Camelia, the Perl 6 bug

IRC log for #mojo, 2012-01-05

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

All times shown according to UTC.

Time Nick Message
00:00 sri but
00:00 sri i'm up for an experiment
00:00 marcus Mind if I create an experimental branch sometime?
00:00 marcus Going to bed right now tho.
00:00 sri feel free to fork the whole thing, rename it and release as separe distributions
00:01 sri if it works out we can consider it for master
00:02 marcus Might be we could bundle it in master.
00:02 * marcus will ponder a bit.
00:02 sri why not separate?
00:03 marcus Sri: avoid degraded installation performance for mojolicious.
00:03 sri then there's no risk and it won't interfere with normal development
00:03 marcus Don't mean for the experimentation.
00:04 marcus Neeways, typing on iPhone is tedious. Nite.
00:04 sri well, that's an option should the experiment work out
00:04 marcus Yep
00:04 marcus Naming suggestion?
00:04 sri hmm
00:05 Psyche^ joined #mojo
00:07 marcus Yoyo? :)
00:08 sri YoYoYo
00:08 sri Maya
00:08 sri i hear that's in fashion this year
00:10 sri Maya-2012.tar.gz
00:21 sri marcus: you should also blog/post to list/open issue for it, i think there will be some heated discussions
00:22 sri maybe invite those that complained
00:29 herrclark joined #mojo
00:43 sri oh shit
00:43 sri i just tweeted it and there's already a lot of fear
00:43 * sri deletes tweet quickly
00:51 sri marcus: that shitstorm is all yours, i'm gonna stay the hell away from it :)
01:01 alnewkirk joined #mojo
01:25 herrclark joined #mojo
01:37 j3nnn1 joined #mojo
01:49 gshank joined #mojo
01:55 hshong joined #mojo
02:12 alnewkirk joined #mojo
02:20 alnewkirk joined #mojo
02:33 xaka joined #mojo
02:43 bvierra joined #mojo
03:15 tempire The separation argument doesn't make any sense to me
03:15 tempire You have to download the "WHOLE FRAMEWORK"
03:15 tempire so what
03:16 tempire you mean the no dependencies framework that downloads an installs in less than a minute?
03:16 tempire it's a perception from an old world, where the cpan dependency hell was everywhere.
03:16 tempire that issue has been removed from mojolicious.
03:17 tempire the entire argument is based on a perception that has no foundation
03:17 tempire That being said, I'm open to someone countering those points.  Maybe I'm missing something.
03:18 crab i think it's pointless to separate them
03:20 crab and just causes added pain for the maintainers with no real benefit to anyone else
03:24 D4RK-PH0ENiX joined #mojo
03:28 d4rkie_ joined #mojo
03:29 d4rkie joined #mojo
03:37 Psyche^ joined #mojo
03:58 mire joined #mojo
06:57 Vandal joined #mojo
07:24 batman joined #mojo
07:28 sugar joined #mojo
08:23 hshong` joined #mojo
08:26 berov joined #mojo
09:07 Vandal joined #mojo
09:27 cosmincx joined #mojo
09:35 Vandal joined #mojo
09:37 marcus memowe: can bootylite renders statically?
09:41 stephan48 currently you would need to have it running as far as i know
09:41 batman joined #mojo
09:47 * sri agrees with tempire and crab
09:54 marcus sri: I'll start the discussion with a detailed blog post and open a ticket for discussion linking to it in the blog post sometime this weekend. Will try to lay out all the various arguments.
09:54 marcus if nothing else, it might kill some myths about mojolicious in the wider perl community.
10:12 D4RK-PH0ENiX joined #mojo
10:34 sri marcus: well, i'll be staying away from it
10:53 d4rkie joined #mojo
11:02 marcus sri: good times.
11:02 purl DON'T OPEN IT!  IT'LL CRASH AOL!
11:19 ki0_ joined #mojo
11:32 sri marcus: btw. since LWP split up their test fails have gone up significantly
11:33 marcus LWP split up?? They were my favorite band :-(
11:34 marcus seems to be failing on windows - http://www.cpantesters.org/distro​/L/libwww-perl.html#libwww-perl-6.03
11:34 marcus not sure if that's related.
11:34 sri comparing 5.x and 6.x releases
11:35 sri especially the time between releases is significant
11:36 sri how has splitting up made LWP any better?
11:36 marcus I didn't even know they split up, so no idea :)
11:36 marcus not been very interested in LWP since I started using Mojo.
