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

IRC log for #mojo, 2014-03-30

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

All times shown according to UTC.

Time Nick Message
00:08 steveeirium joined #mojo
01:36 davido___ joined #mojo
01:45 klapperl_ joined #mojo
01:46 steveeirium ah so websockets have a ->max_websocket_size you have to partition your messages into
01:48 steveeirium or you get a "failed to inflate frame" error, points for surrealism.
01:50 jberger couldn't you just up the max size a little?
02:03 tempire sri!
02:03 tempire halp
02:03 tempire iphone app: eventbrite
02:03 tempire font used
02:03 tempire you have an eye for that sort of thing
02:10 tempire jberger: suggestions as to the message on mojocasts.com?
02:10 tempire I'll put it right on top, not sure about the wording
02:12 sleigner joined #mojo
02:51 steveeirium jberger: doesn't stop chrome erroring
02:54 sri "GET /vehicles/{id}/command/door_unlock"... the future is exciting! http://www.dhanjani.com/blog/2014/03/curosry-evaluation-of-the-tesla-model-s-we-cant-protect-our-cars-like-we-protect-our-workstations.html
02:55 sri tempire: i don't use an iphone anymore
02:55 tempire really. android?
02:55 * sri hugs his nexus 5
03:03 steveeirium is morbo making my app harder to kill when it's going crazy?
03:05 steveeirium hey we should try and make computers less shite to work with
03:07 hesperaux_ joined #mojo
03:17 dj_goku I thought I found a bug in M:L but I was using a text_field and trying to set it to a number type and it wasn't working. then I saw number_field exists. :D
03:19 dj_goku wow my M:L app is 425 lines with all html included. :D
03:27 dj_goku any that uses mongo how do you build your front-end view? using pure js or html?
03:37 mattastrophe joined #mojo
03:39 dj_goku_ joined #mojo
03:41 damaya_ question: what domain name do you like better: socialnot.es or mehnotes.com ?
03:42 damaya_ for a webapp.
03:47 jberger socialnot.es
03:49 damaya_ ok, I like that one too, but my wife and friends like mehnotes.com for some reason
03:49 damaya_ thanks for the input jberger :D
03:49 jberger tempire: something along the lines of: "mojocasts were made for mojo version X. since then a few things have changed, see the errata here (link)"
03:50 jberger either link to the wiki our add your own page
03:50 jberger what is meh?
03:50 purl well, meh is 'unexciting or unappealing' or http://www.youtube.com/watch?v=zSDj7bjAv2s or the MTV generation's opinion on everything
03:50 jberger hahahahaha
03:50 purl LOLCON 5 reached.
03:51 damaya_ meh, said in a certain way, kind of sounds like my, and meh is also the name of my balding, quite disgusting to look at, cat.
03:51 damaya_ And, as purl said, it is a sign of indifference, used by Gen Xers I would assume purl is saying.
03:54 avenj in general I would avoid giving products kind of depressing names
03:54 damaya_ hehe, yeah... it's too bad too, because there's a humor element in there that's kind of lost, unless of course you know my balding cat.
03:54 jberger I knew that meaning, but I'm with avenj, gotta be positive
03:55 damaya_ well, socialnot.es really is descriptive of what I've designed anyway.
03:55 jberger or else be really negative to be funny/memorable
03:55 damaya_ I think the biggest issue is that it is not a .com domain, and I hear non-techies don't often like that.
03:56 jberger can you get socialnotes.com too?
03:56 damaya_ taken, sadly... It's a nice site where you can send out announcements, such as "We are pregnant!"
03:57 jberger they are going to get a third of your traffic
03:57 damaya_ it's weird to think that a domain name can really make or break a webapp.
03:57 jberger totally
03:57 damaya_ a web site I mean.
03:58 dj_goku is not.es taken?
03:58 damaya_ and trying to come up with something that A) doesn't suck, and B) isn't taken is almost like trying to invent some new gadget, such as an amazing toaster that never burns bread,
03:58 damaya_ haha, yeah... taken
03:58 damaya_ I didn't even think of that one honestly. I just checked though, and yeah
03:59 damaya_ however, this domain registrar suggests that not-dead.org is still available.
03:59 damaya_ Register that then have it taken down when you are in fact dead.
04:00 dj_goku single word is the best!
04:01 damaya_ github started with less than $500 and made $15M last year... that's incredible.
04:02 jberger ournotes is taken
04:03 damaya_ Has anything recognizable been developed in Mojolicious?
