The web in a box - a next generation web framework for the Perl programming language

IRC log for #mojo, 2013-11-22

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

All times shown according to UTC.

Time Nick Message
00:05 sri of course
00:35 ronsavage Via PayPal?
00:53 crab i suspect sri was referring to code contributions
00:57 ronsavage Perhaps, or perhaps not
01:00 btyler joined #mojo
01:01 crab are you saying my uninformed speculation could be wrong? unthinkable.
01:03 sri oh, i was referring to code contributions
01:04 sri paypal and gittip links are on the cpan profile https://metacpan.org/author/SRI
01:06 mire_ joined #mojo
01:08 Adurah_ Bitcoins?
01:08 sri haha, people donate bitcoins?
01:09 Adurah_ Bitcoin users like using bitcoins for anything they can.
01:09 Adurah_ My only consistant usage is Namecheap.
01:24 laouji joined #mojo
01:25 d4rkie joined #mojo
01:25 ronsavage Done. But in Australia (don't know about other places) PayPal warns it takes 3-5 days to clear. Pathetic...
01:29 sri ronsavage: thanks, inbox says it arrived
01:29 * sri loves gmail on android
01:40 crab you use android now?
01:40 ronsavage That was quick. But don't spend it quite so fast!
01:45 sri hookers and blackjack!
01:46 Adurah_ Now we're talking bitcoin!
01:46 sri crab: yea, got a nexus 5 recently
01:57 sri hmm, is there a way to split up long L<...> sections in pod into multiple lines?
02:04 sri btw. if anyone really wants to send bitcoins... the id is 1Cid78CmK4hvf78Ry8K2XeDx8pQHNh4hbz
02:04 sri apparently metacpan is not cool enough to have a field for that yet!!!1
02:06 jberger does anyone think it would be worth rewriting AnyEvent::Util::fork_call using Mojo::IOLoop (and perhaps File::Map for interprocess messaging)?
02:07 sri sounds painful
02:08 sri sockets are prolly best for portable ipc
02:08 jberger fork_call uses storable
02:09 sri to files?
02:09 jberger yeah, its only for the result, not really two-way IPC
02:09 sri Oo
02:10 sri i guess it makes sense for one shot operations
02:10 jberger so far fork_call is the easiest way I can find to do async wrapping of generic blocking operations
02:11 jberger its really just AE::io watcher and storable and some pure-perl wrappings
02:11 sri it's also the only one i know that's really portable
02:11 jberger exactly
02:11 * sri had trouble in some form or another with everything else
02:12 jberger now that I'm done with YAPC::Brasil, I'm starting to plan part 3 of my async series
02:12 sri dunno why you would want to use File::Map though
02:12 jberger it was just a thought
02:13 jberger seemed faster than serializing to disk
02:13 sri having something like that mojolicious optimized would surely make a few people happy
02:14 sri you need to read and write the whole file, i don't see it being faster
02:14 jberger File::Map allows mapping anonymous memory without a file backing it
02:15 sri or am i missing something?
02:15 sri hmm
02:15 jberger though admittedly I haven't actually used that
02:15 jberger so I am just hoping it would work like I imagine that would
02:15 sri does that work on windows?
02:15 jberger it claims to
02:35 duncanthrax2 joined #mojo
02:37 jzawodn jberger++
02:37 jzawodn moar async series
02:37 jberger \o/
02:38 jberger I also need to do a post about having several delay steps, the second post only used one step for parallel processes
02:39 jberger so these are 3 and 4 in some order
02:39 hotsnow joined #mojo
02:39 jzawodn yeah, the delay stuff was interesting and educational for me
02:39 jzawodn I got a good prototype going today of a bit of infrastructure using Mojo::IOLoop and Mojo::UserAgent after sri told me I was being dumb
02:39 jberger hehe
02:39 jberger I think the delay is one of my favorite utilities, possibly in all of cpan
02:40 jberger I do like File::chdir
02:40 jberger hmmmm, list of favorite utilities ...
02:41 jzawodn there's a lot to like on cpan.  and even more to avoid!
03:24 jberger oh, god, I can't read that code anymore
03:24 jberger ok AE fork_call for me
03:25 d4rkie joined #mojo
03:29 jberger otherwise known as ... windows <s>sucks</s> is hard
03:51 jberger hmmm, fork_call doesn't serialize to disk
03:51 jberger nevermind, I missed a big something
03:52 jberger File::Map with anonymous memory might be more portable/less confusing
03:52 jberger but might not be any faster
03:53 sri storable+sockets is what i would use for portability
03:59 beyondcreed joined #mojo
04:06 abra joined #mojo
04:07 preflex_ joined #mojo
04:21 ronsavage joined #mojo
04:33 gtodd joined #mojo
05:15 laouji joined #mojo
05:41 btyler joined #mojo
05:57 ryozi joined #mojo
05:59 dotandimet joined #mojo
