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

IRC log for #mojo, 2014-09-25

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

All times shown according to UTC.

Time Nick Message
00:22 doublelel joined #mojo
00:26 woz joined #mojo
00:55 zivester joined #mojo
00:58 woz joined #mojo
01:15 Mso150 joined #mojo
01:41 d4rkie_ joined #mojo
01:45 woz joined #mojo
02:04 dvinciguerra joined #mojo
02:13 noganex_ joined #mojo
02:45 woz joined #mojo
02:47 Adura joined #mojo
02:57 disputin joined #mojo
03:05 basic6_ joined #mojo
03:07 marty joined #mojo
03:19 sujithm joined #mojo
03:43 Adurah joined #mojo
03:44 sri quick heads up, i'm most likely going to stop maintaining Mango
03:45 woz joined #mojo
03:47 sri i'm tired of the shit mongodb inc is pulling... instead of fixing things they just push more responsibility on driver authors
03:47 sri this is where i draw the line http://emptysqua.re/blog/server-discovery-and-monitoring-spec/
03:48 sri i'll write something to the list when i know exactly what i'm going to do
03:49 sri (there is more to this, some folks made promises earlier this year and broke them...)
03:54 sri there won't be much drama though, i'll just move on to something more fun
04:04 firnsy fwiw /me loves Mango
04:05 firnsy sri++ # less drama && more fun
04:16 ningu joined #mojo
04:17 ningu hi, I have a Mojo::DOM question, anyone around?
04:18 ningu I have some HTML like so: <a href="...">foo bar <b>baz</b> quux</a> and I want to loop through the child notes of <a> including the text nodes
04:19 ningu apparently $a->children will not do this
04:19 good_news_everyon joined #mojo
04:19 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/GONE6g
04:19 good_news_everyon mojo/master d1fd06e Sebastian Riedel: Mango is no longer recommended
04:19 good_news_everyon left #mojo
04:20 ningu aha, I need to do contents and not children, looks like. :)
04:26 irq joined #mojo
04:45 woz joined #mojo
04:48 preaction joined #mojo
04:56 sujithm joined #mojo
04:59 doublelel joined #mojo
05:24 tianon sri: how in the heck is that easier than just letting the sysadmin say, "here's the IPs of the nodes in my cluster that you should try" ?  it's certainly not easier to debug...  >.<
05:45 woz joined #mojo
06:10 sujithm joined #mojo
06:11 dod joined #mojo
06:12 KCL_ joined #mojo
06:19 irq joined #mojo
06:20 mr-foobar joined #mojo
06:40 dp_ joined #mojo
06:45 woz joined #mojo
06:47 jamesaxl joined #mojo
06:53 sugar joined #mojo
07:00 bobkare joined #mojo
07:04 Vandal joined #mojo
07:16 sujithm joined #mojo
07:19 marcus sri: Sorry about all the work you put into Mango :-/
07:25 trone joined #mojo
07:28 bc547_ too bad.. I was just about to start developing a large project with Mango :-(
07:30 axlrose joined #mojo
07:34 jamesaxl joined #mojo
07:35 sugar_ joined #mojo
07:37 _eugen joined #mojo
07:38 fhelmber_ joined #mojo
07:42 Mso150 joined #mojo
07:44 jamesaxl joined #mojo
07:44 rawler joined #mojo
07:45 woz joined #mojo
08:08 woz joined #mojo
08:12 basiliscos joined #mojo
08:12 woz joined #mojo
08:15 irq_ joined #mojo
08:17 irq_ joined #mojo
08:40 mjb152 joined #mojo
08:43 aleksey joined #mojo
08:55 mjb152_ joined #mojo
09:02 mjb152_ I'm looking to find a route which would catch  /json   /json/  /json/xx  , is that possible with a wildcard placholder ?  or do I have to define more than 1 route ?   the bit I struggled to catch is /json   (with no trailing /)
09:02 mjb152_ I'd forward that placeholder to a controller , and deal with the different options inside there.