04:04 damaya_ Say for example something kinda big, like Github.
04:07 dj_goku pap.er would be cool but...
04:08 jberger I think I agree though, non hackers don't even like .net sites. you need .com at least as a backup to the 1337 name
04:11 damaya_ pap.er is awesome actually... but yeah, I need a .com domain
04:11 damaya_ hm, well, back to thinking about it
04:11 damaya_ these are the silly things that stall my stuff from getting online
04:12 damaya_ heh
04:12 dj_goku are you done?
04:13 damaya_ hm, yeah almost... I am just working on the design at this point, but the backend code, database stuff, that is all done.
04:13 damaya_ now it's html, css, jquery, all that jazz
04:13 damaya_ and my domain name
04:13 damaya_ ;)
04:14 damaya_ Actually, I may change up the backend design a bit and use websockets. I wasn't sure how widely supported websockets are, but I decided recently that I don't really care.
04:20 jberger web sockets are supported widely enough
04:22 damaya_ yeah, I think so as well. I read an article recently about how one should not use websockets just yet, or at the very least support some sort of fallback method (e.g., long polling). Then I realized it was written by someone associated with Microsoft and I tossed it out of my brain.
04:23 dj_goku lol
04:24 damaya_ Once you realize it was written by someone using IE, working for MS, and arguing that older versions of IE are still widely enough used, you just kind of go, "Oh, that old argument."
04:25 damaya_ The only thing moving technology forward is that people finally started ignoring IE a bit. Stopped worrying so much about making it IE compatible.
04:25 jberger http://caniuse.com/#feat=websockets
04:26 jberger Google threw out ie6 support
04:26 jberger that was the big move
04:27 damaya_ Can you believe that of the IE usage, it is IE11 mostly used, followed very shortly by IE8?
04:27 damaya_ Wow, I guess I am one of the few people left still using Firefox.
04:28 jberger chrome is a powerhouse
04:28 jberger I do use Firefox for a few of the plugins
04:28 dj_goku this is a weird question, but can foreign keys be primary keys too?!
04:28 damaya_ I can honestly say I've never used it. It's never come standard with any Linux distro I know of.
04:29 jberger chromium is in the ubuntu repos now right?
04:29 damaya_ dj_goku, I believe they can
04:29 jberger when I say chrome, I use chromium
04:29 damaya_ Not sure actually. I've been using Gentoo for years, just switched to Xubuntu about a month ago (I love it). I will install and use it.
04:29 damaya_ I'm gonna miss firebug, but I imagine Chrome has something similar.
04:30 jberger chrome has it built in
04:30 dj_goku damaya_: developer tools
04:30 purl developer tools is free.  Download it and voila
04:30 dj_goku lol
04:30 dj_goku purl++
04:30 jberger except purl is wrong
04:31 jberger no need to download, built right in
04:31 dj_goku lol
04:31 dj_goku right that is what it is called.
04:31 dj_goku :D
04:31 dj_goku damaya_: chrome developer tools
04:31 dj_goku view -> developer -> BOOM
04:33 damaya_ I'm relieved my design looks the same in Chrome :D
04:33 dj_goku lol damaya_ yeah looks like I can do that.
04:33 jberger my only problem is that some of the ... shadier plugins haven't been allowed
04:33 dj_goku fk, fk + pk (fk,fk)
04:33 jberger videodownloadhelper is the big one for me
04:33 dj_goku well in sqlite
04:38 dvinciguerra joined #mojo
04:40 damaya_ I think I may actually like developer tools better than firebug
04:46 jberger it's pretty powerful
04:46 jberger But admittedly I haven't used firebug in ages
04:47 damaya_ they are roughly equivalent, but developer tools has some more features, and I think the interface is just a bit more intuitive and easy to work with
04:54 steveeirium coding is like teaching
04:54 steveeirium but you have to learn
04:54 steveeirium the google scramble
04:58 jberger learning to Google effectively is key
04:58 jberger ok nn all
04:59 mfung joined #mojo
05:05 mfung I want to remove the Date header, tried $c->res->headers->remove('Date')  at after_dispatch, but doesn't work. Which hook should I use?
05:12 jack joined #mojo
05:57 xantus mfung: before_dispatch?
05:57 purl somebody said before_dispatch was what I want
05:58 mfung xantus: ok, let me try...
06:01 mfung xantus: still doesn't work.
06:01 purl doesn't work. is having two has_many rels in one class legal?