06:11 laouji joined #mojo
06:52 hrupp joined #mojo
06:58 * batman doesn't get File::Map. is there any better synopsis, or do i know "something else" to get what it's doing?
06:59 Adurah_ "loose the mapping"
06:59 Adurah_ Guess it has tighter mapping.
07:02 crab batman: mmap?
07:02 purl hmmm... mmap is recent
07:03 crab from what jberger said he wanted to use it for, i'd guess mmap with MAP_ANON|MAP_SHARED
07:06 batman crab: guess i need to learn something new :)
07:09 dod joined #mojo
07:12 dotandimet joined #mojo
07:14 KindTwo joined #mojo
07:17 dod joined #mojo
07:25 dpetrov_ joined #mojo
07:29 fhelmber_ joined #mojo
07:31 Vandal joined #mojo
07:40 basiliscos joined #mojo
07:41 Adurah_ map_anonymous, such memory savings.
07:41 Adurah_ Not sure if that's jb's concern, though.
07:44 batman Adurah_: don't you have tab completion on nicks in your irc client?
07:44 Adurah_ That I do.
07:44 batman ok.
07:44 Adurah_ Under some circumstances I avoid highlighting.
07:44 batman i see...
07:46 Adurah_ If his concern was memory savings, perhaps I'd have.
07:46 dotandimet joined #mojo
07:55 Adurah_ https://gist.github.com/anonymous/7596385 Something like that, batman.
08:02 batman i think you should give up. i really don't see how that is helpful.
08:02 maxhq joined #mojo
08:02 Adurah_ I'm saying that's likely his intent to communicate between processes.
08:03 Adurah_ Instead of sockets, shared memory.
08:06 Adurah_ My link was an attempt at a synopsis that'd make more sense.
08:08 denis_boyun joined #mojo
08:09 zanity joined #mojo
08:16 dotandimet joined #mojo
08:27 trone joined #mojo
08:35 denis_boyun_ joined #mojo
08:48 hummeleBop joined #mojo
09:11 themage joined #mojo
09:19 nicomen is it possible to use around_action, and capture the render output of a call to $action->()?
09:19 nicomen I meant $next->(), not $action->()
09:36 dbr joined #mojo
09:39 nicolaas joined #mojo
09:57 dotandimet joined #mojo
10:00 batman not if it's async
10:01 batman nicomen: what do you really want?
10:02 nicomen alter format on the way in, add text around the rendered content on the way out
10:09 kwa joined #mojo
10:10 batman why can't you just add text in the fly?
10:11 batman inside the template or whatever.
10:11 hotsnow joined #mojo
10:19 nic joined #mojo
10:28 basiliscos joined #mojo
10:44 mire_ joined #mojo
10:44 gianpaulo joined #mojo
10:49 basiliscos1 joined #mojo
11:01 basiliscos joined #mojo
11:08 athenot joined #mojo
11:20 denisboyun joined #mojo
11:39 basiliscos joined #mojo
11:40 stig joined #mojo
11:52 basiliscos joined #mojo
11:59 punter joined #mojo
12:03 punter joined #mojo
12:06 ver joined #mojo
12:10 fhelmbe__ joined #mojo
12:11 Adurah_ joined #mojo
12:40 fhelmber_ joined #mojo
13:03 fhelmbe__ joined #mojo
13:26 cooper joined #mojo
13:42 dotandimet joined #mojo
14:02 dvinciguerra joined #mojo
14:11 denisboyun joined #mojo
14:17 dotandimet joined #mojo
14:20 bluescreen joined #mojo
14:29 arthas joined #mojo
14:38 Adurah joined #mojo
14:39 fhelmber_ joined #mojo
14:45 btyler joined #mojo
14:50 marty o/
14:50 tianon "Don't communicate by sharing memory; share memory by communicating. "
14:51 * tianon hides
14:51 marty for anyone interested.  I ran across a pretty good open source  mongodb admin gui today.    http://robomongo.org/
15:01 human39 joined #mojo
15:03 btyler batman: I was just thinking about writing something smaller and crappier than Mojo::IOLoop::ReadWriteFork the other day. thrilled to see you beat me to it :)
15:07 denisboyun joined #mojo
15:08 mire_ joined #mojo
15:17 nicomen batman: because I don't want to have to add the same thing in every controller
15:17 nicomen or even better, I don't want to rely on the fact that other people will remember to do that
15:18 nicomen and it would be really stupid, I just want to wrap all json controllers with the possiblity to serve it as jsonp if wanted
15:18 sh4 joined #mojo
15:20 basiliscos joined #mojo
15:24 tholen42 joined #mojo
15:28 gryphon joined #mojo
15:39 rem_lex joined #mojo
15:54 rem_lex joined #mojo
15:55 hesperaux joined #mojo
16:01 winslow4 joined #mojo
16:04 salparadise 16:45:55 <@sri> paypal and gittip links are on the cpan profile https://metacpan.org/author/SRI << cool
16:07 S joined #mojo
16:07 S hi
16:08 S how to get larger json data fastly in mojolicious
16:09 S in frontend i am using extjs and using mojolicios as rest service
16:11 btyler S: have you hit a bottleneck yet? if you know for a fact that Mojo::JSON is the thing slowing you down, dropping in JSON::XS is pretty easy. I think there's even a mojo wrapper around JSON::XS so the API is the same