11:37 sri well, that's the kind of question i'll be asking in the discussion ;p
11:37 marcus ah, HTTP-Message
11:37 marcus sri: thought you said you'd stay away from it :D
11:37 marcus glad you reconsidered
11:38 marcus My mind isn't set on the issue at all, I just want a thorough discussion.
11:39 * marcus searches for discussions for why LWP split up.
11:39 sri well, i'll be asking as a normal user :)
11:39 sri HOW DOES THIS MAKE MY LIFE EASIER?!
11:40 marcus http://code.activestate.co​m/lists/perl-libwww/7338/
11:40 marcus MARK OVERMEER SAYS IT WILL!
11:42 sri as a developer i also don't see how 50 respositories are easier to maintain than one
11:42 sri maybe if you've given up on architecture development and just concentrate on maintenance
11:43 marcus Yeah, not understanding that either.
11:44 sri i mean, have ever looked into LWP? it's a real mess
11:44 sri it has evolved over 10 years and nobody ever cleaned up the code
11:45 foomator mess, ack
11:45 sri as an exercise, try locating the code that handles timeouts
11:46 crab marcus: i find running away screaming to be a decent first response to marc overmeer.
11:46 marcus crab: ever met him? :)
11:47 crab no.
11:47 marcus and did you run away screaming?
11:47 marcus oh :/
11:47 crab i haven't actually met any perl hackers, only damian once (in 2001)
11:50 sri also worth investigating, how are web frameworks in other languages maintained?
11:50 marcus most ruby frameworks depends on rack?
11:51 netfeed rails has a gem for each part of the fremwork afaik, one for model and for controller ans so on
11:51 sri the framework developers don't maintain rack
11:51 sri we don't maintain Plack, yet support it too
11:51 marcus ~% gem install tzinfo builder memcache-client rack rack-test rack-mount erubis mail text-format thor bundler i18n
11:52 marcus that seems to be the dependencies for rails 3
11:52 sri and look up how that has worked out for rails3
11:52 sri how many people are whining to get rails2 ease of use back
11:52 sri i've seen prominent ones like thomas fuchs do it
11:54 sri i'd go as far as to say that rails3 has helped sinatra *a lot*
11:55 sri and that's even though ruby gems don't even run tests on installation
11:56 sri our installation time would prolly go from < 1 to like 5 minutes
11:57 sri tests would get a lot more fragile too, since useragent couldn't use lite apps anymore
12:00 sri marcus: none of those dependencies are maintained or developed as part of raisl either as far as i can see
12:01 sri rails itself is one big repo https://github.com/rails/rails
12:03 sri the equivalent for us would be one build script that splits up our current repo into 50 small dists without unit tests
12:04 netfeed and each one of those action* dirs is a gem
12:04 marcus sri: yeah, that's a very good point.
12:05 marcus I think I'm just going to give up on this idea, and write a blog post about why it's stupid to split mojolicious =)
12:07 sri https://code.djangoproject.com​/browser/django/trunk/#django # one repo as well
12:08 marcus sri: none of those frameworks include a http stack tho.
12:08 marcus or a http client
12:08 sri https://github.com/joyent/node # has even v8 bundled
12:08 sri and libev and stuff
12:09 sri https://github.com/joyent/node/tree/master/deps
12:11 marcus http://twitter.com/#!/marcusram​berg/status/154897721453592576
12:12 sri actually, i would be interested in seeing a project where splitting up has really worked out
12:12 sri something that's still actively developed on an architecture level
12:13 marcus catalyst runtime and catalyst devel :D
12:13 sri not just deps, but core devs maintaining all components
12:13 sri ;p
12:13 sri you're supposed to argue against me *cough*
12:13 marcus I've had my share of problems with Class::MOP and Moose as well. :-/
12:13 marcus since they always need to be in sync anyways.
12:14 sri didn't they actually plan to merge mop into moose?
12:14 sri i vaguely remember a blog post that talked about the future of moose and modules getting merged
12:15 * marcus is all bat and no belt
12:15 marcus http://twitter.com/#!/miyaga​wa/status/154898470099099648
12:17 sri dunno what he means
12:18 marcus seems to be a lot about feeeeelings
12:19 sri http://search.cpan.org/~miyagawa/Plack-0.9985/ # Plack::Util::Accessor ;p
12:19 marcus most of the reason I was interested in that is that people tell me they would like to use Mojo but not Mojolicious, but since discussing it I've never heard any convincing arguments about *practical* issues. it's all about a feeling of hubris from installing to much.
12:19 marcus oh noes :o
12:20 marcus this is why we need https://github.com/stevan/p5-mop in core :)
12:20 sri i'm all for that
12:21 sri also, i'm pretty sure Mojo but no Mojolicious is a strawman argument
12:21 sri in the end, for me it comes down to the question "what makes our users happy?"