06:04 russum joined #mojo
06:06 russum joined #mojo
06:11 xantus you'd have to paste some code somewhere (not here)
06:12 hesperaux__ joined #mojo
06:36 mfung seems fix_headers is called somewhere after dispatch...
06:37 mfung and Mojo thinks 'Date' is mandatory
06:44 xantus the rfcs say it is
07:08 Vandal joined #mojo
07:19 Adurah joined #mojo
08:48 laouji joined #mojo
08:56 malikai joined #mojo
09:16 mfung joined #mojo
09:18 mfung joined #mojo
09:19 malikai so in the wikipedia page on SSE, mojo is listed.. http://en.wikipedia.org/wiki/Server-sent_events
09:19 malikai but i don't see SSE in any mojo docs.. is it supported in mojo?
09:20 denisboyun joined #mojo
09:20 malikai also, if anyone knows, is it correct to say that SSE just feeds the client a chunk of data, then closes the socket?
09:21 malikai if so, how the hell do you run a websocket on that?
09:22 malikai (SSE says you can do a websocket on it)
09:32 denis_boyun__ joined #mojo
09:34 stephan48 malikai: maybe you are searching vor EventSource?
09:34 stephan48 for*
09:35 stephan48 http://mojolicio.us/perldoc/Mojolicious/Guides/Cookbook#EventSource_web_service
09:40 malikai oh, no, actually not SSE, SDPY is what i'm talking about
09:40 purl talking about is "Yeah, you're alway talking and talking and talking. Shut up and code."
09:41 malikai thanks, purl
09:41 purl de rien malikai
09:41 * malikai oils purl
09:44 malikai i just don't understand what net benefit you get from sending a chunk of data and then closing the socket.. furthermore i don't get how you can reasonably run a websocket on that?
09:45 malikai maybe i just don't understand what problem SPDY is meant to solve
09:47 basiliscos joined #mojo
09:58 sh4 joined #mojo
10:19 stephan48 what do you use for testing database backed applications?
10:20 denis_boyun joined #mojo
10:40 malikai most of my testing is with users.. i don't write too many public modules and these people won't pay for TDD
10:49 hummeleBop joined #mojo
10:53 alnewkirk joined #mojo
11:08 mire joined #mojo
11:24 al joined #mojo
11:30 ver joined #mojo
12:02 priodev joined #mojo
12:04 sri mojolicious will prolly never support spdy, just http/2 https://github.com/kraih/mojo/issues/423
12:07 sri to my knowledge nobody is working on it yet though
12:08 sri patches welcome
12:08 purl rumour has it patches welcome is always true or even for metacpan.org or swahili for "Put up or shut up."
12:29 malikai sri, in your opinion, what problem does SPDY solve?
12:40 punter joined #mojo
12:51 __noidd__ swahili for "Put up or shut up."
12:51 __noidd__ that's funny
12:52 __noidd__ HTTP/2.0 sounds like ipv6
12:53 __noidd__ it needs to happen but everyone's staying on ipv4 regardless
12:56 * moritz is on IPv6
12:57 sri malikai: https://www.varnish-cache.org/docs/trunk/phk/http20.html
12:57 sri i think it's mostly bullshit... but if it happens we have to support it
12:58 bedivere joined #mojo
12:59 bedivere left #mojo
13:02 sri lets reinvent TCP on top of TCP!
13:04 moritz TCCP!
13:06 sri and lets pretend SCTP never happened, because reinventing stuff is the google way!
13:07 sri (most of the http working group is now google folks)
13:10 __noidd__ right
13:14 steveeirium hey I'm making lysergic duct dape for everything
13:16 steveeirium protocols are protocols, we need a way to reel stuff in across cyberspace and figure out how the language works in a useful way
13:16 steveeirium it's coming together
13:17 d4rkie joined #mojo
13:17 steveeirium I'm listening to strange guitar music and rolling a smoke
13:18 steveeirium I think I want to put a perl debugger and code muncher in the browser with all my interpretations of data structures
13:22 sweet_kid joined #mojo
13:24 malikai sctp looks cool.. but it's unrealistic to believe it'll replace tcp anytime soon
13:24 malikai header compression would cause havoc too
13:25 malikai i could just imagine how hacky things would get when people try to be compatible
13:25 steveeirium yeha, the future is further up the stack
13:26 steveeirium it's like a big pyramid spurting humanity out the top
13:26 malikai what kinda smoke was that you were rollin?
