Camelia, the Perl 6 bug

IRC log for #mojo, 2010-12-08

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

All times shown according to UTC.

Time Nick Message
00:24 sri hmm, firefox 4 has websockets disabled by default now... good bye draft 76
00:24 sri calling it a security risk is smart though, makes it much easier to get rid of it
00:25 MojoGuest523 joined #mojo
00:25 MojoGuest523 From: http://www.google.com/url?sa=t&source=we​b&cd=4&ved=0CCoQFjAD&url=http%3A​%2F%2Fdev.xantus.org%2F&rct=j&q=webs​ocket%20demo&ei=VND-TNLbFIOClAf7xKDiCA&a​mp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (1 hits)
00:26 MojoGuest523 join
00:26 MojoGuest523 awesome
00:26 MojoGuest523 left #mojo
00:30 MojoGuest918 joined #mojo
00:30 MojoGuest918 From: http://www.google.com/url?sa=t&source=we​b&cd=4&ved=0CCoQFjAD&url=http%3A​%2F%2Fdev.xantus.org%2F&rct=j&q=webs​ocket%20demo&ei=VND-TNLbFIOClAf7xKDiCA&a​mp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (2 hits)
00:30 MojoGuest918 left #mojo
00:30 MojoGuest410 joined #mojo
00:30 MojoGuest410 From: http://www.google.com/url?sa=t&source=we​b&cd=4&ved=0CCoQFjAD&url=http%3A​%2F%2Fdev.xantus.org%2F&rct=j&q=webs​ocket%20demo&ei=VND-TNLbFIOClAf7xKDiCA&a​mp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (3 hits)
00:30 MojoGuest410 left #mojo
00:36 MojoGuest561 joined #mojo
00:38 MojoGuest561 left #mojo
00:41 MojoGuest186 joined #mojo
00:41 MojoGuest186 From: http://www.google.com/url?sa=t&source=we​b&cd=4&ved=0CCoQFjAD&url=http%3A​%2F%2Fdev.xantus.org%2F&rct=j&q=webs​ocket%20demo&ei=VND-TNLbFIOClAf7xKDiCA&a​mp;usg=AFQjCNGw6nZv0wO42uzUnHhWcD0vcKNULg (4 hits)
00:42 MojoGuest186 left #mojo
01:41 sri growing guide is finished, but i'm not entirely happy with it, gonna reread tomorrow and clean up
04:03 tardisx left #mojo
04:04 tholen42 joined #mojo
04:06 tholen left #mojo
04:40 achromic left #mojo
06:26 koban joined #mojo
06:28 ltriant left #mojo
06:56 ysyrota joined #mojo
07:16 kvorg left #mojo
07:34 Christian joined #mojo
07:34 Christian good morning
07:37 und3f[A] is now known as und3f
07:44 Alias_ joined #mojo
07:47 marcus sri: websockets is a security risk, but flash is ok?
07:47 marcus funny
08:03 und3f is now known as und3f[A]
08:08 tholen42 left #mojo
08:16 breaker313 joined #mojo
08:19 breaker313_ joined #mojo
08:20 breaker313_ left #mojo
08:21 tholen42 joined #mojo
08:24 espent joined #mojo
08:30 tholen42 left #mojo
09:02 su-bzero joined #mojo
09:02 su-bzero_ left #mojo
09:02 spleenjack joined #mojo
09:06 tholen42 joined #mojo
09:20 crab what is the growing guide?
09:20 crab or where is it?
09:24 yko in re-reading at the moment i think
09:29 crab do you know what it's about? growing what?
09:34 yko i think it would be aboput growing as a preson, obtain respect of other people and also i think it would answer on The Ultimate Question of Life, the Universe, and Everything
09:41 marcus it's about growing your app from lite to full app
09:41 marcus FULL ON APPVERTISING!
09:42 yko but then... how about Ultimate Question of Life? Ist sebastian that machine that was designed to find answer on it? :)
09:43 yko *autofix many typos
09:43 gizzlon "The Upgrade-based handshake is vulnerable to attack in network
09:43 gizzlon configurations involving transparent (or intercepting) proxies."
09:43 gizzlon http://www.ietf.org/mail-archive​/web/hybi/current/msg04744.html
10:04 crab ah.