09:06 doublelel joined #mojo
09:07 denis_boyun joined #mojo
09:11 ver joined #mojo
09:16 Adura joined #mojo
09:31 batman mjb152_: $app->routes->get('/json/:placeholder', [ placeholder => '' ])->to("ctrl#action");
09:34 mjb152_ thx, trying it out now
09:38 laouji joined #mojo
09:39 mjb152_ hmm, I get "none of these route could generate a rsponse"  for .. $r->route->get('/json/:jtext', [ placeholder => '' ])->to("site#json");
09:39 mjb152_ +/json/:jtext
09:39 mjb152_ GET
09:39 mjb152_ jsonjtext
09:41 sugar__ joined #mojo
09:45 GillesM joined #mojo
09:47 batman mjb152_: don't paste
09:47 mjb152_ sorry
09:48 basiliscos joined #mojo
09:48 meshl joined #mojo
09:48 batman mjb152_: it should be $app->routes->get('/json/:placeholder', { placeholder => "" })->to("ctrl#action");
09:48 batman sorry about that
09:48 batman mjb152_: http://mojolicio.us/perldoc/Mojolicious/Lite#Optional-placeholders
09:50 mjb152_ :-)  works perfectly, thanks very much.
09:50 sugar joined #mojo
09:53 mjb152_ got to confess I did look in that area, but didn't make the leap that setting it to "" would be the right thing to do.  thanks again
09:53 bayashi joined #mojo
09:54 batman note: you can set it to any value you like. "" or is just easier to see if anything is present: do_stuff() if length $self->param('placeholder');
10:03 Adura joined #mojo
10:06 cosimo joined #mojo
10:16 Adurah joined #mojo
10:51 cfedde joined #mojo
10:51 cosimo joined #mojo
11:00 ua joined #mojo
11:04 mr-foobar joined #mojo
11:04 Adura joined #mojo
11:07 sugar_ joined #mojo
11:12 aleksey joined #mojo
11:12 Adurah joined #mojo
11:14 ichesnokov joined #mojo
11:22 Adura joined #mojo
11:26 D4RK-PH0ENiX joined #mojo
11:29 Eitz joined #mojo
11:31 neilhwatson joined #mojo
11:45 nicomen marcus: batman: it seems like I'm stuck in a loop if trying to connect to an irc server with an errournous nickname, sounds familiar?
11:45 nicomen (convos)
11:45 Nei cant you change the nick?
11:46 batman i've heard about it, but i can't remember the details. have you tried checking your nick in the connection details and then restart the connection?
11:48 batman nicomen: ^
11:50 nicomen I'm stuck in the wizard per se
11:50 nicomen I'm looking into the code of course, it seems there is no check that irc->nick(...) actually succeeded
11:50 nicomen but I'll dig further
11:51 jamesaxl joined #mojo
12:01 nicomen batman: is there an admin to remove other people's connections?
12:02 batman no. https://github.com/Nordaaker/convos/issues/99
12:03 meshl joined #mojo
12:19 lipizzan joined #mojo
12:37 Kripton joined #mojo
12:51 basiliscos1 joined #mojo
13:00 sri marcus: me too :(
13:01 sri but if i keep going it will only become more painful
13:01 marcus sri: I agree, I don't trust mongodb
13:01 sri tianon: yea... mongodb inc just keeps making stupid decisions...
13:02 marcus sri: Postgresql http api seems promising tho! :)
13:02 marcus <3 postgres
13:02 sri and when i talk to folks working there in private... it's just like... "yea...we know"
13:03 * Lee wonders how much of those decisions come from the founder
13:11 sujithm joined #mojo
13:17 genio ugh. mongo--
13:17 Kripton joined #mojo
13:18 mishantil I was very enthusiastic about mongodb, and then I read "MongoDB - The Definitive Guide" (O'Reilly). Enthusiasm-B-Gone
13:18 sujithm joined #mojo
13:18 jberger_ sri: so do we have a recommended async db solution?
13:19 * jberger_ votes that sri should write an asynchronous database!