13:26 malikai :)
13:26 steveeirium tubes of what where & how
13:27 steveeirium you only need to be sober if you get tangled in the depths with a debugger or something
13:28 malikai SPDY looks like a wet dog with fleas
13:29 malikai i do not see the problem SPDY solves that wasn't already solved by websockets.. why bother with headers when you have messages?
13:30 steveeirium yeah the why bother aspect is important
13:33 malikai i look at SPDY's so-called speedups.. i see gains of like up to 40%..
13:34 malikai i look at my market data platform doing websockets.. i see hundreds of sessions per t1.micro and practically no load
13:34 d4rkie joined #mojo
13:35 sri the http and websocket working groups don't seem to get along very well
13:35 malikai i wonder what rails numbers i'd see if i had users banging on a rest interface for the same data
13:37 sri and nobody seems to realize yet how bad http/2 will be for perl
13:38 sri (mojolicious is one of the few projects that will be fine though)
13:38 malikai you mean how a lot of CPAN stuff will be broken because of it?
13:38 sri no
13:38 sri http/2 uses concurrent streams, so you need threads or non-blocking i/o in every server
13:39 malikai ugh.. threads
13:39 malikai that would be good for mojo though :)
13:39 sri perl has no threads... so only servers that use non-blocking i/o will be fine after http/2
13:40 moritz well, perl has threads... kinda
13:40 sri GET OUT!
13:40 malikai perl wants to have threads
13:40 * moritz kackles evily
13:40 malikai i'd prefer to call perl threads 'hell'
13:41 sri call them what they are... fork emulation
13:42 sri the world would be much better if fork emulation was available on *all* perls... so you could test your portability code on os x and linux too... but i'm heading off topic...
13:44 malikai i think the world would be better if fork emulation was simply forgotten..
13:45 malikai anything that makes ugly code necessary is probably a problem
13:48 stephan48 81
13:50 jberger that post from varnish is good reading
14:32 __noidd__ the more complex we make the underlying protocols the more opportunity there will be for people to attack it
14:32 __noidd__ are we really at the point where we need to worry about header **compression** ?
14:33 __noidd__ I mean really, how big is a header in comparison to the insane bandwidths we have now
14:34 jberger and since the "headers" on websocket messages are basically nil anyway
14:34 jack_ joined #mojo
14:36 __noidd__ right - so why?
14:39 sri all that matters to google is mobile
14:39 dvinciguerra joined #mojo
14:40 sri jberger: problem then is caching of assets, you need meta data for that
14:55 __noidd__ mobie is the future for sure
14:56 sri solving the mobile latency problem on layer 7 seems silly though
15:00 sri http://www.guypo.com/technical/not-as-spdy-as-you-thought/ # also an intersting post about spdy/http2
15:01 sri the fact that http/2 as implemented in browsers will require every server to have an ssl certificate is going to be fun too
15:02 __noidd__ it's a cash cow
15:04 bedivere joined #mojo
15:06 bedivere hi, i'm stuck with route dispatch again :(
15:06 bedivere i have a route defined, but when i try to access the route, it shows me the "Page not found ... yet!" page
15:07 bedivere in the routes table on that page the route i am trying to access is listed
15:07 bedivere the controller and action behind that route are available to and they work because i am able to access them with a different route
15:09 bedivere debugging with a "die" shows that the action is not run at all, so my guess is, that the dispatch matches something else before actually matching the route i am trying to access
15:10 bedivere but how can i figure out what the dispatch mechanism is trying to do? are there some ways to get more information with a custom exception page?
15:35 jberger batman, around?
15:35 jberger I think I want to work on a "unique" attribute for fields in mandel
15:36 xchg joined #mojo
15:38 jberger and perhaps extend that to any kind of index creation too
15:38 xchg hi, is mojolicious session expiration time based on cookie expiration time or can I count on it?
15:41 * sri wonders if he should put minion into his mojoconf talk
15:41 sri suppose i'm going to consider it another spin-off project like mango... so it would make sense
15:41 tianon xchg: I'm fairly certain it's both (ie, it goes into the cookie expiration time and is stored in the session itself for mojo to validate)
15:42 tianon the source of "Mojolicious::Sessions::load" appears to agree with me, too
15:42 tianon return if defined $expires && $expires <= time;
15:42 xchg tianon: thanks
15:43 batman jberger: yeah
15:43 batman jberger: cool :) wanna do index as well while you're at it?
15:44 batman jberger: field "foo" => (isa => Int, unique => 1); ?
15:44 jberger a unique index is a special case of index, so it won't be hard to generalize
15:44 jberger yep, that's how I would call it
15:44 batman right. you could also do field_index "foo"; or something...
