Camelia, the Perl 6 bug

IRC log for #mojo, 2011-03-12

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

All times shown according to UTC.

Time Nick Message
00:02 mattastrophe joined #mojo
00:10 ok22 left #mojo
00:57 metaperl left #mojo
01:00 metaperl joined #mojo
01:19 minimalist left #mojo
01:22 minimalist joined #mojo
01:42 * tempire is sad to see client go
01:43 * tempire is a team player, though
01:43 tempire GO TEAM!
01:43 * tempire embraces ->ua
02:03 marcus $team->usa
02:07 tempire http://www.youtube.com/watch?v=LJOBw9fCkeE
02:44 mateu norwegians are golden in the biathlon
02:47 mateu edenc: Eu comi a sua mãe até o rabo e ela realmente gostou.
02:49 mateu bunda dela é apertado, se sente bem.
02:59 edenc mateu: sua mãe eu não quero, folgada demais
02:59 edenc mateu: mas sua irmã tem uma xaninha apertadinha e gostosa
03:00 edenc mateu: quando eu encostei a cabeça ela ficou toda molhadinha!
04:48 mateu edenc: good to see you still got it :P
05:40 Hammit joined #mojo
05:40 Foxcool joined #mojo
05:55 Foxcool left #mojo
06:07 Foxcool joined #mojo
06:34 tholen_ joined #mojo
06:36 sri mateu: why do you need invalid json?
06:36 sri anyway, we don't do that
06:37 sri marcus: nope
07:07 tholen_ left #mojo
07:19 tholen_ joined #mojo
08:16 marcus sri: no what?
08:16 marcus oh
08:17 sri NO YOU CAN'T HAVE NICE THINGS
08:17 marcus http://nowjs.com/
08:18 sri already seen it, interesting
08:22 marcus could it work on mojo?
08:22 sri why shouldn't it?
08:22 marcus the examples are all node.js?
08:23 sri who cares? node.js has to use the same transport layer we have to
08:24 marcus yeah, I guess it wouldn't be as seamless with a mojo backend tho
08:24 sri of course not, it would user perl on the server side
08:24 sri 2 languages = less seamless than one
08:26 sri the only "advantage" you get from node.js is the ability to reuse some of your code on both sides
08:30 marcus yeah, it makes for a nice illusion
08:31 marcus Wonder how it's going with our project to get ipad2s
08:31 marcus We sent arne's brother to NYC. ;)
08:32 marcus so hopefully we will have some on monday.
08:41 sri ohoh, http://www.twitvid.com/LICNU this does not look good at all
08:47 marcus shit
09:03 metaperl1 joined #mojo
09:06 metaperl left #mojo
09:20 jamesw uses socket.io though
09:24 metaperl joined #mojo
09:27 Foxcool left #mojo
09:27 metaperl1 left #mojo
09:30 minimalist left #mojo
09:40 Foxcool joined #mojo
10:07 jpablobr joined #mojo
10:37 sigue left #mojo
10:54 GitHub95 joined #mojo
10:54 GitHub95 mojo: master Sebastian Riedel * 22dc223 (2 files in 2 dirs): more ioloop tweaks - http://bit.ly/fhRrQB
10:54 GitHub95 left #mojo
11:36 * sri would always go with plain old long polling instead of socket.io
11:36 sri because scaling is much more predictable
11:41 MisterHa1t anyone know if the haml renderer is stable/complete enough for production yuse?
11:41 MisterHa1t -y
11:42 sri vtiware, don't even think he maintains the haml module anymore
11:42 MisterHa1t doubting it as well
11:43 sri haml is out anyway ;p
11:44 sri there's this new hipster whitespace sensitive language
11:44 MisterHa1t any other nifty renderers I should be looking at? trying to find something that my designers dont suicide at, even TT they didnt like
11:44 Hammit i didn't even know there was a haml module
11:44 sri well, you could make a plugin for mustache or so
11:45 MisterHa1t never heard of it
11:45 * MisterHa1t googles
11:45 sri logicless templates with mustaches
11:46 sri http://mustache.github.com/mustache.5.html
11:46 MisterHa1t did you mean sass by :hipster whitespace sensitive language"?
11:47 sri nope, let me search
11:50 jpablobr yeah, I heard the mustache is what twitter is using for the newtwitter, apparently is one of the fastest template engines...
