Camelia, the Perl 6 bug

IRC log for #mojo, 2010-09-01

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

All times shown according to UTC.

Time Nick Message
02:26 GabrielVieira2 joined #mojo
02:32 GabrielVieira left #mojo
02:35 janus left #mojo
02:48 janus joined #mojo
02:57 baton8 left #mojo
03:11 yakudzo joined #mojo
03:19 Foxcool joined #mojo
04:03 tl joined #mojo
04:05 tholen42 left #mojo
04:48 sri joined #mojo
04:49 perlrocks joined #mojo
04:58 yakudza joined #mojo
05:17 sri yay, most of my servers are back
05:18 sri no more slacking, i want patches!
05:18 * sri cracks the whip
05:21 marcus frisky dingo?
05:32 tempire http://video.adultswim.com/fri​sky-dingo/meet-awesome-x.html
05:52 sri oooh, apple event today :o
05:53 marcus and they are streaming it live
05:53 marcus LIVE
05:53 sri OMG
05:53 marcus tempire: that url is not approved for european audiences
05:54 marcus http://www.adultswim.co.uk/videos/frisky-dingo/xpo found something on the .uk site tho
05:55 und3f[A] is now known as und3f
06:18 koban joined #mojo
06:19 marcus !tweet TO INFINITY AND BEYOND!
06:19 perlrocks marcus: http://twitter.com/perlrocks/status/22684550191
06:23 marcus I think I'm switching to node.js, too many cool projects at http://nodeknockout.com/teams
06:23 marcus is there a node.js blog engine yet?
06:24 marcus http://thechangelog.com/post/52664​6366/wheat-blog-engine-for-node-js
06:31 ltriant left #mojo
06:33 ltriant joined #mojo
06:35 nuclon joined #mojo
06:35 nuclon morning
06:35 nuclon I couldn't find Mojo::Server::FCGI at CPAN, any ideas why?
06:41 marcus weirdness
06:41 nuclon Actually I couldn't find Mojo::Server::FCGI at all
06:42 nuclon http://github.com/acme/mojo-fcgi but it looks very old
06:44 marcus nuclon: I have a distro by sri in my Source directory
06:44 marcus but it seems he deleted it or something
06:44 nuclon interesting
06:46 nuclon so sri suggests to use pure-perl Mojo::Server::FastCGI
06:46 marcus I think he suggests using the prefork daemon
06:46 marcus or hypnotoad, when he finishes that
06:50 janus hm
06:51 janus can i walk recursive structures with blocks somehow?
06:51 janus oh wait, our instead of my could do the trick
06:52 nuclon marcus: what is "hypnotoad"?
06:57 marcus nuclon: http://irclog.perlgeek.de/search.pl?c​hannel=mojo&nick=&q=hypnotoad
06:59 und3f is now known as und3f[A]
07:09 ysyrota joined #mojo
07:11 sri actually i made someone else maintainer of Mojo::Server::FCGI, he let it die
07:11 sri personally i have no interest in maintaining it
07:12 marcus who was it?
07:12 nuclon marcus: http://groups.google.com/group/mojolicio​us/browse_thread/thread/f97141bee791bbaf
07:12 sri ispy
07:12 janus q: how can i call templates recursively?
07:12 nuclon sri: but Mojo::Server::FastCGI doesn't have prefork mode
07:12 marcus sri: seems xantus volunteered.
07:13 sri xantus never took it over
07:13 sri i don't even know where to find the latest version
07:13 sri those under my name prolly won't work at all
07:14 tl left #mojo
07:15 marcus I've got Server::FCGI in production at $client
07:15 marcus so I have some interest in letting it live.
07:16 sri the prefork part might still depend on Mojo::Server::Daemon::Prefork
07:16 sri was a very ugly hack
07:16 sri that would need to be forked from the core
07:17 perlrocks Twitter: "@sukria Competition is a good thing. :) @PerlDancer vs #mojolicious" --aer0 http://twitter.com/aer0/status/22687269626
07:17 nuclon sri, so is there any recommended way to deploy Mojolicious with the best possible performance?