13:20 genio DynamoDB!   *ducks*
13:20 jberger_ ;-P
13:21 jberger_ Or maybe i should just have a server instance with zmq transport
13:21 jberger_ Using whichever server
13:22 genio at least I'm early enough along into this project to stop using MongoDB+Mango.  I wasn't a fan of Mongo anyway
13:22 jberger_ *using whichever db
13:23 jberger_ Fwiw,  I like Mongo/Mango enough for small projects
13:27 dvinciguerra joined #mojo
13:30 genio I had the (dis)pleasure of hearing a MongoDB talk at a conference I went to a few years ago on application security.  There wasn't much in the sense of user privileges at the time and his suggestion was to rename functions so nobody would know how to communicate with the server if they got in
13:34 basiliscos joined #mojo
13:34 sri jberger_: we do not
13:39 odc so who is gonna maintain Mango now? :D
13:45 disputin joined #mojo
13:45 disputin1 joined #mojo
13:52 aleksey joined #mojo
14:05 rem_lex joined #mojo
14:12 mib_terxgl joined #mojo
14:13 sh4 joined #mojo
14:16 ryozi joined #mojo
14:17 davido___ joined #mojo
14:25 sugar__ joined #mojo
14:28 sri odc: i suspect it might get abandoned
14:30 odc not while I still use it in production!
14:31 Kripton joined #mojo
14:32 odc i hope there will be someone more competent than me to take care of it though
14:37 sri you're of course welcome to fork it
14:39 sri but honestly, mongodb 2.8 is just around the corner with new wire protocol changes
14:39 sri so i'd rather deprecate it gracefully than let it rot with bad maintenance
14:40 odc right
14:40 odc sri, so what's your new DB of choice? Postgre?
14:40 sri i don't know
14:41 odc text files!
14:41 purl text files are gzipd, and it uses Compress::Zlib to uncompress?
14:41 sri aside from databases that use HTTP, using pretty much anything from Perl non-blocking sucks
14:42 sri and redis with it's super minimalistic protocol is allright too
14:43 sri one thing is for sure, i'm done with custom protocols
14:43 sujithm joined #mojo
14:56 dod joined #mojo
15:03 jamesaxl joined #mojo
15:03 disputin joined #mojo
15:04 Kripton_ joined #mojo
15:09 axlrose joined #mojo
15:13 jamesaxl joined #mojo
15:13 mad_hatter joined #mojo
15:13 mad_hatter hey folks
15:13 mad_hatter can anyone point me to a good tutorial on doing REST with mojo?
15:21 basiliscos joined #mojo
15:21 marcus sri: elasticsearch.
15:21 purl hmmm... elasticsearch is nice and easy too
15:21 marcus ;)
15:24 sri marcus: then i might as well use redis
15:24 marcus sri: They are very different approaches.
15:25 sri what they have in common is that both are not real databases
15:25 marcus sri: what does that even mean
15:25 marcus redis is single threaded and lets you do atomic transactions with lua
15:25 sri i vaguely remember you laugh about mongodb being fragile
15:26 marcus it's an ops nightmare, more lik
15:26 marcus like
15:26 sri i'd like my database to actually make sure my data is safe
15:26 sugar joined #mojo
15:26 sri neither elasticsearch not redis do that
15:26 sri *nor
15:28 sri you're insane if you store sensible data in either
15:29 marcus sri: I guess it all depends on your requirements. I wouldn't use either to store economic transactions.
15:29 marcus I'm fine with using either for storing my IRC tho. :)
15:30 marcus But then, I'd rather shoot myself than use mongodb for economic transactions either.
15:30 sri then why would we recommend a database like that?
15:31 sri the question is what we will use in example apps
15:31 marcus sri: systems that do provide that kind of security often has other limitations.
15:31 marcus for instance, postgresql is very hard to scale up.
15:31 marcus (to multi master scenarios, slaves are fine)
15:39 batman Databases are hard. Let's go shopping.
