Camelia, the Perl 6 bug

IRC log for #mojo, 2010-11-01

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

All times shown according to UTC.

Time Nick Message
00:00 spleenjack1 left #mojo
00:18 forwardever left #mojo
00:18 und3f is now known as und3f[A]
00:28 GitHub107 joined #mojo
00:28 GitHub107 mojo: master Sebastian Riedel * 0c05461 (1 files in 1 dirs): cleanup - http://bit.ly/clTtqm
00:28 GitHub107 left #mojo
00:35 perlrocks Twitter: "Released #mojolicious 0.999934 with a bugfix that makes the client noticeably faster for a few big websites, enjoy. #perl" --kraih http://twitter.com/kraih/status/29326140715
02:20 meraxes left #mojo
02:20 und3f[A] left #mojo
02:20 rhaen left #mojo
02:20 ryanc left #mojo
02:20 fish_ left #mojo
02:20 memowe left #mojo
02:20 hidek___ left #mojo
02:20 charsbr left #mojo
02:20 stephan48 left #mojo
02:20 diegok left #mojo
02:20 kthakore left #mojo
02:20 crab left #mojo
02:20 Kulag left #mojo
02:20 jamesw left #mojo
02:20 Alias__ left #mojo
02:20 polvo left #mojo
02:20 viking-3 left #mojo
02:20 marcus left #mojo
02:20 ashleydev_logbot left #mojo
02:20 chansen left #mojo
02:22 tardisx left #mojo
02:24 ashleydev_logbot joined #mojo
02:25 marcus joined #mojo
02:25 chansen joined #mojo
02:25 jamesw joined #mojo
02:28 viking-3 joined #mojo
02:28 tardisx joined #mojo
02:29 MojoGuest728 joined #mojo
02:29 MojoGuest728 From: http://www.google.fr/search?rlz=1C1S​VED_enFR394FR394&sourceid=chrome​&ie=UTF-8&q=demo+websockets (1 hits)
02:29 MojoGuest728 kkkkkkkkknn
02:29 MojoGuest728 left #mojo
02:32 polvo joined #mojo
02:33 Drossel joined #mojo
02:33 Alias_ joined #mojo
02:33 memowe joined #mojo
02:33 meraxes joined #mojo
02:33 und3f[A] joined #mojo
02:33 diegok joined #mojo
02:33 rhaen joined #mojo
02:33 ryanc joined #mojo
02:33 kthakore joined #mojo
02:33 fish_ joined #mojo
02:33 crab joined #mojo
02:33 stephan48 joined #mojo
02:33 hidek___ joined #mojo
02:33 charsbr joined #mojo
02:51 xantus did anyone see this: http://aws2.xantus.org:3000/
02:55 xantus I'm running apache bench ever 5s
02:55 xantus every
02:56 sri xantus: unreachable
02:57 xantus I restarted it
02:57 xantus try again :)
02:58 tempire You know what we need.  An 2010 mojolicious advent calendar.   Now would be about the time to start thinking about topics.
02:58 sri xantus: lol
02:58 xantus sri: like it?
02:58 sri indeed
02:59 xantus I subclassed the loop to track connections
03:02 sri tempire: not sure we have enough writers yet
03:02 xantus I think I'll make an animation toggle, maybe add some click handling to show info about each client
03:03 xantus and the requests it has made, state, etc
03:03 sri i need to work on more guides rather than a calender
03:05 sri tempire: i'm afraid it might be like the first catalyst calender i organized, which was hell
03:05 tempire what made it hell?
03:06 sri making people deliver what they promised and step in half of the time write a new entry myself a few hours before deadline ;p
03:07 tempire hell indeed
03:07 tempire well you're right about enough writers.
03:07 tempire an advent week, maybe.
03:07 tempire next year, then.
03:07 sri or take the guides apart
03:08 sri present a new recipe every day
03:08 tempire that's a decent idea
03:08 tempire would be a good template for an intro-to-mojolicious series
03:09 sri till then i'll also have two more guides ready i think
03:09 sri most people won't have read those, lots of new stuff to pick apart
03:18 tempire excellent.  you'll have something to do while resisting the urge to fix bugs so quickly ;)
03:29 sri ;p
03:29 sri i'll also have to wrap up 1.0 :)
03:37 xantus everyone should go to http://mojolicio.us/ and donate if you're feeling generous :)
03:40 sri I'm Bender, Baby, Please insert liquor!
