for storing numeric database id's
10:53 yko db id doesn't act in css selectors i think, only for jquery, etc
10:53 Su-Shee ha. now I have my feature matrix.
10:55 dotan yko: I get it. It's a nice idea (there's probably a hopelessly complex way to do that "correctly" in html5...) - I just used id="a1234" and added/dropped the prefix in my javascript.
11:04 yko dotan: i did just the same, but then once made some tests. Using addictional attrs just works with jquery, which i usually use. And yes, it's not valid, just works :D
11:06 GitHub65 joined #mojo
11:06 GitHub65 mojo: master Sebastian Riedel * 4a3a374 (2 files in 2 dirs): improved unicode support in Mojo::DOM again - http://bit.ly/aqZsVb
11:06 GitHub65 left #mojo
11:07 sri .♥ is actually a valid class selector :)
11:07 yko madness :D
11:08 dotan http://mathiasbynens.be/notes/html5-id-class
11:08 garfield [ The id attribute just got more classy in HTML5 · Mathias Bynens ]
11:08 Su-Shee pretty! I can write hippie css now!
11:08 sri dotan: thats what motivated me :D
11:09 yko i don't like using unicode in selectors as i dont like using unicode in perl function names :D
11:10 sri i just want to be able to scrape every html5 site out there with Mojo::DOM :)
11:10 Su-Shee me neither, but the thought counts and looks pretty :)
11:10 yko yep :D
11:10 yko sub привет { скажи "Hello World" } isn't good idea :D
11:11 sri sites like reddit and hackernews will use it to screw around with web scrapers i bet
11:11 Su-Shee but sub spread_♥ {} is :)
11:11 sri romantic code :o
11:11 yko hehe i don't see that char in my font :p
11:11 sri I WROTE YOU A SCRIPT! ♥
11:11 Su-Shee well nerds have to express their love somehow :)
11:12 Su-Shee sri: *haha* YES!
11:12 yko :DD
11:12 yko lol, yes that's fun
11:13 yko damn, i totally need perl woman on #ru.pm :\
11:14 yko we had one, but now she's gone :/
11:14 Su-Shee I WROTE YOU FLOWERS! ⚘
11:15 Su-Shee yko: there's like maybe 30 on the entire planet. ;)
11:16 yko I need two or three russian-speaking. Please send :))
11:17 sri endangered species
11:17 yko God made woman fom a rib. We can make perl-woman from ...\ what?
11:19 * sri sets the creationist on fire to prove evolution
11:21 xantus_ joined #mojo
11:59 Su-Shee yko: from =~.
11:59 sri http://github.com/creationix/nstore # hmm...node.js now has its own key value store
11:59 garfield [ creationix's nstore at master - GitHub ]
11:59 yko Su-Shee: ?
12:00 yko Ah. I got it
12:00 Su-Shee yko: the perl version of a rib ;)
12:00 yko aha
12:00 Su-Shee yko: just add the approriate pattern on the right ;)
12:06 dotan joined #mojo
12:21 dotan sri: isn't html5 a superset of every document currently found on the web? Doesn't it have a parsing model that explains how the DOM looks like for basically any string of characters?
12:22 Su-Shee little decision helper for more selling points of mojo: http://www.innoq.com/blog/st/2010/07​/rest_litmus_test_for_web_frame.html
12:22 garfield [ REST Litmus Test for Web Frameworks - Stefan Tilkov's Random Stuff ]
12:22 garfield http://xrl.us/bhsrsa
12:23 sri dotan: html5 as a whole is monstrous and adds tons of stuff that has never been done before
12:24 sri when i say html5 i also mean css3
12:25 sri parsing html5 is actually really easy
12:25 sri css3 not so much
12:25 Su-Shee sri: "web 3.5" ;)
12:25 sri ;p
12:25 Su-Shee well let's be faster than tim o'reilly this time. ;)
12:27 sri dotan: my focus for Mojo::DOM is not html5 exclusive, ideally i want to be able to parse everything that even remotely looks like xml
12:27 dotan sri: html5 is a marketing label, so its meaning changes depending on who is using it. So for example Apple will say it includes all the CSS transition stuff, Mozilla will pick a different subset of features.