15:45 * sri uploads his daily minion release while thinking some more about it
15:45 tianon what about a multi-key index? :)
15:45 tianon and then a reverse index?  ie, { "bar": -1 } ?
15:45 jberger tianon: hmmmm, good point
15:45 purl nice and sharp
15:45 sri and index hints ;p
15:46 sri and text/geo indexes!
15:46 tianon YES! :D \o/
15:46 jberger in that case, perhaps batman is right to think about a separate keyword
15:46 tianon doesn't Mongo also support compound unique indexes?  or am I remembering that wrong?
15:46 batman what is reverse index ?
15:46 tianon http://docs.mongodb.org/manual/core/index-types/#ordering-of-indexes
15:46 jberger I still like allowing unique => 1 on field though
15:47 tianon "in compound indexes, the direction of the sort order can have a greater impact on the results"
15:48 batman jberger: it might get confusing. i think i vote +1 on a new keyword.
15:48 tianon http://docs.mongodb.org/manual/tutorial/create-a-unique-index/ -> "You can also enforce a unique constraint on compound indexes, as in the following prototype:"
15:49 jberger ok, then
15:49 jberger that will need more thinking about then
15:49 * sri is rather happy with the basic design of minion now... time to think about feature set again
15:51 sri building a sweet dashboard ui for minion will be fun
15:51 batman jberger: we might want to make initialize() easier to use instead. maybe we want something like sub initialize { my($self, ...) = @_; $self->ensure_index(...); } instead..?
15:51 gryphon joined #mojo
15:52 jberger batman, you and I are on the same wavelength, that was exactly the goal
15:52 batman \o/
15:53 batman not often i'm on the same wavelength with anyone :P
15:53 jberger calling initalize was going to walk the fields and call ensure_index to create the unique index
15:53 batman jberger: something else...care to add Toadfarm to the "new module" list which get announced by the bot?
15:54 * sri hooks up a radio jammer to the channel
15:54 jberger when I first created cpan_mojo I asked sri what its limits should be
15:54 jberger he requested, and I agreed, qr/Mojo/ seemed sane
15:55 sri a check for modules that depend on mojolicious would be more fun though ;p
15:55 batman okey
15:55 batman jberger: this is what i do now: http://p.thorsen.pm/8c063e8aabfb
15:56 batman which is indeed clunky...
15:56 jberger sri, I'm not opposed, but that wasn't originally what we discussed
15:56 batman anyhow... i need to go now.
15:56 sri jberger: i don't think you gave me that option :)
15:56 sri (it seems ridiculously expensive)
15:57 jberger it would need an addtion to the bot, it would have to call metacpan to get the deps
15:57 jberger there aren't that many cpan releases per time
15:57 jberger metacpan might care more than I do
15:58 sri true... it's not that many per day https://metacpan.org/recent
15:59 sri btw. job queues as a core feature for frameworks is in fashion
15:59 sri there's been talk about adding it to rails for quite some time and laravel already has it http://laravel.com/docs/queues
16:00 jberger I follow @cpan_new on twitter (which was the inspiration for cpan_mojo bot), every now and again, someone re-releases all of their modules, but that is rare
16:00 sri very common in the php world... prolly because of zend... but i'm only guessing
16:01 sri jberger: you could also keep a Mojo::Cache with the last 1000 or so modules so you don't have to recheck all the time
16:03 sri i'd guess you're left with maybe 10 requests per day then
16:04 jberger wow, til!
16:04 jberger I thought I knew the entire Mojo:: core namespace, I haven't ever seen Mojo::Cache
16:04 jberger sri++
16:04 sri it's tiny... easy to overlook :)
16:05 sri https://github.com/kraih/mojo/blob/master/lib/Mojo/Cache.pm
16:05 sri it's where we keep the last 100 compiled templates and matched routes
16:06 jberger yeah, it made sense as soon as I saw it
16:06 jberger I just didn't know it had been formalized into a module
16:07 jberger I like it, simple and effective
16:08 jberger anyway, using metacpan I would only get direct dependencies, not deeper ones, does that sound correct
16:09 jberger otherwise I would have to look into the cpantesters system which has deep dependencies
16:10 jberger for something like Mojolicious one level of dependency depth makes sense, but for say Mojo::IOLoop, deeper dependency is totally possible
16:13 jberger batman, do you use initialize for anything other than index creation?