15:39 batman :(
15:40 batman It sucks big time that mongodb is awful. I was really thinking Mandel was getting cool :)
15:41 odc i know the perl guys at Weborama are pretty happy with Riak
15:41 batman I'm sorry you have spent so much time on rot sri. That is indeed bad news ????
15:44 mst they're optimising for simpler deployments at the expense of driver authors
15:44 mst basically because a lot of mongodb's potential user base are fucking terrible at ops
15:45 mst it's ... a trade-off
15:45 mst I can see why they made it. I can see why sri's pissed off by it.
15:46 sri that's incorrect, the latest fuckup is no tradeoff, just bad decisions
15:48 odc i think what sri wants is CouchDB. It's just like MongoDB but with an http interface
15:48 odc and failover is handled by an http proxy
15:48 sri no ad-hoc queries is a non-starter
15:48 odc what's a ad-hoc query?
15:51 mst sri: I think the rationale is "making non-shard deployments simple"
15:51 sri mst: there is a simple solution https://jira.mongodb.org/browse/SERVER-1594
15:51 mib_0en9eo joined #mojo
15:52 sri but they said "meh... better have all the driver authors implement it individually"
15:52 sri dozens of times, in 3 variations
15:53 mst right, but then somebody'd have to deploy mongos
15:56 mst whereas if you make the client authors do all the work, they don't
15:56 mst SIMPLER
15:56 mst http://trout.me.uk/mongodb.jpg applies
16:05 c-- joined #mojo
16:16 c-- how would a helper know that it has been called like app->helper() rather than controller->helper() ?
16:17 moritz c--: shift->isa('Mojolicious::Controller')
16:18 _eugen joined #mojo
16:21 c-- moritz: seems the helper always gets a Mojolicious::Controller object as this reference, even if app->helper was called...
16:22 moritz c--: huh.
16:22 c-- yeah i'm confused too, maybe i'm doing it wrong
16:25 aleksey joined #mojo
16:26 Mso150 joined #mojo
16:26 sri i think i'm going to dig into the async features of certain DBD modules for now
16:28 sri they all perform i/o blocking, but allow async waiting for queries... it's not elegant, but performance wise it's not terrible
16:29 c-- moritz: have you tried this? http://paste.debian.net/123064/ i thought $self in the helper should be a Mojolicious (app) object rather than a controlle
16:29 odc sri, just sleep it off, calm yourself, and in a few days you can start hacking on Mango again
16:31 sri odc: i've been thinking about this for a few days actually
16:31 odc fuck
16:33 sri in fact, i'm writing the mango deprecation announcement right now
16:36 saki joined #mojo
16:43 disputin joined #mojo
16:44 basiliscos joined #mojo
16:48 preaction joined #mojo
16:49 aleksey joined #mojo
16:50 sri https://groups.google.com/forum/#!topic/mojolicious/GFsXFQF3t-k
16:58 saki bummer
17:01 brainbuz joined #mojo
17:02 ichesnokov joined #mojo
17:04 * sri is not sure how to properly deprecate a github repo https://github.com/kraih/mango
17:10 doby I'd say what you did is pretty effective
17:10 moritz sri: you could also add "deprecated" to the repo description
17:13 * sri feels really shitty
17:24 tempire whoa
17:24 tempire That's quite the decision
17:25 tempire So now we're back to having no non-blocking data store.
17:25 * tempire cries
17:26 * batman gives sri a beer
17:26 batman ????
17:27 * tempire gives sri a baby seal
17:27 tempire ᶘ ᵒᴥᵒᶅ
17:28 tempire Well I forced myself to learn mongodb, but only because of mango.