07:18 sri it always depends on what your app does
07:18 sri i will recommend hypnotoad once it's released though
07:18 nuclon say, nothing special. DB/templates :)
07:18 sri (the next incarnation of Mojo::Server::Daemon::Prefork)
07:19 sri DBI?
07:19 sri doesn't matter what you use then
07:21 sri cute how the dancer folks still think we are competing
07:21 sri the only competitiong out there is node.js
07:22 sri and it is no longer framework vs framework but language vs language
07:24 sri http://www.yuiblog.com/blog/2010/08/30/​yui-theater-douglas-crockford-crockford​-on-javascript-scene-6-loopage-52-min/
07:24 sri excellent talk
07:25 sri whats amazing about node knockout is not so much the entries but the sheer number of sponsors
07:25 marcus downloading the .mov for later perusal
07:26 marcus did anyone build a decent mvc framework on top of node.js yet?
07:26 sri i think its express vs geddy atm
07:28 nuclon Mojo::Server::Daemon:237 [21116]: Event "read" failed for connection "IO::Socket::UNIX=GLOB(0x9141ba4)": Bad arg length for So
07:28 nuclon cket::unpack_sockaddr_un, length is 16, should be 106 at /usr/local/lib/perl5/5.8.8/mach/Socket.pm line 380
07:28 sri lol
07:28 sri freebsd and perl 5.10?
07:28 marcus nuclon: on freebsd with domain sockets?
07:28 nuclon sri: freebsd and perl5.8
07:28 sri it's a perl bug
07:28 marcus :p
07:29 marcus seems it affects all versions of perl on freebsd
07:29 ltriant left #mojo
07:29 tl joined #mojo
07:30 sri marcus: i should have listened to you and used node.js for mojolicious -.-
07:33 marcus sri: yepp
07:33 marcus then you would be world famous today ;)
07:33 sri :,(
07:38 Alias_ node.js?
07:38 sri really wondering what will happen with perl/python/ruby now
07:38 Alias_ hrm?
07:38 sri node.js is making us obsolete
07:39 Alias_ Never heard of it
07:39 sri javascript for the server side done right
07:39 sri http://www.yuiblog.com/blog/2010/08/30/​yui-theater-douglas-crockford-crockford​-on-javascript-scene-6-loopage-52-min/ # very good introduction
07:40 Alias_ Can it bind to native C libraries?
07:40 sri yes
07:40 * Alias_ tries to think of things he truly hates about JavaScript
07:40 sri it is basically libev + libeio + v8
07:41 sri c++ mostly
07:41 sri an event loop scriptable with javascript
07:42 Alias_ oh, it's eventy?
07:43 sri top to bottom
07:43 Alias_ And it uses non-blocking mysql/postgres/oracle?
07:43 sri yes
07:43 sri everything that would block runs in an external thread pool
07:43 Alias_ The one thing I've generally found weak about event stuff is that you generally don't reuse the same code in non-eventy places
07:43 Alias_ And that's awefully handy
07:44 Alias_ Also, non-eventy code is easier to write
07:44 Alias_ And god, debug
07:44 sri thats the big question
07:44 sri what if your language didn't give you the choice to block
07:45 Alias_ Then it's harder to write stuf
07:45 Alias_ Which is why functional and erlang'y languages aren't popular
07:45 Alias_ Despite being superficially superior
07:45 sri javascript people are used to events
07:45 Alias_ Haskell people are used to functional, that doesn't make it easier for everyone else to move to it
07:46 sri everybody knows javascript already from the browser
07:46 Alias_ If PHP and Python have shown us nothing else, it's that dumbing down programming and creating a low barrier to entry and learning creates popularity
07:46 sri you don't have to learn anything new
07:46 Alias_ Lots of people have done little bits of non-event JavaScript
07:46 Alias_ Not everyone has done "serious" JavaScript
07:47 sri using the same language on the client and server side makes a huge difference
07:47 Alias_ Agreed
07:48 marcus you basically have to write eventy javascript on the browser
07:48 marcus imo
07:48 Alias_ Nah
07:48 Alias_ Lots of people do simple procedural stuff for rendering whatever in the browser, with the odd minor random on_load triggered by the browser
07:49 marcus even $('foo').click(function(){}) is eventy
07:49 Alias_ That's not really understanding writing full blown event apps
07:49 marcus understanding it and doing it isn't the same.