12:22 sri i'm sure there's a great jobs quote for that
12:27 sri http://damienkatz.net/2012/0​1/the_future_of_couchdb.html # couchdb is toast
12:30 marcus old news
12:30 purl Dewey defeats Truman! ; Hindenburg explodes over New Jersey! ; Lincoln shot sitting in John Wilke's Booth! ; Climate Change Campaigners blame upcoming "Little Ice Age" on not enough petrol combustion  or debian OpenSSL PRNG hole
12:31 sri yea, but now it's official
12:33 sri http://www.cpantesters.org/distro​/M/Mojolicious.html#Mojolicious-2.42 # glad we are still supergreen
12:35 sri now… would you mind if i get back to working on the mongodb driver? :)
12:36 marcus sri: nothing would make me happier :)
12:37 marcus with a possible exception of blackjack and hookers.
12:38 marcus http://www.youtube.com/watch?v=z5tZMDBXTRQ
12:41 marcus We really need to get this quote into the core http://www.youtube.com/watch?v=​swDpWNKB5Co&amp;feature=related
12:43 sri please open an issue :D
12:44 netfeed i thought couchdb moved from a c-version to erlang because of threading, is he giving c another go now then?
12:44 sri think it's already mixed in couchbase
12:55 marcus sri: http://twitter.com/#!/miyaga​wa/status/154908570352222210
13:11 sri marcus: that dependency could be made soft thought
13:11 marcus yeah
13:12 memowe marcus: good idea! #bootylite #static
13:13 sri https://metacpan.org/favorite/leaderboard # ++ is back, don't forget to vote ;)
13:13 sri think you can still manually manipulate it though, there is just a captcha
13:16 sri marcus: if you're still writing about it, i think the important point is that we are not contra dependencies, but pro maintainability and user happyness
13:18 sri i think splitting off mosules makes most sense when you're giving away responsibility, like we did with Mojo::Server::FastCGI
13:18 sri s/s/d/
13:22 alnewkirk joined #mojo
13:23 marcus yea
13:24 tarski joined #mojo
13:24 memowe marcus, stephan48: it will, maybe in a few days. Please remember me in a week or so if nothing happens. :) #bootylite #static
13:24 marcus memowe: good times.
13:24 purl DON'T OPEN IT!  IT'LL CRASH AOL!
13:25 memowe Y U ACCIDENTALLY AOL!?
13:30 stephan48 memowe: great idea!
13:31 stephan48 yea crash aol!
13:34 memowe aol still exists?
13:50 tudorconstantin joined #mojo
13:51 tudorconstantin hi, does someone used other types of sessions with Mojolicious than the one builtin with the signed cookie?
13:54 tudorconstantin I am trying with something like:
13:54 tudorconstantin my $session = MojoX::Session->new(         #tx        => Mojo::Transaction->new(),         #tx        => $self->build_tx,         #store     => MojoX::Session::Store::DBI->new(dbh  => $dbh),         store     => [dbi => {dbh => $dbh}],  # use MojoX::Session::Store::Dbi         transport => 'cookie',                # this is by default         ip_match  =>
13:54 tudorconstantin hmmm.....bad formatting...sorry
13:54 sri please do not paste into the channel
13:56 tudorconstantin sorry for that - I am writing the question on stackoverflow and put here the link ... just a second
13:58 sri i'm not aware of anyone using MojoX::Session anymore though
14:04 jnap joined #mojo
14:06 tudorconstantin oh...f..
14:06 tudorconstantin here is the link to the question http://stackoverflow.com/questions/​8743822/mojolicious-custom-sessions
14:08 sri and if there is one you'd be much more likely to find him on the mailing list than stackoverflow
14:08 tudorconstantin are there any plugins that can replace the default sessions?
14:09 tudorconstantin I wrote the question there to be able to format the code better
14:12 stephan48 why you would want that?
14:13 tudorconstantin what? to replace the default session?
14:14 tudorconstantin i don't like the idea of letting the user to be able to see internals of the application
14:15 stephan48 why internals?
14:15 stephan48 you could just store a session id into the cookie
14:15 stephan48 and load data for it by yourself
14:16 stephan48 you then also get the advantage that the user cant fiddle with session id
14:16 stephan48 yoyu could make it into a plugin
14:17 tudorconstantin yep, I want exactly that: just store a session id into the cookie and load data from somewhere else ... but done transparently for me by sessions
14:18 stephan48 you can make a pluginn for it
14:19 stephan48 f.e. use a FileCache as session store
14:19 stephan48 and use session id as key
14:19 tudorconstantin I thought that MojoX::Session was doing exactly that
14:20 stephan48 its old
14:20 tudorconstantin initialising it with different stores, you would get different storage locations
14:20 tudorconstantin oh, ok
14:21 tudorconstantin is there a fast way of getting the session id?