12:27 nagual joined #mojo
12:29 dotan sri: so all the horrible tag soup on pages where people publish important bioinformatic data, which I scrape with HTML::Tree (and XPath!), will Mojo::DOM choke on that?
12:29 sri maybe, maybe not :)
12:29 dotan I should write test cases, but looking at the source of those pages makes me want to cry ;)
12:29 sri some crazy tests already in dom.t
12:30 sri we do html1!
12:31 sri it's so liberal chances are high that it will just work with your data
12:31 sri only thing i would like to improve is the ability to repair broken structures
12:31 sri missing tags and stuff
12:32 dotan I recall many irregular tables and links filled with javascript. And then, merciful oblivion...
12:32 sri but i don't think it's possible to crash the parser at all
12:33 sri just try it
12:33 dotan Can you dump the parse results?
12:33 sri easy enough
12:33 sri just try it ;p
12:34 sri $dom->tree
12:35 vti you should add js-like dom support too :) getElementById, insertChildBefore, findElement etc
12:36 sri how about no? ;p
12:36 sri nobody uses those anymore
12:37 Su-Shee what for? those aren't usally exposed anymore but hidden behind the js framework..
12:37 sri you got damn css3 selectors!!!
12:37 Su-Shee sri: AND HOPEFULLY NEVER HAS AGAIN.
12:37 vti you're boring
12:37 Su-Shee vti: no, we're extremely cool and avangarde. it just _looks_ boring from the outside.
12:38 Su-Shee avant.
12:38 sri but i might add a way to expose the SAX events used inside Mojo::DOM
12:38 dotan I hate that Java-head DOM API. The first thing any Javascript library does is find better names for those, even once getElementById got universal support..
12:39 sri we simply don't have them at all
12:39 dotan yay!
12:39 dotan :)
12:40 sri css3 or die!
12:40 sri someone should add more pseudo classes though *hint hint*
12:41 sri perl -MMojo::DOM -MData::Dumper -e 'warn Dumper(Mojo::DOM->new->par​se("<div>lalala</div>"));'
12:41 sri thats the internal structure for you slackers
12:41 sri very simple
12:42 sri i would have liked it to be even simpler, but reversibility is not free
12:43 vti somebody has a patent on reversibility? ;p
12:44 * sri calls the kgb and tells them the location of vti
12:45 vti somebody is knocking on the door
12:46 vti *at the
12:46 sri http://dev.w3.org/html5/html4-differences/ # for those interested, html5 is not based on sgml
12:46 garfield [ HTML5 differences from HTML4 ]
12:53 sri Su-Shee: re your rest litmus test, i'm not sure automatically calculating etags is the right thing to do
12:58 Su-Shee it's just an inspiration. you could add "RESTful foobar handling of abc" to your list of features.
12:59 Su-Shee if I have $r->route('/volkswagen/de/mem​ber/:id')->to('member#show');
13:00 Su-Shee how do I fetch my id in Member::show?
13:00 sri it's all in the stash
13:00 sri the stash is always the answer :)
13:01 sri i thought the routing guide made that very clear
13:01 yko :)
13:01 Su-Shee I know where it is but not I can grab it?
13:01 sri $self->stash('id')
13:02 sri or even $self->param('id') where all incoming parameters (query/body/route) get combined
13:03 Su-Shee ah. no, the routing guide didn't really make that clear.. you're pretty much assuming your reader know how web frameworks of this kind generally work. :)
13:03 yko route params are goint to stash also?? Oo
13:03 sri oO
13:04 sri you meant to ask params right?
13:04 sri they always went to the stash
13:04 sri since day 1 ;p
13:05 yko i ment params. yes
13:05 Su-Shee sri: it's somewhat implicitly clear if I think really hard and remember that the stash is something hashy which I can manipulate everywhere.
13:06 * Su-Shee has two working routes now. o/ :)
13:09 GitHub91 joined #mojo
13:09 GitHub91 mojo: master Sebastian Riedel * 087cad6 (1 files in 1 dirs): added stash example - http://bit.ly/a754Jn
13:09 GitHub91 left #mojo
13:09 sri Su-Shee: that might help the next reader :)
13:11 vti we just translated that piece of docs!