11:50 MisterHa1t mustache looks pretty cool, you just pass a single hash to the renderer with all your tags?
11:50 sri http://jade-lang.com # this was the new js hipsters template system
11:51 sri jpablobr: that's not true
11:51 MisterHa1t owait, make a plugin, so nobody's working on one or written one already, fun
11:52 sri yea, be the first!
11:53 sri most mustache implementations are actually quite slow
11:53 MisterHa1t you have much more confidence in me than I do
11:53 MisterHa1t ah so another chance for mojolicious to be fastest
11:53 sri just look at the latest tt plugin, it's quite easy
11:54 MisterHa1t so with mustache, you still write all your html, it just reads tags from a hash?
11:54 sri mojolicious templates are most likely the fastest pure perl ones
11:54 sri you'll have to go xs for more
11:54 Hammit if your passing a hashref to ur template, sounds like HTML::Template
11:54 sri MisterHa1t: not tags, data
11:54 sri you hardcode tags
11:55 MisterHa1t right
11:55 MisterHa1t is now known as MisterHatt
11:55 sri that's the point of logicless templates ;p
11:55 MisterHatt ah
11:55 jpablobr sri: well that what I thought from this video conf
11:55 jpablobr http://confreaks.net/videos/476-rubyconf2010​-mustache-in-the-front-mustache-in-the-back
11:55 Hammit left #mojo
11:56 sri ah, frontend mustache
11:57 sri that's one of the advantages, you can use mustache.js and render some json serialized hash client side using the same template you would normally user server side
11:58 sri i think those coffeescript template systems have a brighter future client side though
11:59 sri https://github.com/sstephenson/eco
12:00 sri really like that one
12:10 Foxcool left #mojo
12:12 MisterHatt hmm, i18n is handy for UI translation but what about actual content?
12:12 MisterHatt any suggestions?
12:12 purl "I suggest you don't smoke cigarettes ... and don't touch yourself when you're chillin' at the bus stop."
12:13 MisterHatt that's some quality advice right there, smart bot~
12:15 sri actual content is a problem for you to solve i'm afraid
12:16 sri it's usually model layer specific
12:17 MisterHatt I thought as much, I was just curious how other people handled it
12:17 MisterHatt site I'm working on right now needs en/jp/he interfaces and no language id in the url
12:18 * MisterHatt isn't a webdev so shouldn't even be doing this, but mojo is fun
12:19 sri it's one of the harder problems to solve
12:20 MisterHatt so it seems
12:20 MisterHatt especially as the pages (for some users) have inline editing of content, but should't really have someone updating content in english and not japanese or hebrew...
12:21 MisterHatt of course dont have that inline editing written either, joy
12:21 Foxcool joined #mojo
12:22 MisterHatt anyone tried phenona yet by the way?
12:22 sri beta has not started yet
12:23 MisterHatt ah
12:23 MisterHatt any idea what they're waiting for?
12:24 sri i guess they are just not ready yet
12:26 MisterHatt sure is jumping the gun and allowing signups, sort of implies they had an active closed beta at some stage
12:27 sri i think they are trying to solve the hard problem
12:27 MisterHatt which is?
12:27 sri a free tier
12:27 MisterHatt ah
12:28 sri ^
12:28 MisterHatt one of my work projects is somewhat similar, free tier wasn't really a problem in our stuff though
12:28 sri it's the secret that made heroku so successful
12:29 MisterHatt so it seems
12:29 MisterHatt their addon stuff is quite good too though, not quite sure why phenona wants to implement their own mysql service, or any other service that is already managed by someone else to a higher degree in the same datacentre and thereby not costing them anything in bandwidth
12:30 MisterHatt our biggest problem at least in code isolation, looked into a bunch of things but they all seemed too open or full of overhead making them slow...
12:30 MisterHatt s/in/is/
12:32 sri yea, managing databases is annoying, i would outsource that part too
12:32 sri with isolation you mean sandboxing perl?
12:32 MisterHatt yes
12:32 MisterHatt perl itself doesnt really have to be sandboxed I guess
12:32 MisterHatt just the app and any non-system modules, as well as allocating ram/cpu
12:33 sri i think a good ptrace module would be enough
12:33 MisterHatt are there any?