10:04 crab i never really used lite at all
10:16 yakudzo how do I know current url in template ?\
10:23 yakudzo found (:\
10:28 alnewkirk joined #mojo
10:40 crab what was it? $self->tx->req->url or something?
10:41 yakudzo url_for ('current')
10:48 espent left #mojo
10:58 ajgb joined #mojo
11:33 otaviof left #mojo
11:38 yakudzo how can I get name of the route if I know path ? (:
11:40 snearch joined #mojo
11:59 Akron joined #mojo
12:09 tardisx joined #mojo
12:09 MojoGuest286 joined #mojo
12:09 MojoGuest286 From: http://www.google.de/search?sourceid=ch​rome&ie=UTF-8&q=websockets+demo (1 hits)
12:10 MojoGuest286 left #mojo
12:27 yko yakudzo: just mojo generate app and change 'welcome' action https://gist.github.com/733217
12:28 yko if you want to get name of current route that matched url from action - you can just access $c->match->endpoint
12:28 yko but if you want to get name of route that will match random path/url - that part is more tricky
12:33 yko i think, matching is not very userfriendly now. users are not supposed to rummage in routes by themselves
12:36 dotan joined #mojo
12:51 perlrocks Twitter: "perl-Mojo-e 'a ("/", sub {$ s = $ _ [0]; $ s-> render (format => "txt", text => $ s-> req-> headers-> to_string )})-> start 'daemon prints the browser title" (ru) --lattenwald http://twitter.com/lattenwa​ld/status/12489458163122176
12:52 fhelmber_ joined #mojo
12:58 otaviof joined #mojo
13:02 espent joined #mojo
13:57 koban left #mojo
13:58 koban joined #mojo
14:23 sri finding a route name by path makes no sense
14:24 sri if you want the current route you can always just do $self->url_for without a name
14:30 sri oh, salesforce bought heroku
14:33 dotan about disabling websockets, an Opera guy writes "This attack affects Flash and Java as well, but we have higher standards for browsers" ( http://annevankesteren.nl/2010/12​/websocket-protocol-vulnerability )
14:37 sri it's all basically bullshit
14:38 sri it's a proxy bug basically
14:38 sri they just want people to stop supporting draft 76
14:38 und3f[A] is now known as und3f
14:39 sri proxies can't handle upgrade headers, thats what it's all about
14:42 breaker313 left #mojo
14:45 sri they really just want to make sure people are going to adopt the next version
14:46 koban left #mojo
14:49 koban joined #mojo
15:29 gizzlon sri: why would'nt ppl adopt the next version?
15:33 spleenjack left #mojo
15:35 vel joined #mojo
15:47 espent left #mojo
15:50 sri gizzlon: because it breaks everything
15:50 sri and we'll have a lot of browsers stuck with draft 76 for a few months, since people upgrade slowly
15:51 sri every single web socket app will break
15:52 alnewkirk left #mojo
15:59 gizzlon really? Because of the CONNECT thing or something else?
16:05 gizzlon anaways, should'nt ppl expect things to change when they use something called "draft .." ;)
16:05 koban left #mojo
16:06 Christian have a nice evening
16:13 Christian left #mojo
16:24 sri gizzlon: the handshake and framing change
16:24 sri basically everything
16:26 sri the security problem will make it easier to drop backwards compatibility
16:26 sri now there is a reason why we can just ignore draft 76
16:26 sri for me as an implementor it's good
16:27 sri i don't have to support multiple versions
16:27 sri mojolicious will just move to ietf-4
16:28 sri ietf-4 + draft 76 support would have been very painful
16:30 sri "draft 76 is unsecure" is pretty much the perfect excuse :D
16:35 Akron left #mojo
17:06 perlrocks Twitter: "Mojolicious:: Lite you can create a large-scale Web apps - Getting Started with Perl code samples http://d.hatena.ne.jp/perlco​desample/20100913/1278596435" (ja) --nekoruri http://twitter.com/nekorur​i/status/12553618448187392
17:06 perlrocks Twitter: "RT @ nekoruri: Mojolicious:: Lite you can create a large-scale Web apps - Getting Started with Perl code samples http://d.hatena.ne.jp/perlco​desample/20100913/1278596435" (ja) --AE_PerlIsm http://twitter.com/AE_PerlI​sm/status/12553619492577281
17:10 perlrocks Twitter: "RT! Nekoruri: Mojolicious:: Lite you can create a large-scale Web apps - Getting Started with Perl code samples http://d.hatena.ne.jp/perlco​desample/20100913/1278596435" (ja) --perlism http://twitter.com/perlism​/status/12554513114206209
17:20 dotan left #mojo
17:26 otaviof left #mojo
17:30 ajgb left #mojo
17:49 ysyrota left #mojo
17:53 stephan48 sri: how does this bot works?