13:11 sri maybe i shouldn't call it action so early in the guide
13:14 GitHub115 joined #mojo
13:14 GitHub115 mojo: master Sebastian Riedel * 3b2dc17 (1 files in 1 dirs): cleanup - http://bit.ly/aZRB9S
13:14 GitHub115 left #mojo
13:15 sri vti: btw. ;p
13:16 yko i think sri dont like russians
13:16 * sri loves russia!
13:17 sri it will be the last country my terminators wipe out
13:19 vti sri loves nobody
13:20 sri "we robots have no feelings, sometimes that makes me very sad :("
13:20 sri hmm, maybe i've overused that quote a bit :/
13:24 sri http://www.couch.io/ # haha
13:24 garfield [ CouchDB 1.0 Released ]
13:26 sri but they didn't use trajan...thats not a movie poster!
13:29 spleenjack1 joined #mojo
13:30 su-bzero joined #mojo
13:51 alex joined #mojo
14:06 ashleydev joined #mojo
14:23 Su-Shee would I write a renderer helper if I want a nice mechanism to attach js and css to specific templates?
14:23 koban joined #mojo
14:24 sri Su-Shee: sure
14:24 yko Su-Shee - attach - you mean include?
14:25 sri Su-Shee: or you just use a partial template
14:25 Su-Shee yko: ideally yes. but I want to keep it separately on the server or I get insane.
14:25 sri <%= include 'lulz' %>
14:25 yko sri++
14:25 sri templates/lulz.html.ep
14:25 yko but why not just link it like <script ...> ?
14:26 sri what yko said
14:26 sri i would just use them as static assets from the layout
14:27 Su-Shee wow, I just really had forgotten about script and link. :)
14:27 yko hehe, but sometimes you might whant to include different resources for different templates
14:27 sri Oo
14:28 yko for example, template foo.html.ep requires BIG_LARGE_HUGE.css
14:28 yko and template bar.html.ep - doesn't
14:28 sri btw. i can confirm the file descriptor thing in Mojo::Client
14:29 sri when you just keep cloning a client instance you'll run out of file descriptors
14:29 sri it's not a bug though
14:30 sri simply every client having its own keep alive queue
14:33 yko Su-Shee: you may like linked_content plugin, mb i should rename it somehow
14:34 yko http://bit.ly/csBl5d *self-pr*
14:34 garfield [ lib/Mojolicious/Plugin/LinkedContent.pm at master from korshak's mojolicious-plugin-linkedcontent - GitHub ]
14:35 vti sri: oO
14:35 sri vti: Oo
14:35 vti sri: OO
14:35 sri X_X
14:36 yko YOU KILLED HIM!
14:36 yko :'((
14:36 Su-Shee sri: you're keeping the "perl 5.8.x" thing up?
14:37 sri Su-Shee: for now
14:37 Su-Shee *pew* I might get near to something I could call a "migration path" finally.
14:38 sri we just bumped it from 5.8.1 to 5.8.7
14:38 Su-Shee This is perl, v5.8.9 built for x86_64-linux
14:38 Su-Shee ;)
14:39 sri thats what most use :)
14:39 sri it's mostly redhat enterprise holding us back
14:39 vti and my mipsel router, it has 5.8.8
14:40 Su-Shee poor $coworker is working on updating our apache 1 and mod_perl 1 and perl 5.8.9.. ;)
14:41 sri good old times, 10 years ago, those were the days...
14:43 Su-Shee well this app is constantly running since 1997.
14:44 Su-Shee and we signed a contract until 2024 recently. so it has to make at least another decade. ;)
14:49 koban left #mojo
14:58 yko sri: can i force mojolicious render template without calling action?
14:58 yko hmm. i think i got it...
14:58 yko ignore me
14:58 Su-Shee didn't I just read that in the Mojolicious::Lite docs...
14:58 sri get '/foo' => {template => 'lulz'};
14:59 * Su-Shee dares to go outside now. if I never say anything again, please call 112. ;)
14:59 yko yes, exactly
15:05 vti sri: get '/foo' => 'lulz'; too?
15:06 sri vti: oh right :)
15:06 * sri blames the heat
15:07 sri i've written that example like 100 times :S
15:15 yko gg
15:19 * yko got some icecream.
15:23 sri :o
15:55 _mpu joined #mojo
15:55 Su-Shee BAH :/
15:55 Su-Shee how's the weather in russia?