12:33 sri then you just whitelist the syscalls you want to allow
12:33 sri there is one for freebsd only
12:33 MisterHatt we looked into lxc, cgroups, uevalrun, and mentioned ptrace but nobody wanted to mess with it on grounds of "too complicated"
12:34 sri it doesn't look so complicated, just a bit of xs
12:34 MisterHatt xs?
12:34 purl somebody said xs was hard for everyone. (except Nick Ing-Simmons) or a new kind of fun
12:34 sri :)
12:34 MisterHatt never even heard of it
12:35 sri the glue layer between perl and C
12:36 sri it's not very pretty and you always need a compiler :S
12:36 * sri still hopes ctypes will make it into perl 5.16
12:36 MisterHatt ah, yeah, that doesnt sound nice
12:37 sri http://cpansearch.perl.org/src/DANKOG​AI/FreeBSD-i386-Ptrace-0.03/Ptrace.xs
12:38 sri i guess making it work on linux and os x would take a weekend
12:39 MisterHatt our server is running linux but not like we'd be implementing on anything but
12:47 sigue joined #mojo
12:57 crab hi
12:57 purl que tal, crab.
13:00 jpablobr left #mojo
13:05 jpablobr joined #mojo
13:20 weestrom joined #mojo
13:20 weestrom hey sri: got a problem that seems to have no good solution and wanted ur take if youll lend it
13:21 sri shoot
13:22 weestrom i have a .net 3.5 dependency that only seems to like running on win32, i prefer ubuntu/fedora as dev environments, so i got my eclipse and mojolicious fired up in a vmware player image on a win32 box
13:23 weestrom my dependency api executes on bar.com/foo.application?c:/path/to/baz.xml
13:23 weestrom and returns a .net executable, i cannot figure a good way to execute that remotely because i cant launch a non tty browser to do the put
13:24 weestrom tried cywin/sshd, and Net::XML::Perl, but no joy
13:25 weestrom and wine/.net 3.5 and mono are no joy as well for keeping it in the VM, so I think im stuck on win32, any options i havent explored that you are familiar with?
13:28 weestrom it actually returns an executable xml that msxml parser pulls the .net exe, so... maybe xml::simple can help get me to the next step (totally forgot to closely inspect the executable xml on the vm side.....)
13:28 weestrom sometimes explaining a problem is more valuable than help :)
13:30 weestrom are there good recipes for the new useragent stuff? for the immediate moment im going with system(firefox.exe uri) but thats just for quick hackery
13:33 Foxcool left #mojo
13:33 * sri doesn't really understand the problem
13:34 sri the cookbook is full of good recipes
13:37 weestrom i got a good idea while explaining it to you to try
13:38 weestrom thus i stopped trying to explain...
13:38 weestrom i realized that i wasnt parsing the xml that is likely a pointer to the correct .net executable
13:39 weestrom so i can retrieve that in the vm and see if i can either execute the .net code there or remotely
13:44 Foxcool joined #mojo
13:44 weestrom left #mojo
13:56 crab weestrom: your problem is full of scary words i don't understand
14:13 Foxcool left #mojo
14:22 Foxcool joined #mojo
14:51 mateu sri what's invalid about { "content": "<b>Mojo</b>" } ?
15:00 sri mateu: actually it's valid, just less strict
15:00 sri both are valid
15:05 mateu and I can't get json=> to not alter valid JSON?
15:06 sri ?
15:06 mateu have json => not escape the slash as it's not necessary to produce valid JSON.
15:06 sri no you can't
15:06 sri why would you care?
15:07 marcus but it should deserialize to the same?
15:07 sri indeed
15:07 mateu I'm doing some testing that involved  JSON::encode_json and json from Mojo and they act different
15:07 sri it's exactly the same
15:08 sri different result same meaning
15:08 mateu I guess I should not test the string but the resulting JSON.
15:08 sri indeed
15:08 mateu sri: thanks for taking a moment to explain  the situation.