17:31 d4rkie joined #mojo
17:33 D4RK-PH0ENiX joined #mojo
17:34 sanya_com_ua2 joined #mojo
17:35 dexteruk joined #mojo
17:35 dod joined #mojo
17:35 tempire so does postgres have an http api or not
17:35 tempire I see mentions of it
17:35 tempire but not in the docs
17:36 sri it doesn't, there is only notes on the wiki for a possible future feature
17:42 jamesaxl joined #mojo
18:00 disputin joined #mojo
18:05 irq joined #mojo
18:06 dexteruk_ joined #mojo
18:12 KCL joined #mojo
18:23 Adurah joined #mojo
18:54 * sri has reached the stage where he's kicking himself in the butt for investing so much into not-really-open project
18:55 sri lets see if it's possible to make SQL fun
18:55 irq_ joined #mojo
18:56 irq_ joined #mojo
19:07 Mso150 joined #mojo
19:10 bc547_ sri: i'm curious what you will come up with :-)
19:12 sri you'll prolly be disappointed
19:12 sri the closest to something i like is actually Anorm https://www.playframework.com/documentation/2.4.x/ScalaAnorm
19:13 mst Sequel in ruby is nice
19:17 sri i don't want sql generation
19:17 marcus sri: I'm with you on that.
19:17 marcus I actually like writing sql and very often feel like orms get in the way.
19:18 sri the big problem is to make sql look nice when embedded in perl
19:18 marcus row objects don't map well to the data sets
19:19 marcus sri: LINQ style?
19:20 marcus ORMs also have big problems with async :)
19:20 mst marcus: Sequel is LINQ style, so presumably not
19:20 marcus mst: The way we've been using Sequel it is very similar to DBIC
19:21 mst oh, ok
19:21 mst I've mostly seen LINQ-ish uses of it
19:21 sri this one looks funny too http://hiddentao.github.io/squel/
19:22 marcus squeal would have been funnier.
19:22 sri no btw. not a fan of linq style either
19:22 mst I think what I want is something like
19:23 sugar_ joined #mojo
19:23 mst $db->users->where($Q->name eq 'Joe')->start::all->on_done(sub { <callback here> });
19:23 mst or similar
19:23 mst but I'm sure lots of people won't like that
19:24 sri hehe
19:24 mst I quite like writing APIs that are sync by default, and then using the start:: trick to put them into async mode
19:25 mst means the easy things are still easy without littering the world with ->get or ->await or ->recv calls
19:26 sri of course you can't have that with DBI
19:26 sri just the time you wait for the query to finish is async, receiving records is sync again
19:26 sri according to my benchmarks that's not bad though
19:27 sri as long as the connection to the database is low latency
19:28 mst right, usually once the DB is sending data it can send it as fast as you can read it
19:28 mst imperfect but not as bad as it looks like to begin with
19:28 * sri nods
19:31 moritz sri: does minion need its backend primarily for storage? or rather for communication?
19:31 moritz sri: I guess what I'm really asking is: would a rabbitmq backend make sense?
19:33 marcus zeromq \o/
19:33 moritz well, zeromq and rabbitmq (or amqp in general) are very different beasts
19:34 marcus yes
19:34 marcus One is not horrible
19:35 marcus the other is a message queue
19:35 moritz zeromq is quite horrible when your network topology doesn't allow p2p connections between all relevant hosts
19:36 ua joined #mojo
19:36 marcus moritz: that is true.
19:36 marty joined #mojo
19:36 avenj zeromq isn't a complete solution to much of anything
19:36 avenj it's a set of tools for DIY
19:37 marcus But I don't really like brokers.
19:38 marcus Guess it smells too much enterprise architecture for me :-/
19:38 noganex joined #mojo
19:39 moritz marcus: I'm currently drifting towards enterprise architecture :(
19:39 moritz marcus: not because I like it, but because all the other approaches don't seem to solve our problems
19:40 moritz marcus: it's scary how many buzzwords float around in our dev team lately, and none of us are happy with that
19:40 marcus moritz: I'm sorry. =]
19:41 marcus Seems I'm moving into ops at a large institution soon, so I guess I'll have to learn to like it too.
19:46 marcus I see Lestrrat are doing the official zmq perl bindings
19:48 marty sri: Sorry to hear the new about Mango.  But you have to do what you feel is best.
19:48 basiliscos1 joined #mojo
19:49 marty I wonder what rethinkdb is doing these days
19:50 marty I'll have to convert my entire app to whatever the solution is.  eeeek!