17:54 sri stephan48: what do you mean?
17:54 gabb its not a bot, sri pays a chinese kid to pretend to be a bot and search for mojolicious messages on twitter every 10 seconds
17:55 sri ^^^
17:55 stephan48 k
17:55 stephan48 how much do you pay?
17:55 stephan48 does it have any free ressources?
17:55 gabb don't tell him sri, he might be from the Chinese-Child-Labour-Union
17:55 sri :O
17:56 gabb <insert banksy simpsons intro>
17:56 sri the simpsons are made in korea though
17:57 gabb I think "doh!" is the correct response ^^
17:58 gabb I plan a massive mojolicious + graph-based db project which I shall spend all january on, can't friggin wait to code again for once
17:58 sri neat, which database?
17:58 gabb thinking neo4j via REST
17:59 sri ewww java
17:59 gabb not many strong graph db's out there
18:00 gabb flockdb is another candidate and so is infogrid, if I can work out licensing issues
18:02 gabb there really are quite few example apps and case studies / experience reports with graph db projects out there  :|
18:02 sri why does it have to be a graph db?
18:03 MadManchegan joined #mojo
18:03 gabb the stuff I want to store are basically graphs, a native way of storage seems like the best way to deal with it
18:04 sri i see
18:04 gabb I have shitloads of 1:n, n:n vertices to edges relations
18:04 sri huge graphs?
18:04 gabb ye, simple meta model, but the instances evolve into giant graphs
18:05 sri ah ok, for smaller ones you could also use riak
18:05 gabb to model it into a relational db would be rather painful
18:05 sri riak is a key/value store that allows linking
18:06 sri but once you go above 100 links for one object it becomes too much
18:06 gabb hm, but I need attributes on links, does it support that?
18:07 sri http://vimeo.com/14563219
18:07 sri there's a webinar for that
18:07 gabb cool, cheers
18:17 MadManchegan left #mojo
18:17 sri i like how simplistic it all is
18:17 sri yet damn powerful
18:18 rhaen hm.
18:19 sri neo4j should be better if you want to go all out with graphs
18:19 sri riak on the other hand gives you dead simple scaling
18:20 sri you just add more machines basically
18:20 sri the videos are all quite good btw
18:24 perlrocks Twitter: "Really good #riak screencasts and webinars. http://bit.ly/eiXRhC #mojolicious" --kraih http://twitter.com/kraih/status/12573203427758081
18:30 perlrocks Twitter: "RT @kraih: Really good #riak screencasts and webinars. http://bit.ly/eiXRhC #mojolicious" --insurgentsoft http://twitter.com/insurgents​oft/status/12574876858585088
18:32 sri damn, who would have thought, $212 million for heroku
18:42 janus gabb: depends how much you into sql and if your implementation supports recursive queries :)
18:43 janus s,i,are i,
18:44 janus i wouldn't say its dead simple, but as long as you only involve one recursive relation its pretty easy imho
18:45 gabb hmm, I think a schemaless solution would work better
18:46 gabb the data model is complex as it is, dealing with implementation restrictions along the evolution of this thing would be a nightmare
18:46 janus multi-dimensional trees or graphs are more challenging... but i guess that's the same on all architectures
18:48 janus sure, but imho it all depends on how you want to process your data
19:03 perlrocks Twitter: "RT @kraih: Really good #riak screencasts and webinars. http://bit.ly/eiXRhC #mojolicious" --zaa http://twitter.com/zaa/status/12583103209930752
19:10 MojoGuest863 joined #mojo
19:10 MojoGuest863 left #mojo
19:38 spleenjack joined #mojo
19:53 spleenjack left #mojo
20:00 fhelmber_ left #mojo
20:19 und3f is now known as und3f[A]
20:21 und3f[A] is now known as und3f
20:26 und3f is now known as und3f[A]
20:27 viking-3 joined #mojo
20:29 viking-3 good evening! i have one question about nginx deployment of Mojolicious, i'm not sure how it should work for serving static files, right now it doesn't, i have subfolders and files under the public folder, but requesting something there gives 404 in mojolicious logs
20:30 viking-3 the nginx config is basically proxying to the mojolicious daemon
20:30 sri i'm very sure serving static files works