03:43 sri http://www.youtube.com/watch?v=QYQ4uV8NDJo
03:56 ash_ joined #mojo
03:59 perlrocks Twitter: "Is feeling mojolicious tonight" --thugboss626 http://twitter.com/thugboss626/status/29341047475
04:03 tholen42 joined #mojo
04:05 tl left #mojo
04:45 Alias joined #mojo
05:10 perlrocks left #mojo
05:25 perlrocks joined #mojo
06:07 giwi joined #mojo
06:23 koban joined #mojo
06:24 ysyrota joined #mojo
06:57 ash_ left #mojo
07:02 polvo Can't locate object method "prepare_pid_file" via package "Mojo::Server::Daemon::Prefork" <- eh?
07:03 sri polvo: did you update recently? the prefork daemon is not part of mojolicious anymore
07:04 sri see Changes
07:04 polvo sri, yes, i has updated recently. i was just going to try the prefork daemon. :)
07:04 sri leftover files then
07:05 polvo sri, what (where) are these files?
07:05 omega xantus: haha, love the bouncing cubes! :)
07:05 sri wherever you installed mojolicious previously
07:06 polvo sri, i mean `app help` lists 'daemon_prefork' as an option
07:06 sri because the .pm file is still there
07:06 sri from the old installation
07:07 polvo hm, cpan left it i think
07:07 omega sri: how about including that file in one last dist, with a helpfull error?
07:07 sri omega: that would prevent someone from taking it over
07:07 omega not, if it was only for one dist?
07:07 sri i still hope someone will adopt the prefork daemon ;)
07:08 sri omega: ?
07:08 omega I don't see the difference between the situantion now, and including that command for one more release, with a helpful error?
07:09 sri i'm looking for someone to adopt the Mojo::Command::DaemonPrefork and Mojo::Server::Daemon::Prefork
07:09 sri if i include the command it will be blocked
07:10 omega I'm not talking about forever, I'm talking about for one release
07:10 omega It could even display a message that it is looking for maintainers
07:10 sri releases happen every second day atm ;p
07:10 omega yeah, but as polvo just proved, the files stay even after upgrade
07:11 sri if we go that route we would have to block it for at least a few months
07:11 omega but its not a big deal, was just a suggestion to make the experience for upgrading users like polvo easier
07:11 polvo i've just removed and reinstalled Mojolicious -> solved
07:12 polvo btw, do fastcgi instances share the same data? e.g. db connection
07:12 sri no
07:13 * polvo got stuck in a situation when ORM is the bottleneck
07:13 sri databases are *always* the bottleneck
07:14 polvo not the database :)
07:14 polvo sri, http://pastebin.com/JSDAA1yj
07:14 sri no big surprise there either
07:15 sri you trade developer time for cpu time
07:15 polvo solutions? i'm about to use DBI for the critical code
07:16 sri don't think there is a "solution"
07:16 sri it's a fair tradeoff
07:16 sri is dbic always that ugly? Oo
07:16 polvo ok, workarounds? :)
07:16 sri use DBI ;p
07:17 polvo that's what i'm going to do
07:17 sri or some fancy nosql database
07:17 sri redis is as fast as it gets
07:17 polvo it's not a db issue like i said
07:17 sri but pay a different price
07:18 sri in the end all half decent code is a tradeoff
07:18 sri i tend to favor developer time for my code
07:20 polvo dbic is handy when it comes to creating and updating
07:26 marcus there are a few more lightweight orms out there.
07:27 MojoGuest692 joined #mojo
07:27 MojoGuest692 From: http://www.google.de/search?sourceid=chr​ome&amp;ie=UTF-8&amp;q=html+5+irc+client (3 hits)
07:28 marcus yay
07:36 und3f[A] is now known as und3f
07:39 MojoGuest692 left #mojo
07:40 sri one could built a neat little async orm DBD::Pg
07:40 sri *build
07:42 sri wonder if the tornado folks are right with their take on blocking database apis
07:42 perlrocks Twitter: "allframeworks / Perl Mojolicious:: Lite стартуем! http://ff.im/sYi3Z" (sv) --planetperlru http://twitter.com/planetperlru/status/29351749335
07:42 polvo noob question: how to get rid of $a in $a = shift; $b = $$a;
07:42 sri they basically favor blocking database apis because they get automatic throttling
07:43 perlrocks Twitter: "Mojolicious:: Lite 'll start! / Mojolicious / allframeworks.ru - Blog about modern web development and .. http://bit.ly/bWSzH8" (ru) --sharifulin http://twitter.com/sharifulin/status/29351790176
07:44 sri considering that databases are the bottleneck, overloading them can of course turn out to be a massive problem
07:45 sri especially if you have a cluster of web servers
07:46 perlrocks Twitter: "Mojolicious:: Lite 'll start! / Mojolicious / allframeworks.ru - Blog about modern web development and tools http://vk.cc/6agng" (ru) --xomaa http://twitter.com/xomaa/status/29351906024
07:46 sri oh, this week it's the russians :)
07:47 perlrocks Twitter: "Mojolicious::Lite СТАРТУЕМ! http://is.gd/gyG5n #perl #mojolicious #mojo" --xomaa http://twitter.com/xomaa/status/29351959456
07:48 baton8_ left #mojo
07:48 baton8 joined #mojo
07:50 polvo sri, imho, "hello world"-like programs are annoying
07:51 sri polvo: you can't be serious
07:51 polvo i've followed the blog on twitter
07:51 sri hello world is the one most important learning tool
07:52 polvo not when there are hundreds of the same articles -)
07:52 sri of course once you learned the basics you don't care
07:52 polvo which almost copy the manual
07:52 sri make better blog posts :D
07:53 polvo i don't write to blogs and read them seldom
08:00 polvo besides, i'm not a programmer :0))
08:00 sri :o
08:01 sri i'm not even human!