07:50 sri what do perl/python/ruby offer that could outweight the benefits of javascript?
07:50 Alias_ tons of people can load google adwords and boot it, doesn't mean they could write an event-driven server app :)
07:50 Alias_ sri: That's what I'm pondering
07:50 sri i'm afraid this could very well kill the good old scripting languages for mainstream use
07:51 Alias_ Package repository
07:51 Alias_ OpenJSAN notwithstanding
07:51 sri you automatically get people hooked with the browser
07:51 Alias_ Lack of ability to bind to any random C library
07:51 sri they've started a new package repository i think
07:51 sri npm
07:51 Alias_ How many significant applications do you see that are mostly about the browser
07:51 marcus Alias_: I can't do that with perl either. It requires me to learn XS
07:52 Alias_ marcus: But at least you CAN use those libraries
07:52 Alias_ Creation of FFI modules will come
07:52 sri i predict libffi support in the near future
07:53 Alias_ How on earth do you do libffi in an event server
07:53 Alias_ It doesn't support threads does it?
07:53 Alias_ Being event?
07:53 Alias_ or forks
07:53 sri it has the libeio thread pool
07:53 sri it does evented io on all platforms with emulation
07:53 Alias_ I don't know what that is... does that let you use more than one core?
07:54 sri http://search.cpan.org/~ml​ehmann/IO-AIO-3.65/AIO.pm # thats libeio
07:54 sri yes it does
07:55 Alias_ That, btw, is why we ignored POE or AnyEvent etc for Padre and went with threads
07:56 sri thats the big war going on atm
07:56 sri events vs threads
07:56 marcus but I thought perl's threading support sucked.
07:57 sri it's also in the crockford video btw
07:57 sri he compares events and threads
07:57 sri very objectively
07:58 cognominal left #mojo
07:59 sri the biggest loser is of course php
07:59 sri it is pretty much toast
07:59 Alias_ It's been on borrowed time the same way Perl was with CGI
07:59 Alias_ Just waiting for something that solves the same problems as well, but sucks less, to arrive
08:00 Alias_ I'm not sure if JavaScript can achieve the same level of integration between code and interface though
08:00 Alias_ Which is really the one main advantage of PHP
08:00 sri crockford also predicts the end of the server side *template*
08:00 sri instead there will be dom sharing
08:01 sri syncing DOMs between server and client
08:01 sri YUI apparently moves into that direction
08:03 marcus search engines really love that ...
08:03 sri hehe
08:04 sri http://www.youtube.com/watch?v=-BsB0HpS768 # try indexing this guys amazing html5 demo
08:04 marcus It's pretty funny. I did lots of serverside javascript development back in 1999
08:04 marcus using netscape server
08:04 sri (make sure to not miss the video effects at the end)
08:05 * sri started programming in 2000
08:06 marcus started programming pascal in 1992 or so.
08:06 marcus I did some basic on c64 and amstrad before that.
08:07 sri i've just been playing games until then, was never really interested in programming
08:10 marcus I had programming classes in high school
08:13 janus new test for template.t: http://gist.github.com/560392 am i doing something wrong?
08:14 sri that seems quite wrong
08:15 janus please elaborate :)
08:16 sri printing the code might give you an answer
08:16 sri $mt->code i think was it
08:17 sri you really should use the new begin/end to build reusable blocks
08:18 janus ->code it is
08:18 janus oops
08:24 janus i've initially tried it with blocks, but those can't call themselves :/
08:24 sri they can't?
08:24 janus a closure knows it's own method?