19:54 marty sri: if/when you find something to replace Mango I'll be happy to to donate to development.
19:54 woz joined #mojo
20:02 stephan48 sad to hear that Mango is not continued any more... but on the otherside, yay postgres!
20:13 sri moritz: minion is a storage/transport hybrid
20:14 moritz sri: might make sense to separate the two aspects (but it's also quite some work)
20:14 sri and quite some complexity
20:14 moritz sri: I know that TaskFlow (the thing that openstack uses) does that
20:14 sri just look at how massive celery is (two layer design)
20:15 moritz muhahaha
20:15 moritz celercy looks like a pancake in comparison to taskflow :-)
20:15 sri -.-
20:16 sri my point!
20:16 purl i heard your point was that you can use a non-multipart request
20:16 sri minion is like 500 lines of code
20:17 sri anyway, postgresql with its notify mechanism could replace mongodb pretty well
20:19 sri but i guess you're right, minion as we know it today is dead
20:23 sri interesting how many DBD::Pg users we have on the mailing list
20:24 woz joined #mojo
20:29 chansen sri: for the uninitiated it would be welcomed if you shed some light on why MogoDB's new driver specification sucks. Sometime standards sucks, but a standard that sucks is usually better than no official standard!
20:31 * chansen please replace standard in the last sentence with specification
20:36 mst so far as I can tell, clients have always been expected to handle those things
20:36 mst and they've decided to make all their official clients handle them the same way, rather than 'mostly the same'
20:36 mst so they've published a specification that describes how the official clients are going to be brought in line with each other
20:37 mst I'm still not entirely sure why that's such a bad thing
20:37 hertzhaft joined #mojo
20:39 chansen that was my understanding also of the meaning of the spec and blog post
20:40 Akron joined #mojo
20:41 mst I mean, I think the fact that clients have to do this is kinda fucking dumb, but the bug sri pointed out says that's been an issue since at least 2010
20:43 sri the spec on its own is not bad, the fact that clients are supposed to implement it is
20:50 mst right, but clients have always been supposed to implement something, so far as I can tell
20:50 mst all this does is mean you don't have to reverse engineer what
20:51 mst "we decided to make the drivers we provide less inconsistent, and so we've released the spec we're using so that driver authors have to do the same amount of work they always did, but now have a way to tell if they did it right" doesn't seem like a bad thing
20:51 mst I mean, the 'clients need to implement this' part of true when you started Mango, AFAICR
20:51 mst I'm not sure how them trying to make it a bit easier suddenly makes it bad :)
20:52 marty I was considering Postgres during my development period but the pure Perlness of Mango was just awesome sauce.
20:52 mst I wonder if DBD::PgPP still works
20:53 neyasov joined #mojo
20:55 sri mst: the whole read-preference shit is new
20:56 mst if the existing drivers don't implement it, why do you feel like you'd have to now?
20:56 sri discovering the cluster topology and maintaining a table of roundtrip times... and so on
20:56 mst I mean ... why listen to them instead of your users?
20:56 sri they are implementing it right now
20:57 woz joined #mojo
20:57 mst sure. but the question is: does implementing it help *your* users
20:57 mst if not ... don't
20:57 mst nobody's making you
20:57 sri i won't
20:57 sri i'm out of the game
20:58 sri why should i put up with their bullshit at all?
20:58 sri it makes me unhappy
20:59 sri i was in it purely for fun, it's not fun anymore, so i'm out
20:59 sri mst: if you feel like it's worth keeping Mango maintained, feel free
21:00 mst nah
21:00 mst just wanted to make sure you'd thought it through
21:00 marty Mango was fun regardless of how long it lasts.   No regrets here.
21:00 mst I think your reasoning is faulty, but it's your project, you can be irrational about it if you want to :)
21:00 marty I'm going to check out rethinkdb again, see if anything interesting is happening.
21:01 marty But Mango is solid enough to last me quite a while longer.
21:02 marty Maybe things will change in the future and Mango will LIVE AGAIN!!!   :P
21:03 marty would be kewl, my Mongo DB query zen is actually not to bad now.