15:56 sri wait, it's not snowing all the time in russia?
15:56 Su-Shee yes, but it could be a sunny icy day or snow storm.
16:01 vti ye, we are freezing
16:01 Su-Shee vti: can I move in?
16:01 Su-Shee wait, why am I asking! I INVADE!
16:01 vti Su-Shee: sure, take a trans-siberia railway
16:01 Su-Shee MOVE OVER.
16:02 Su-Shee vti: have you read that china is planning to connect everything between beijing and berlin with high speed trains? :)
16:02 vti oO
16:02 vti no
16:02 vti i am in kiev now anyway ;p
16:03 Su-Shee what? forever? or just for vacation?
16:07 vti until october
16:07 Su-Shee .oO(does he actually has a nice summer house and just left the city..? ;)
16:13 vti just running from the polizуi!!!
16:13 vti polizei!!!
16:14 GitHub15 joined #mojo
16:14 GitHub15 mojo: master Sebastian Riedel * bc2caab (1 files in 1 dirs): cleanup - http://bit.ly/dxLKmE
16:14 GitHub15 left #mojo
16:16 sri :O
16:17 yko vti: stolen watermelon?
16:17 vti yko: :D
16:17 yko *stole
16:19 yko damn... ->to(template => 'foo') doesn't work as expected if waypoint was before
16:20 vti bug bug bug
16:20 yko so it's to(template => 'foo', action => undef)
16:20 yko it's not a bug actually
16:20 yko you just should know what you are doing
16:20 yko it's not as easy as steal melon :p
16:22 yko hmm. bridge is nice
16:25 sri actually i'm not sure action should be inherited...hmmm
16:27 yko sri: any way you do, there will be someone who dissatisfied
16:27 vti it depends... if youwant to confuse your users.. it is really good :)
16:28 arthas joined #mojo
16:28 yko i think, inheriting action can be used. just not in my case.
16:28 sri controller yes, but action is rather unlikely
16:28 yko hmm. maybe youre right
16:29 * yko spent 20min diging what's wrong...
16:29 sri oh well, it's too hot to investigate... so i'll just wait and see if someone cares enough to send a patch ;p
16:29 Su-Shee is there an example out there somewhere which includes grabbing something from a db? I vaguely remember someone doing something with mojo and couch.
16:30 sri couchdb is as easy as calling $self->client->...
16:30 sri database would be used just like in any other script
16:30 sri we have no helpers for that
16:31 vti i have ;p
16:31 Su-Shee I know how to use a DB, I'd like to see more of the life cycle or best practices or "do it this way" or something like that.
16:31 Su-Shee vti: is it in your github?
16:31 vti no =/
16:31 yko then you haven't :p
16:32 vti ok, ok
16:32 vti wait 3 secs
16:32 sri wasn't that twitter clone a great example for sqlite?
16:32 vti ye
16:32 * yko giggles
16:32 Su-Shee ah, where?
16:32 sri http://onionstand.blogspot.com/2010​/06/tweetylicious-twitter-like.html
16:32 Su-Shee I really just need some architectural inspiration.
16:32 garfield [ The Onion Stand: Tweetylicious - a Twitter-like microblogging app in just one file! ]
16:32 garfield http://xrl.us/bhon63
16:33 sri http://github.com/garu/tweetylicious
16:33 garfield [ garu's tweetylicious at master - GitHub ]
16:33 vti http://github.com/vti/mojolicious-apps
16:33 garfield [ vti's mojolicious-apps at master - GitHub ]
16:34 yko http://github.com/korshak/Mojolicious-Plugin-Db  also :D
16:34 garfield [ korshak's Mojolicious-Plugin-Db at master - GitHub ]
16:34 yko mb not the best example, but works
16:34 vti it is async *cough*
16:35 vti i have even attachments...
16:35 Su-Shee "korshak"? seriously? man, I'm so going to name my modules something painfully girly...
16:36 Su-Shee I'm already clicking. :)
16:36 vti Su-Shee: it's his last name ;p
16:36 Su-Shee yeah I just realized.. I thought you're naming your module funny. ;)
16:37 vti i have to finish that website.. i just don't have motivation :D
16:37 yko Su-Shee : didn't get you :)
16:38 Su-Shee yko: was a heat related knot in my brain, you can gladly ignore it. ;)
16:38 yko korshak is really my last name, actually it means 'kite bird'
16:39 vti Su-Shee: there is also a good testing/development/production infrastructure
16:40 Su-Shee vti: where?