08:03 polvo hail the robots!
08:04 su-bzero_ joined #mojo
08:06 sri good meat bag
08:09 su-bzero left #mojo
08:10 marcus I for one welcome our new bit munching overlords.
08:10 sri *munch*
08:32 perlrocks Twitter: "#Mojolicious::Lite стартуем! http://goo.gl/fb/W2P7v #perl #mojo" --allframeworks http://twitter.com/allfram​eworks/status/29353885011
08:38 polvo heh, using pure DBI made big difference
08:50 su-bzero_ left #mojo
08:50 ysyrota left #mojo
08:50 ryanc left #mojo
08:50 Drossel left #mojo
08:50 fish_ left #mojo
08:50 rhaen left #mojo
08:50 memowe left #mojo
08:50 und3f left #mojo
08:50 hidek___ left #mojo
08:50 meraxes left #mojo
08:50 stephan48 left #mojo
08:50 kthakore left #mojo
08:50 crab left #mojo
08:50 diegok left #mojo
08:50 charsbr left #mojo
08:50 Alias_ left #mojo
08:50 polvo left #mojo
08:52 su-bzero_ joined #mojo
08:52 ysyrota joined #mojo
08:52 charsbr joined #mojo
08:52 hidek___ joined #mojo
08:52 stephan48 joined #mojo
08:52 crab joined #mojo
08:52 fish_ joined #mojo
08:52 kthakore joined #mojo
08:52 ryanc joined #mojo
08:52 rhaen joined #mojo
08:52 diegok joined #mojo
08:52 und3f joined #mojo
08:52 meraxes joined #mojo
08:52 memowe joined #mojo
08:52 Alias_ joined #mojo
08:52 Drossel joined #mojo
08:52 polvo joined #mojo
09:01 spleenjack joined #mojo
09:23 und3f is now known as und3f[A]
09:33 Alias_ left #mojo
09:33 Alias_ joined #mojo
09:39 marcus /win 8
09:44 memememomo_ left #mojo
10:00 stephan48 where is vti?
10:21 sri he's taking a break
10:23 sri hope he'll be back soon
10:25 polvo i'm trying to set flash variable in a bridge, will it be accessible later?
10:26 sri polvo: why shouldn't it?
10:26 polvo sri, because it isn;t for some reason :)
10:26 sri you might be misunderstanding it then
10:27 sri people tend to use flash instead of stash sometimes
10:27 sri mixing up scopes
10:28 polvo sri, http://pastebin.com/x7UD92y0
10:29 sri yea, thats totally wrong
10:29 sri read the session section of the tutorial
10:30 spleenjack left #mojo
10:31 fhelmber_ joined #mojo
10:31 polvo sri, do you mean perldoc Mojolicious::Lite  ?
10:32 sri yes
10:32 sri let me guess, you ignored it because you are starting with normal mojolicious right away? :)
10:32 sri you are supposed to learn lite first
10:32 polvo sri, yes :0)
10:33 sri haha... always the same ;p
10:33 sri people skip lite and ask those questions
10:34 crab well, where does it say you're supposed to pay attention to lite first?
10:34 crab i did, but i surely didn't know i was supposed to
10:36 polvo "Signed cookie based sessions just work out of the box as soon as you start using them. The flash can be used to store values that will only be available for the next request (unlike stash, which is only available for the current request), this is very useful in combination with redirect_to."