16:19 voegelas joined #mojo
16:22 sri hmm, celery stores the task return value as job result
16:26 sri http://docs.celeryproject.org/en/latest/getting-started/first-steps-with-celery.html#application
16:27 sri in ruby and php that seems to be uncommon
16:27 jberger sri: it should be pretty easy to serialize and store the result of the job right/
16:27 jberger ?
16:27 sri trivial
16:28 jberger if there was a non-blocking way to watch a job for completion, then with that, ForkCall would be obsolete
16:28 sri also trivial
16:28 sri at least if you're ok with polling
16:29 jberger timers are basically polling right?
16:29 sri ye
16:29 jberger so yeah
16:29 sri push is also possible, but a little harder
16:29 purl okay, sri.
16:30 sri with a capped collection the workers would push ids of jobs that reached failed/finished state
16:30 sri to
16:31 sri the big problem with that however is testing
16:31 sri jobs with results you wait for are cute, but terrible to test
16:32 sri https://github.com/kraih/minion/blob/master/t/minion_lite_app.t#L40
16:32 jberger I guess polling jobs is more costly than timers, since it has to talk to the db
16:32 sri these kinds of jobs are easy
16:33 sri when your jobs have results and you non-blocking wait... then the ->get_ok() will never return, and you can't start the worker
16:33 jberger oh, right
16:34 jberger because the worker isn't in another process in the tests
16:34 sri right
16:34 * jberger is still getting his mind around the model here
16:35 sri one of the goals i had for minion was to make jobs very easy to test
16:35 jberger a worthy goal
16:41 dj_goku so if I get an db exception I am catching it, but not sure how to handle it with mojo. I just want to go back to a route and display a flash message. But what keeps happen with either a redirect_to or render is it keeps going through the subroutine.
16:48 asarch joined #mojo
16:49 punter joined #mojo
17:18 dj_goku ahh I just wrapped an if/elsif around the part and now it is working better
17:26 jberger dj_goku, a typical pattern is `return $c->redirect_to(...)`
17:27 jberger dj_goku, if I may also self-promote: https://metacpan.org/pod/Mojolicious::Plugin::Humane
17:35 russum joined #mojo
17:36 russum joined #mojo
17:44 dj_goku jberger: interesting.
17:47 dj_goku jberger: that sounds better than how I am thinking of doing it.
17:55 dj_goku jberger: what about stashing other messages?
17:55 dj_goku err
17:55 dj_goku other iteams.
17:55 dj_goku items.
18:00 dj_goku lol well I feel dumb
18:01 dj_goku jberger++
18:03 jberger :-)
18:07 dj_goku now I just need to figure out how to make it a bad alert!
18:07 dj_goku red!
18:10 jberger hmmmm, I could probably have a more extensible thing for humane
18:10 jberger { message => "failed login", action => "error" }
18:11 jberger otherwise you can provide your own template
18:12 dj_goku maybe, still trying to figure this all out. :) mojo included.
18:19 sujithm joined #mojo
18:20 basiliscos joined #mojo
18:30 punter joined #mojo
18:34 dj_goku jberger: thanks again!
18:37 mire joined #mojo
19:04 * jberger is testing the next iteration of cpan_mojo
19:17 noidd joined #mojo
19:18 noidd Do people actually use mojo as the actual webserver in production systems with SSL?  ... or do people always front through apache or ngix etc?
19:24 dj_goku jberger: cpan_mojo?
19:34 mire joined #mojo
19:53 ua joined #mojo
19:59 denis_boyun joined #mojo
20:05 espent joined #mojo
20:12 punter joined #mojo
20:47 mudler_ joined #mojo
20:48 jhg joined #mojo
20:50 meshl joined #mojo
20:56 marty_ joined #mojo
20:57 marty_ joined #mojo
20:58 marty__ joined #mojo
21:04 marty joined #mojo
21:04 jack joined #mojo
22:18 jhg joined #mojo
22:27 iftekhar25 joined #mojo
22:32 jhg_ joined #mojo
22:57 cpan_mojo joined #mojo
22:58 cpan_mojo joined #mojo
22:58 jberger well I couldn't finish in time, so the old cpan_mojo is back for now
23:55 * sri thinks he might have a good solution for the testing problem :)
23:55 sri $t->app->minion->auto_perform(1);
23:55 sri then whenever you call $c->minion->enqueue() there's an implicit ->all_jobs
23:56 sri makes waiting for return values possible :)
23:56 sri you set it once in your tests, and no more manual ->one_job/all_jobs either

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