15:10 marcus mateu: json generators could use various amount of whitespace as well, for instance, but still produce the same structure
15:13 sri you could also get a different encoding
15:17 sri json.t has some funny examples, like the faihu roundtrip ones, with  surrogate pairs in utf-16
15:27 yegor joined #mojo
15:28 yegor left #mojo
16:10 oleber joined #mojo
16:16 Foxcool left #mojo
16:24 perlrocks Twitter: "Sao Paulo Perl Mongers: Scraping with Mojolicious easy (and Atom feeds!): Http://disq.us/1e9nb7" (pt) --mantovanidaniel http://twitter.com/mantovanida​niel/status/46607394582691840
16:25 sri garu++
16:55 Foxcool joined #mojo
17:42 garu :)
17:43 garu it should read "Scraping made easy with Mojolicious" but I guess it's understandable
17:43 garu I used Mojo::Client since UserAgent is not on CPAN yet
17:44 perlrocks Twitter: "Scraping with easy Mojolicious http://bit.ly/gj9Yqv perl # [# Equinox - London Perl Mongers] (v @ garu_rj)" (pt) --_sppm http://twitter.com/_sppm/status/46627582149541889
17:59 minimalist joined #mojo
18:22 oleber left #mojo
18:37 snearch joined #mojo
18:48 perlrocks Twitter: "Mojolicious Perl Web Framework is the first choice ... (personal opinion)" (zh-TW) --gslin http://twitter.com/gslin/status/46643863921754112
18:50 kaare_ joined #mojo
18:53 kaare_ left #mojo
18:54 perlrocks Twitter: "Said lemacf, @ gslin push perl web framework http://tinyurl.com/47m46fz (Mojolicious Web Framework), although less likely to write perl, ... http://plurk.com/p/b4mgmr" (zh-TW) --fcamel http://twitter.com/fcamel/status/46645334314721281
18:57 kaare_ joined #mojo
19:00 kaare_ left #mojo
19:13 kaare_ joined #mojo
19:16 kaare_ left #mojo
19:17 snearch left #mojo
19:23 perlrocks Twitter: "just minted a shiny new ubuntu 10.10 server vm for messing around with Mojolicious" --jasoncrowther http://twitter.com/jasoncrowt​her/status/46652453332062208
19:23 kaare__ joined #mojo
19:26 perlrocks Twitter: "In Mojolicious the irc to see their tweet twitter bot translated by throwing up, feeling quite strange XD" (zh-TW) --gslin http://twitter.com/gslin/status/46653199448408064
19:26 sri haha
19:26 kaare__ left #mojo
19:29 gslin haha
19:32 perlrocks Twitter: "http://www.pubbs.net/201103/debian/7899-bug617​946-itp-libtest-www-mechanize-mojo-perl-module​-for-testing-web-applications-built-using-mo" --bullpubbs http://twitter.com/bullpubb​s/status/46654748627185664
19:37 kaare__ joined #mojo
19:37 kaare__ is now known as kaare
19:38 kaare left #mojo
19:47 Foxcool left #mojo
20:02 Foxcool joined #mojo
20:23 stephan48 i have a template which i could rather be standalone(and then it should include the layout) or as partial(and then not include layout), is there some other way than seperate the layout including part in another template which then includes the partial template?
21:05 garu stephan48: you mean like template blocks?
21:05 stephan48 mh?
21:05 purl mh is a unix commandline-based mail user agent or at http://www.ics.uci.edu/~mh/ or Men's Health magazine or a "women's magazine" for men or Getty's way of saying hm or your way of saying hm
21:07 stephan48 how would that help me?(ok i could put the layout and title commands in a block and only exec the block on a condition, but thats sounds awfull too)
21:08 tholen_ left #mojo
21:09 garu stephan48: I'm not sure that will help you since I don't think I understood your question. I was talking about this, though => http://paste.scsys.co.uk/92589
21:09 stephan48 ye
21:09 stephan48 look: i have a template called ship_info, which displays some infos about a ship based on a stash value
21:10 stephan48 this template could be either displayed directly.. to only view the ship info, where it should also include the layout
21:11 stephan48 but i also have a event_info template which should show the ship_info + some infos about a event(based on stash)
21:11 stephan48 now when i include ship_info in event_info i get douplicated layout(once from event_info and once from ship_info)
21:17 Foxcool left #mojo
21:22 garu stephan48: is event_info always shown together with ship_info?
21:22 stephan48 ye
21:22 stephan48 but ship_info can be standalone
21:23 garu right
21:24 kaare joined #mojo
21:26 garu I guess the simplest solution would be set a stash value and process event_info under ship_info whenever it's true. Not really elegant, but does the trick
21:27 kaare left #mojo
21:28 kaare joined #mojo
21:30 kaare left #mojo
23:31 Alias__ joined #mojo
23:35 Alias left #mojo

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