20:31 sri you prolly just misunderstood something
20:31 sri static files have no prefix by default
20:32 viking-3 yes, the proxy pass is done matching on a location, as i have the webapp running under /something
20:32 viking-3 so i request /something/css/file.css
20:33 viking-3 i don't have any rewrite in nginx, and i do see the request in mojolicious logs
20:34 sri well, you need to rewrite somewhere
20:34 viking-3 i'm basically doing the example in the cookbok
20:35 und3f[A] is now known as und3f
20:35 sri i don't think you are
20:36 und3f is now known as und3f[A]
20:36 sri there is no prefix in the cookbook
20:36 viking-3 i also thought i need a rewrite for serving static files directly from nginx, but from what i understood, mojolicious should serve those aswell
20:36 viking-3 i meant, same thing but with location /something
20:37 sri well, thats not the same thing
20:37 viking-3 cause my app should serve that url, not /
20:38 viking-3 ok, then the default static works only for /
20:38 und3f[A] is now known as und3f
20:38 und3f is now known as und3f[A]
20:38 sri all your routes should be screwed
20:39 sri since the prefix changes everything
20:39 viking-3 routes work fine
20:39 sri ok, then yours are liberal enough
20:39 sri most of the time they would be screwed too though
20:40 viking-3 not sure what you mean, i have routes like /something/else, they work
20:41 sri right, your routes are deployment environment specific then
20:41 sri thats equal to moving all your static files into a /something directory
20:42 viking-3 under public you mean?
20:42 sri yes
20:42 viking-3 hmm... ok
20:42 viking-3 it's a bit messy indeed
20:43 und3f[A] is now known as und3f
20:43 viking-3 the thing is that i'm thinking of serving different paths with different mojolicious apps
20:43 sri app->static->prefix('/something') should work too
20:44 viking-3 i was reading that module, didn't try tho
20:44 viking-3 that's cleaner
20:44 und3f is now known as und3f[A]
20:45 viking-3 what's the best approach if i want /path1 -> app1 /path2 -> app2?
20:45 viking-3 different independent apps
20:45 sri you can always make it configurable
20:45 und3f[A] is now known as und3f
20:45 sri put the prefix in a config file
20:46 sri or do you mean entirely independent apps?
20:46 sri then it's of course two servers
20:46 viking-3 independent apps, that i can start stop without affecting the other parts
20:46 viking-3 yes also different daemons
20:48 sri i've been thinking about a hosting optimized daemon
20:49 sri that would run multiple apps in separate workers, possibly stop/start apps that are not requested for some time
20:50 sri but it's too much work for a side project
20:50 viking-3 so for my case i guess it's fine to have different location /foo, location /bar that proxy to different apps/daemons
20:50 sri i would have to start a hosting business for that :)
20:51 viking-3 with the static prefix thingy works \o/
20:51 sri not unexpected ;p
20:52 viking-3 ofc, you wrote it! :)
20:53 sri you could make the prefix longer to make nginx serve static files in the production setup
20:53 viking-3 longer?
20:53 sri like /something/public
20:53 sri /something would be the app and /something/public the static path
20:54 viking-3 or rewrite the single folders in /something...
20:55 sri there are many possibilities
20:56 sri i think you can also just make rules to cache .css/.js/.png files
20:56 viking-3 yes, i'll definitely do something on nginx for the static stuff
21:03 dvinciguerra left #mojo
21:23 ltriant joined #mojo
22:01 und3f is now known as und3f[A]
22:15 josh joined #mojo
22:31 vel left #mojo
22:39 viking-3 left #mojo
22:41 stephen left #mojo
22:46 stephen joined #mojo
22:53 perlrocks Twitter: "Mojolicious - Web framework for Perl http://www.findbestopensour​ce.com/product/mojolicious # FOSS" (no) --bestopensource http://twitter.com/bestopenso​urce/status/12641034416685056
23:08 MojoGuest653 joined #mojo
23:08 MojoGuest653 left #mojo
23:30 perlrocks left #mojo
23:31 perlrocks joined #mojo

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