16:41 vti Su-Shee: mojolicious-apps with commands like teardown and setup for couchdb and tests
16:41 Su-Shee ah. :)
16:41 Su-Shee I need icecream.
16:41 vti i have a better project.. but it i commercial ;p
16:42 Su-Shee based on mojo? how large?
16:44 vti not very large
16:45 vti only 21 class
16:46 Su-Shee well that's more than a one-file blog :)
16:46 vti bootylicious is great!
16:46 vti and other 100 apps :D
16:48 dvinciguerra joined #mojo
16:48 Su-Shee well I'M going to replace 12 years of Perl .. it's going to be more than one file ;)
16:54 Su-Shee yko: your db handler looks like what I need right now
16:55 yko tbhat's great
16:56 yko i often use it with mysql and sqlite, of any bugs, issues, etc - please write issue on github, ill be happy
16:56 yko *if
16:57 Su-Shee I'll use it with postgres then.
16:58 yko :D great
16:59 siegfri3d joined #mojo
17:01 Su-Shee uh, garu has very nice examples in tweetylicious. I must say the Perl I've seen recently looks _very_ nice, clean, tidy and zenlike.
17:33 awnstudio joined #mojo
17:54 baton8 joined #mojo
18:00 ask joined #mojo
18:12 GitHub127 joined #mojo
18:12 GitHub127 mojo: master Sebastian Riedel * 08cc992 (6 files in 5 dirs): cleanup - http://bit.ly/aOGRiL
18:12 GitHub127 left #mojo
18:12 sri please test this on different platforms :)
18:13 yko waypoint(...)->bridge->...->waypoint('/') was bad idea
18:14 yko waypoint(...)->bridge->...->waypoint # that's how it works :(
18:17 yko sri: that was really 'cleanup' :D
18:17 sri i hope
18:18 sri not sure yet if there are some platforms going berserk without locking
18:19 yko anger users will tell you :D
18:25 sri :)
18:30 chansen It will work on unix platforms, but you may want to google for "thundering herd problem"
18:30 vti wtf was taht?
18:31 vti why did you remove lock?
18:34 yko first one :)
18:48 sri vti: i wanted to try and see where it breaks ;p
18:53 sri as far as i can see nginx has its accept mutex disabled, thats what got me interested in trying
18:53 sri (by default)
18:58 ashleydev joined #mojo
19:07 arthas joined #mojo
19:11 spleenjack joined #mojo
19:13 GitHub140 joined #mojo
19:13 GitHub140 mojo: master Sebastian Riedel * b68d4fd (1 files in 1 dirs): cleanup - http://bit.ly/cnrSkO
19:13 GitHub140 left #mojo
19:17 GitHub49 joined #mojo
19:17 GitHub49 mojo: master Sebastian Riedel * 0c674c6 (1 files in 1 dirs): fixed typo - http://bit.ly/bZO8jX
19:17 GitHub49 left #mojo
19:33 GitHub58 joined #mojo
19:33 GitHub58 mojo: master Sebastian Riedel * d8dc1c5 (1 files in 1 dirs): fixed small prefork bug - http://bit.ly/ax30SX
19:33 GitHub58 left #mojo
19:44 sri hmm, almost looks like rails is dead, everybody arguing if node or clojure win the web war Oo
19:44 sri *wins
20:03 ashleydev So i've been using engineyard for my rails app...
20:03 ashleydev pretty sweet
20:03 ashleydev engine yard that is
20:03 ashleydev i think i'm going to get trailer park for hosting perl (& mojo) apps
20:04 ashleydev trailerpark.com
20:05 ashleydev (not that I have that url :( )
20:05 ashleydev but it's funny
20:17 su-bzero_ joined #mojo
20:24 sri that name is not inspiring confidence ;p
20:34 sri Zzz
21:02 yko zzZ
22:16 ltriant joined #mojo
22:22 dotan joined #mojo
23:03 spleenjack joined #mojo
23:27 baton8 joined #mojo

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