10:36 polvo is all it says :\
10:40 sri makes it pretty obvious
10:40 sri next request vs current request
10:41 spleenjack1 joined #mojo
10:41 sri crab: hmm
10:43 polvo sri, what is a next request?
10:43 sri the next http request
10:44 sri there is just one type of request
10:44 polvo sri, 'there is' <- do you mean my code?
10:45 bugweiser joined #mojo
10:47 polvo ah, never mind
10:47 bugweiser just tested Mojo::Client against Furl's benchmarks: http://pastebin.ca/1978580
10:47 bugweiser is synchronous mode really slower than lwp?
10:48 sri mojo client is entirely unoptimized perl code
10:49 sri why would it be faster than lwp if it has more features?
10:49 bugweiser i see
10:49 crab "oh no! my http client is too slow. i'm going to have to REWRITE ALL MY CODE!"
10:49 marcus wow, curl is 10 times faster than mojo.
10:50 sri c code being slower would be really scary :)
10:52 sri bugweiser: is that lwp or lwp simple?
10:52 bugweiser lwp::useragent
10:53 sri what actually did you test?
10:53 sri hello world?
10:54 bugweiser much simpler test: "Hi" :)
10:54 sri thats rather uninteresting
10:54 bugweiser http://github.com/tokuhirom/p5-Fu​rl/blob/master/benchmark/byown.pl
10:55 bugweiser it's useful enough to indicate what would be better suited for scraping
10:56 sri boring, too easy to build an optimized client for that
10:56 sri i can easily reach 12-14k req/s with an ioloop hack
10:57 spleenjack1 left #mojo
10:57 sri see examples/microhttpd.pl for a server example
10:58 bugweiser that looks like it's async though
10:58 sri all we do is async
10:59 sri sync is emulated
10:59 bugweiser oh, guess that explains it
10:59 spleenjack joined #mojo
10:59 sri btw. regarding your scraping example
11:00 sri thats a false assumption
11:00 spleenjack left #mojo
11:01 sri once latency comes into play everything changes
11:01 sri web crawlers spend most of their time waiting
11:02 polvo sri, i don't get it. i see 2 requests in logs. what am i doing wrong/
11:05 bugweiser left #mojo
11:13 sri polvo: just use stahs instead of flash
11:13 sri *stash
11:13 sri you'll know what flash is once you need it
11:17 polvo sri, can flash be used only in combination with redirect_to ?
11:17 sri why won't somebody benchmark how long it takes to implement some rather complex task with all http clients? :S
11:18 sri polvo: something like that
11:18 sri polvo: "next request"
11:19 polvo sri, what is a 'next request'? would reloading a page be it?
11:19 sri yes
11:19 sri liek i said, next http request
11:20 polvo sri, does it matter if it's get or post ?
11:20 sri no
11:26 crab wait, so if i set something in the flash in redirect_to and someone else makes a request before my browser requests the redirected-to page, what happens?
11:27 sri nothing
11:27 sri flash is stored in the session cookie
11:27 sri you would have to make parallel requests really really quick to screw it up
11:28 sri i'd say it is almost impossible
11:30 crab oh, heh.
11:31 sri the response would also have to be received before the other request starts, otherwise you just have parallel requests that all contain the same flash
11:32 sri i would go as far as saying it is impossible for the flash to get lost after a redirect
11:39 sri maybe i should write examples/microurl.pl :D
11:40 sri if you're just interested in fetching a file there is very little code needed
11:42 sri then again, maybe i should rather finish hypnotoad
11:44 crab i saw a pond full of toads
11:44 sri i could use some more toads
12:07 spleenjack joined #mojo
12:10 y0j left #mojo
12:32 polvo sri, still there?
12:34 polvo sri, my only problem was that i expected flash() to operate same as session(), i.e. $self->flash('a') ne $self->flash->{'a'} but self->session('b') eq $self->session->{'b'}
12:57 y0j joined #mojo
13:28 xantus omega: :)
13:40 dotan joined #mojo
13:54 GitHub105 joined #mojo
13:54 GitHub105 mojo: master Sebastian Riedel * f70cf7a (4 files in 4 dirs): improved Mojo::IOLoop performance by about 20% - http://bit.ly/aMOL5A
13:54 GitHub105 left #mojo
13:55 sri abstraction is so damn expensive
14:00 marcus mm
14:00 marcus assembly for life!
14:01 sri !