08:25 sri let me try adding a test
08:26 janus my $sub = { $sub->() } doesn't work
08:26 sri my $sub; $sub = { $sub->() }
08:27 janus hmkay, let's try again
08:31 GitHub139 joined #mojo
08:31 GitHub139 mojo: master Sebastian Riedel * b3aac62 (2 files in 2 dirs): added recursive template block test - http://bit.ly/9riwBT
08:31 GitHub139 left #mojo
08:31 sri works flawless :)
08:34 janus duh, that's what i get for using git and reading release docs :/
08:34 janus didn't notice the capture_start change
08:34 sri pretty isn't it :)
08:35 janus sri: thanks anyway, it works now :)
08:40 sri Alias_: http://npmjs.org/ # thats their cpan btw
08:50 sri http://peepcode.com/products/nodejs-i # damn, those peepcode guys are fast
08:50 Foxcool left #mojo
09:19 Akron joined #mojo
09:21 koban left #mojo
09:23 nuclon left #mojo
09:31 perlrocks Twitter: "http://bit.ly/dsASq0 Mojolicious Web Framework - Join the revolution!" --pliemskavx6 http://twitter.com/pliemskavx6/status/22693110811
09:43 koban joined #mojo
10:10 GabrielVieira2 left #mojo
10:52 ka2u left #mojo
10:52 ka2u joined #mojo
11:30 janus should i use includes instead of trying to get layouts nested?
11:33 janus with my current fail rate i must be learning like a maniac :>
11:35 sri nested layouts are usually not such a good idea, even though they should work
11:44 yakudzo left #mojo
12:04 janus inline templates are amazing - lovely how one can draft the hierarchy in one file :)
12:07 janus sweet...
12:07 sri :)
12:10 janus !tweet development with liteappning speed... just #mojolicious
12:10 perlrocks janus: http://twitter.com/perlrocks/status/22701465256
12:10 perlrocks Twitter: "development with liteappning speed... just #mojolicious --janus" --perlrocks http://twitter.com/perlrocks/status/22701465256
12:41 dynax60 left #mojo
12:44 Akron left #mojo
12:54 cognominal joined #mojo
12:55 Akron joined #mojo
13:09 jest joined #mojo
13:14 GabrielVieira joined #mojo
13:25 sri back to hypnotoad hacking :)
13:28 awnstudio left #mojo
14:09 marcus sri: I hope you have more than hot deployment with hypnotoad
14:09 marcus I already have that on catalyst with FCGI-Engine
14:40 yakudza left #mojo
14:43 tempire I agree with sri about node.js being a threat.  It's about eliminating barriers of entry.  Server-side javascript may have a learning curve, but I already 'think' I can do it because I'm already exposed to it & have used it.
14:44 tempire And now I'm downloading it to prove it to myself.
14:44 tempire And there you go.  I've been trojaned by node.js.
14:45 tempire Something about that doesn't sound right...
14:48 marcus yeah, you should use goatsed
14:48 marcus :p
15:01 GabrielVieira left #mojo
15:13 koban left #mojo
15:23 tempire sri: I appreciate your eye for design.  It's part of what makes mojolicious appealing.
15:40 koban joined #mojo
15:46 koban left #mojo
15:47 GabrielVieira joined #mojo
16:26 Alias_ left #mojo
16:29 ysyrota left #mojo
16:34 jest left #mojo
17:09 spleenjack joined #mojo
17:17 esskar__ joined #mojo
17:21 esskar_ left #mojo
17:21 esskar__ is now known as esskar_
17:38 Akron left #mojo
17:39 esskar__ joined #mojo
17:41 stephen left #mojo
17:42 stephen joined #mojo
17:42 stephen left #mojo
17:43 stephen joined #mojo
17:43 esskar_ left #mojo
17:43 esskar__ is now known as esskar_
17:52 und3f[A] is now known as und3f
17:53 marcus one more thing?
18:12 Alias joined #mojo
19:08 dvinciguerra joined #mojo
19:17 yakudza joined #mojo
19:33 Htbaa joined #mojo
20:08 su-bzero_ joined #mojo
20:10 und3f left #mojo
20:10 und3f joined #mojo
20:12 su-bzero left #mojo
20:25 spleenjack left #mojo
20:30 perlite left #mojo
20:30 perlite joined #mojo
20:55 Htbaa left #mojo
20:57 dvinciguerra left #mojo
21:28 jest joined #mojo
21:30 tempire The appletv is looking pretty sweet
21:32 jest left #mojo
21:43 und3f is now known as und3f[A]
22:06 ltriant joined #mojo
23:10 sri :)
23:11 sri marcus: what do you hope hypnotoad will provide besides hot deployment?
23:24 stephen left #mojo
23:25 stephen joined #mojo
23:29 diegok left #mojo
23:30 GabrielVieira left #mojo
23:50 GabrielVieira joined #mojo

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