14:21 tudorconstantin I tried with $self->session->transport->get
14:21 tudorconstantin but I got an error from MojoX::Session
14:23 sri rule of thumb, all MojoX::* modules are unmaintained legacy
14:23 sri the good ones usually start with Mojolicious::Plugin::*
14:25 trone That you know, there are Mojolicious examples online that show how to provide files outside public path? (I would to serve them after auth) I was thinking about render_data with file->slurp and format key or set headers before … but looking for easy mode. :)
14:27 stephan48 i got something laying around here based on apache2 xsendfile and mojolicious
14:27 stephan48 need to look throu where
14:30 stephan48 it i would avoid slurping the file
14:30 stephan48 depending on the size this can be bad
14:31 trone stephan48: ah. I was not too worried about this (for ignorance, not for prediction. :-( )
14:31 stephan48 oh ok
14:32 sri slurping is never necessary in mojolicious, Mojo::Asset::File can stream files
14:32 stephan48 great
14:32 sri just look at how Mojolicious::Static does it
14:32 trone sri: indeed I'm using Mojo::Asset::File with path to create the object, and $file->slurp to serve it inside render_data …
14:33 sri which is wrong
14:33 trone ops. ~_~
14:34 sri https://github.com/kraih/mojo/blob/m​aster/lib/Mojolicious/Static.pm#L119
14:36 trone sri: ok, thanks. I go to give a look around there.
14:37 sri basically, if you're loading a whole static file into memory you're doing something wrong
14:38 stephan48 https://tn123.org/mod_xsendfile/ that i meant with xsendfile
14:40 trone sri: Certainly I understand is not an optimal approach, but I still miss a lot of things about IT and don't know exactly as move to make the right thing ™ . Thanks for the advice. (I saw *_chunks methods, and maybe now understand well why they perform.)
14:40 rhaen I love you. All!
14:41 trone stephan48: thank you, I will read also it. :) Never listen about xsendfile
14:42 stephan48 basically you just do auth checks in your app, give a download header and tell the webserver serve this file
14:42 sri rafl: new entries are still not up
14:42 stephan48 and it will support ranges and all optimizations apache2 has
14:49 herrclark joined #mojo
14:50 ZadYree joined #mojo
14:53 sri rhaen: grouphug?
14:53 purl sri: that gave some error
14:53 sri purl: :(
14:53 purl Buck up, kiddo! It's not so bad...
15:03 GitHub75 joined #mojo
15:03 GitHub75 [mojo] kraih pushed 1 new commit to master: http://git.io/MviPmg
15:03 GitHub75 [mojo/master] added streaming response example - Sebastian Riedel
15:03 GitHub75 left #mojo
15:03 stephan48 where do i find a info about url_for and query params
15:03 sri trone: that should explain it
15:05 trone sri++ # too much. :)
15:09 kaare joined #mojo
15:11 GitHub28 joined #mojo
15:11 GitHub28 [mojo] kraih pushed 1 new commit to master: http://git.io/IWLniQ
15:11 GitHub28 [mojo/master] extended url_for example with query params - Sebastian Riedel
15:11 GitHub28 left #mojo
15:11 sri stephan48: and there you go
15:11 stephan48 thanks!
15:20 mire joined #mojo
15:21 noganex_ joined #mojo
15:22 cosmincx joined #mojo
15:31 stephan48 i am creating a url via link_to with query params, url_for('auth', path => $path)->query(sig=> $sig)) but link_to takes the stuff coming out of url_for and puts it behind the current url resulting in something like auth/register/auth/..., url_for gives me auth/...
15:33 tarski joined #mojo
15:43 GitHub23 joined #mojo
15:43 GitHub23 [mojo] kraih pushed 1 new commit to master: http://git.io/KnJ5fg
15:43 GitHub23 [mojo/master] fixed link_to example - Sebastian Riedel
15:43 GitHub23 left #mojo
15:45 stephan48 thanks
15:45 sri considered adding an exception to url_for for Mojo::URL objects… but that could break existing apps
15:47 sri if anyone is interested in that please open an issue, then we can discuss it further
15:48 sri https://github.com/kraih/mojo/blob/mas​ter/lib/Mojolicious/Controller.pm#L552
15:48 stephan48 are there any known problems regarding CGI and basic auth?