14:05 sri removing buffer objects might actually make the code simpler though
14:05 a|newkirk[assoc] left #mojo
14:07 forwardever joined #mojo
14:19 forwardever i have a question regarding "require" a module: the following command works fine: Mojo::Loader->new->load('World::Models::User'); if used directly in the controller, however, I get "Can't locate World::Models::User in @INC" as soon as the load command is used in a helper method, even though I started the app via script/myapp, it seems that the lib folder is not included in INC when the helper is
14:19 forwardever generated
14:21 sri unlikely, that kind of loading is used internally too
14:22 forwardever sri: maybe, it helps if I send you part of the plugin/helper: http://pastie.org/1264021
14:24 sri dunno
14:24 xantus sri: how do you prefer patches?
14:24 sri github pull request
14:24 xantus k
14:26 xantus there are 0 issues in github, do you plan on using its issue tracker at all?
14:26 sri we are
14:26 sri all closed ;)
14:27 marcus sri closes them as soon as they appear :p
14:27 sri indeed
14:28 xantus with fixes I hope :P
14:28 sri NEVER!
14:29 marcus that would be admitting weakness.
14:30 sri Debugging? Klingons do not debug. Our software does not coddle the weak. Bugs are good for building character in the user.
14:31 sri Our users will know fear and cower before our software! Ship it! Ship it and let them flee like the dogs they are!
14:33 forwardever sri: it seems that I forgot to remove "require $class" in my plugin, Mojo::Loader->new->load($class); works fine
14:33 sri shame on you!
14:34 forwardever it took me 15 minutes to find out :(
14:35 sri xantus: what kinda patch do you have?
14:38 baton8 left #mojo
14:43 baton8 joined #mojo
14:46 xantus sri: I noticed warnings from Content.pm when running the test suite, but I can't work on it @ work
14:47 sri xantus: strange, i don't see warnings
14:52 tempire left #mojo
14:53 xantus sri: yeah, strage.  I up-to-date, make clean, and all
14:53 xantus s/I/I'm/
14:53 xantus app.t is the first to warn.  Use of uninitialized value in length at /home/xantus/git-repos/moj​o/blib/lib/Mojo/Content.pm line 176.
14:55 xantus oh, btw, the drain cb to write was a great idea
15:01 ash_ joined #mojo
15:18 koban left #mojo
15:27 sri :)
15:48 perlrocks Twitter: "Uploadify and Mojolicious http://ff.im/sZmC7" --sharifulin http://twitter.com/sharifulin/status/29383827833
15:54 perlrocks Twitter: "Анатолий Шарифулин: Uploadify and Mojolicious http://ff.im/sZnor" --planetperlru http://twitter.com/planetperlru/status/29384244445
15:56 GitHub4 joined #mojo
15:56 GitHub4 mojo: master Sebastian Riedel * 8d207d3 (1 files in 1 dirs): cleanup - http://bit.ly/9Uz5jf
15:56 GitHub4 left #mojo
16:03 giwi left #mojo
16:17 xantus sri: line 177 of Content.pm should use a defined check instead of a length check
16:51 dotan left #mojo
17:07 rhaen meooow!
17:13 forwardever left #mojo
17:22 ysyrota left #mojo
17:26 stephen joined #mojo
17:48 alnewkirk joined #mojo
17:58 spleenjack left #mojo
18:04 Alias__ joined #mojo
18:09 Alias_ left #mojo
18:35 und3f[A] is now known as und3f
18:37 und3f hello
19:06 spleenjack joined #mojo
19:29 perlrocks left #mojo
19:36 forwardever joined #mojo
19:49 fhelmber_ left #mojo
19:54 ash_ left #mojo
19:57 y0j left #mojo
20:53 ash_ joined #mojo
21:20 MojoGuest427 joined #mojo
21:20 MojoGuest427 From: http://news.ycombinator.com/item?id=1277067 (32 hits)
21:21 MojoGuest427 left #mojo
21:26 alnewkirk left #mojo
22:08 xantus ruff
22:34 xantus the boxee box starts shipping nov 10th
22:35 xantus http://xant.us/boxee
22:48 memememomo_ joined #mojo
22:48 memememomo_ left #mojo
22:49 memememomo_ joined #mojo
23:03 forwardever left #mojo
23:28 spleenjack left #mojo
23:30 MojoGuest471 joined #mojo
23:30 MojoGuest471 From: http://news.ycombinator.com/item?id=1277067 (35 hits)
23:31 MojoGuest471 Hey there
23:31 MojoGuest471 What's up?
23:31 MojoGuest471 is now known as Koistya
23:31 Koistya It's working.. cool
23:31 Koistya but.. how fast
23:59 ash_ left #mojo

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