21:03 sri after one representative from mongodb got snippy with me in a private mail, i doubt it
21:03 sri now i have a personal grudge
21:04 mst hey, you got over your grudge against me after half a decade or so
21:04 mst it could happen again
21:04 mst in theory ;)
21:04 marty \o/
21:04 marty Kill them with kindness grasshopper
21:04 chansen mst: LOL
21:05 woz joined #mojo
21:05 chansen sri: why not state it's a personal grudge instead of blaming the spec?
21:06 * marty thinks he needs to install convos so he can keep up with IRC
21:06 sri umm, that happened a few minutes ago
21:06 sri and who said it's about the spec in the first place?
21:06 sri i certainly never did
21:06 sri you're putting words in my mouth
21:07 sri https://groups.google.com/forum/#!topic/mojolicious/GFsXFQF3t-k
21:07 mst most of your ranting in here has been about the spec
21:07 sri this has been building up for a long time
21:07 sri the fact that the wire protocol keeps changing without docs is much more annoying
21:08 mst ...
21:08 mst I ... they ... wat
21:08 chansen sri: yesterday you said:
21:08 chansen 05:47:11 <sri> i'm tired of the shit mongodb inc is pulling... instead of fixing things they just push more responsibility on driver authors
21:08 chansen 05:47:24 <sri> this is where i draw the line http://emptysqua.re/blog/server-discovery-and-monitoring-spec/
21:08 marty wow, I didn't know the wire protocol was changing without docs.  That would most definitely suck.
21:10 chansen marty: protocols change! usually with backwards compatibility.
21:11 sri anyway, discussion is over now
21:11 marty I don't think it is the changing that is the problem.  It is not being documented that would be the issue.
21:11 marty Good news is.  Right now, Mango still rocks!!!
21:12 mst Bad news is. Right now, you're still using MongoDB :(
21:12 marty :(
21:13 sri for the rest of the day only happy discussions
21:13 disputin joined #mojo
21:15 chansen marty: xdg (David Golden) is leading the development of the "official perl" version, I'm sure he is interested in a asynchronous API compatible with Mojo
21:16 mst right, I've already been talking to him about options to provide maximally useful async support
21:35 disputin joined #mojo
21:36 woz joined #mojo
21:37 jberger_ sri++ # deprecation full disclosure
21:37 jberger_ Also maybe 10gen should try this : http://tools.ietf.org/html/rfc6919
21:43 chansen jberger_: I don't think any serious developer or company invested time in a API would be amused with the RFC 6919 April fool joke
21:50 mst jberger_: 'the passive aggressive voice' is an amazing concept
21:58 mgrimes joined #mojo
22:00 jberger_ chansen: imagine that ;-)
22:04 chansen jberger_: truthfully, I'm not sure how to respond a comment like that :/
22:15 cosimo joined #mojo
22:16 sugar_ joined #mojo
22:19 meshl joined #mojo
22:27 woz joined #mojo
22:28 * chansen away
22:31 ribasushi joined #mojo
22:45 sri oh my, that poll thread exploded :O
22:48 firnsy yes it did :)
22:50 disputin joined #mojo
22:56 * sri didn't expect tempire to be a mysql user
23:08 neyasov joined #mojo
23:09 neyasov_ joined #mojo
23:10 neyasov__ joined #mojo
23:12 neyasov___ joined #mojo
23:13 neyasov____ joined #mojo
23:13 basiliscos joined #mojo
23:14 neyasov_____ joined #mojo
23:15 neyasov______ joined #mojo
23:19 disputin joined #mojo
23:26 * sri wonders what a migration system integrated into mojolicious would look like
23:27 woz joined #mojo
23:38 neyasov______ joined #mojo
23:39 D4RK-PH0ENiX joined #mojo
23:47 lestrrat marcus: I'm done with zmq
23:48 lestrrat I just don't see using zmq in any of my projects in the future, so you might want to use the other newer versions, like the ones that use FFI
23:58 neyasov______ joined #mojo

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