15:48 sri basically extend that line with || blessed($target) eq 'Mojo::URL'; or similar
15:49 stephan48 especially Mojolicious::Plugin::BasicAuth
15:49 stephan48 when i run my app as daemon auth works fine when i run it as cgi i permanently get an auth dialog
15:56 sri https://gist.github.com/1565766 # canvas animation of my metacpan logo :)
15:59 stephan48 $self->req->url->to_abs->userinfo  <- in here the username and pw should be or?
16:00 stephan48 it seems like they dont get passed along
16:06 d4rkie joined #mojo
16:16 batman joined #mojo
16:31 Therstrium joined #mojo
16:33 tempire fyi, I monitor stackoverflow for mojolicious questions
16:33 sri tempire++
16:34 tempire marcus: I like your blog post idea, regardless of the tactic you take, for dispelling myths
16:34 tempire GROUPHUG
16:34 purl tempire: that gave some error
16:34 tempire also, aol is a remarkably successful company these days
16:34 tempire focusing on content, mostly the background
16:41 tempire oooh, I love the new modern perl feature: use Modern::Perl 2012;
16:41 tempire I never used feature because of having to type a . in the per version
16:41 tempire *perl
16:45 memowe Maja perl would die on 'use Modern::Perl 2012'.
16:46 memowe s/Maja/Maya/
16:46 * tempire senses a new mem
16:46 tempire e
16:46 * memowe senses a new memowe
16:50 ki0 joined #mojo
16:58 xaka joined #mojo
16:59 ladnaV joined #mojo
17:00 noganex joined #mojo
17:01 pl0p joined #mojo
17:01 mattp_ joined #mojo
17:03 tempire joined #mojo
17:08 mattastrophe joined #mojo
17:09 marcus Did you shave or something?
17:09 memowe marcus: no, but I should. ;)
17:10 tempire hmm.  use Modern::Perl 2012 doesn't work for me.
17:10 tempire :(
17:23 tempire for those interested: http://empireenterprises.com/metacpan_logo.html
17:23 buu joined #mojo
17:23 sri :)
17:36 xaka press start 10 times, looks cool )
17:39 diogo joined #mojo
17:42 mattastrophe joined #mojo
17:43 cosmincx joined #mojo
17:44 diogo1 joined #mojo
17:47 tempire true enough.  added another button
17:48 sri :D
17:48 sri ohoh
17:48 sri i clicked that button 10 times
17:49 abra joined #mojo
17:49 tempire and your browser crashed?
17:49 tempire my fan went crazy
17:49 sri nope
17:49 sri yea, fan started
17:50 sri but, it's really cold outside, warm and cozy laptop is actually pretty nice
17:57 xaka hm, browser got stuck :(
17:58 tempire HORDE!
18:09 nightmare Why does controller request url not match the controller request env variable "REQUEST_URI"?  Trying to run Mojolicious app in CGI environment.
18:15 Therstrium joined #mojo
18:33 alnewkirk joined #mojo
18:34 * tempire likes the term "rogue pixels"
18:34 alnewkirk joined #mojo
18:35 bvierra joined #mojo
18:49 memowe Finally!
18:50 memowe I found the hidden manual!
18:50 memowe http://www.amazon.com/Mojo-How​-Keep-Back-Lose/dp/1401323278
18:50 memowe Mojo: How to Get It, How to Keep It, How to Get It Back if You Lose It [Hardcover]
18:51 memowe I think it's about cpanminus...
18:56 kaare joined #mojo
18:58 Greeny joined #mojo
19:01 diogo joined #mojo
19:06 Psyche^ joined #mojo
19:10 kaare joined #mojo
19:36 ccushing_ joined #mojo
19:52 kaare_ joined #mojo
20:02 grim_fandango joined #mojo
20:31 perlite joined #mojo
20:36 stephan48 any idea to my problem stated earlier with cgi and basic auth, where credentials doesnt seem to propergate to the uri?
20:43 DaTa stephan48: well in a cgi environment you can only access the username, the password isn't propgated usually. I guess you should find the username in $self->req->env->{REMOTE_USER} or so
20:44 stephan48 ye
20:45 stephan48 https://gist.github.com/1567202 in the .htaccess or apache2 config propagated the Authorization header back to mojolicious
20:46 stephan48 would be good to clarify that
21:11 sri http://blog.cloudant.com/the-futue-of-couchdb/ # couch vs couch :o
21:16 kaare_ joined #mojo
21:21 vel joined #mojo
21:41 preflex_ joined #mojo
21:58 diogo joined #mojo
22:17 mattastrophe joined #mojo
22:21 larry joined #mojo
22:50 larry joined #mojo
22:55 diogo joined #mojo

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