16:13 S i am not sure... abt that but i am retriving data from sybase which has millions of records
16:14 S and i will load atleast 40000 records
16:15 S i can try json::xs instead mojo::json
16:15 btyler well, like I said, using JSON::XS is pretty easy to do. I haven't used Mojo::JSON::XS personally, but its tests are all passing
16:16 S okay i will try
16:16 batman nicomen: that's you you refactor your code and write unit tests
16:16 batman *why
16:16 batman btyler: glad you like it! :)
16:17 btyler batman: hit a minor snag, some of the t/cat.t tests are timing out, was just poking at it
16:17 btyler or rather, they're hanging and the test process hits 1800 seconds or whatever and gives up
16:17 batman btyler: please let me know if you use it in an actual web-app. i haven't done that myself, so i'm interested in seeing how it could work smoother with mojolicious::controller
16:17 batman btyler: which version is this?
16:18 btyler cloned latest from github
16:18 btyler 0.03 was the one that cpanm grabbed
16:18 btyler both hung on t/cat.t tests
16:18 batman that's superweird. which operating system?
16:18 btyler OSX 10.9
16:19 btyler perl 5.18.1
16:19 purl i think perl 5.18.1 is released now
16:19 btyler oh, haven't updated my mojo in a few weeks, that might be it
16:19 batman i don't think that's it, but please upgrade it...
16:19 btyler yep, already going
16:20 batman so... if you do prove -vl t/cat.t then you simple don't see any output?
16:21 btyler https://gist.github.com/anonymous/02f21bdf1f01710027a7
16:22 btyler https://gist.github.com/anonymous/c717e92a46119b0b2668 there's what I was getting from make test
16:22 batman btyler: but why doesn't it time out after 3 seconds..?
16:22 batman must be some bad while() logic somewhere
16:23 btyler anyways, if it works for you I'll poke at it after work today, if I figure it out I'll send a PR or confirm that something is screwy with my setup
16:23 batman do you have something like "strace" i osx?
16:23 mire_ joined #mojo
16:24 sri muhahahaha.... now i really want your bitcoins!!! http://www.coindesk.com/virgin-galactic-will-accept-bitcoin-space-travel-says-billionaire-entrepreneur-sir-richard-branson/
16:24 sri SEND SRI TO SPACE!
16:28 batman btyler: http://www.cpantesters.org/distro/M/Mojo-IOLoop-ReadWriteFork.html # waiting for 0.03 test results :)
16:28 btyler batman: no strance, but dtruss is apparently similar, lists system calls. it screwed with the test env somehow though, caused the first test to fail
16:28 btyler but I don't have much experience with those tools, will do some reading later
16:29 batman btyler: if you don't think there's any secrets, you can pastebin it to me.
16:29 batman but run with WAIT_PID_INTERVAL=0.1 dtruss prove t/cat.t;
16:30 sri way to change the topic :(
16:30 batman too low WAIT_PID_INTERVAL will give you a lot of output i would suspect...
16:30 btyler by "do some reading later" I meant "mine some bitcoins to SEND SRI TO SPACE"
16:30 sri -.-
16:31 batman sri might pull Mojo::IOLoop::ReadWriteFork into pieces... maybe that could be useful :)
16:34 asarch joined #mojo
16:34 cfedde joined #mojo
16:38 hrupp joined #mojo
16:45 dotandimet joined #mojo
16:47 delias joined #mojo
17:33 dod joined #mojo
17:33 beyondcreed joined #mojo
17:44 batman odc: the module actually support coderef already :D
17:44 batman i just forgot to document it
17:44 odc o_O
17:44 odc excellent :)
17:45 batman the code is quite old. that's my best excuse :)
17:45 batman and it's from work, so the doc is not as good as my other cpan packages
17:45 d4rkie_ joined #mojo
17:45 batman at least i like to think so.
17:53 bluescreen joined #mojo
17:58 d4rkie joined #mojo
18:26 sawtooth joined #mojo
18:28 batman odc: sorry. it doesn't work. but it will work in the next version.
18:39 Adurah_ joined #mojo
18:39 gryphon joined #mojo
18:40 tholen42 joined #mojo
18:41 sawtooth joined #mojo
18:53 batman btyler++ # testing my module
18:54 batman odc: 0.04 will soon be available with code ref support
18:55 odc batman, are you implying you want me to work during my weekend? ;)
18:56 batman odc: https://metacpan.org/pod/release/JHTHORSEN/Mojo-IOLoop-ReadWriteFork-0.04/README.pod#start
18:56 batman odc: haha. that's up to you.
18:57 batman i'm out of here
18:57 batman metacpan++
18:58 odc bye!
18:58 odc and thanks again
19:10 basiliscos joined #mojo
19:12 Mike-PerlRecruiter_ joined #mojo
19:20 rem_lex joined #mojo
19:21 mire_ joined #mojo
19:25 hummeleBop joined #mojo
20:13 punter joined #mojo
20:40 mire_ joined #mojo
22:38 gryphon joined #mojo
22:38 asarch joined #mojo
23:47 btyler joined #mojo
23:55 ronsavage joined #mojo

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