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

IRC log for #mojo, 2014-06-21

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

All times shown according to UTC.

Time Nick Message
00:18 d4rkie joined #mojo
00:25 nicomen sri: isn't it ok since the keys change on every fetch?
00:28 sri fetching random numbers for crypto from a server, what could possibly go wrong
00:32 tempire ok
00:32 tempire It's time for a game
00:32 tempire called
00:32 tempire "Questions tempire doesn't know the answer to"
00:32 sri where are my keys?
00:33 tempire that's not part of the game!
00:33 tempire They're where you left them.
00:33 sri :(
00:34 tempire https://gist.github.com/tempire/ae63c7692b4f6603433f
00:35 tempire The question is, why doesn't $tx->res->body return an error
00:35 howitdo joined #mojo
00:35 tempire because I would expect it to not be a real transaction object.
00:36 sri syntax error
00:36 purl YOU CAN'T JUST MAKE SHIT UP AND EXPECT THE COMPUTER TO MAGICALLY KNOW WHAT YOU MEAN, RETARDO.
00:36 * sri hugs purl
00:36 * purl nibbles sri's elbow
00:39 tempire oh
00:39 tempire paste fail
00:39 tempire paste fix!
00:39 tempire oh
00:40 tempire nevermind
00:40 purl Well piss off then, tempire
00:40 tempire I do know the answer.
00:40 tempire it's a real object, just has the error as entered by the tx finish event
00:40 nicomen didn't you teach about this?
00:40 tempire I told them what happened. Just not why.
00:41 howitdo joined #mojo
00:41 nicomen heh ok
00:42 sri error handling for transactions is actually still an open problem, what we have now just sucked the least
00:43 sri (do not just throw around suggestions, it's a hard problem, look deeper if you want to really give it a try)
00:46 nicomen tempire: where exactly do you want an error?
00:47 nicomen or, why would you not expect it to be a real tx object?
00:47 tempire I don't necessarily have a problem with it.
00:48 nicomen that's ok, but I'm wondering why, because I would expect it to be
00:48 tempire I originally expected $tx to not be there since $ua went out of scope
00:52 nicomen then you would expect the sub not to be executed either or?
00:54 tempire my expectation didn't go that far.
01:19 klapperl_ joined #mojo
01:27 alnewkirk joined #mojo
03:12 basic6 joined #mojo
03:14 irq joined #mojo
03:28 btyler joined #mojo
03:31 sri i've changed how delayed jobs work in minion, think it makes more sense now https://github.com/kraih/minion/commit/ad001eb9354b54091472277f128806374e371da7
03:31 sri you say {delay => 20} for a 20 second delay, and that applies also when you restart the job
03:34 sri on the other hand... if you restart a failed job you don't want the delay again i suppose :o
03:34 sri hmmmm
03:44 jberger cpan_mojo is seeming to keep very strange hours lately 😫
03:44 sri lazy bots
03:44 purl Simple little hoo-mon!
03:45 sri :o
03:45 jberger Except for mowing the lawn, i have the entire day tomorrow free!
03:46 * jberger is ridiculously excited
03:47 jberger For cpan_mojo i wonder if there is some rate limit ban that it trips on when multiple releases happen at the same instant
03:47 sri \o\
03:47 sri /o/
03:48 jberger I didn't put any rate limit in because there is a relatively low rate of cpan submission
03:49 jberger But i wonder if MagNet bans instantaneous sequential posts
03:49 sri wouldn't surprise me if the irc servers had rate limits
03:51 jberger I can just make a simple message queue and a one second recurring which sends one message per exec
03:51 sri i guess restarting jobs with delays is silly... you'd generally just enqueue a new job from within the old job
03:53 jberger I do have ForkCall test improvements waiting to be released to cpan
03:54 jberger So I'll do that tomorrow
03:55 jberger sri re restarting, i can see the use case both ways
03:56 jberger Which doesn't help you much i guess
04:15 sri was just wondering if i should rename ->restart to ->retry and remove the ability to restart finished jobs
04:15 sri but being able to restart finished jobs in addition to failed jobs seems rather useful for debugging
04:16 jberger hmmmm, that's a tough one
04:16 preaction retry sounds to me as though it would do both
04:16 sri "oh, you did not get your email confirmation? our minion says it went out, but let me just restart this job..."
04:16 jberger ^^ nice
04:17 jberger anyway, my contribution for the day: http://www.reddit.com/r/perl/comments/28kt0s/world_cup_results_in_perl/cid2v9q
04:17 jberger preaction, I even added something in there for you :-)
04:17 preaction heh
04:17 sri :)
04:18 sri can you believe costa rica beat italy?
04:19 jberger knowing nothing about soccer **ahem** football, still, the answer to that is: no
04:19 jberger been a bad go for Europe for the most part, eh?
04:20 sri germany is very strong
04:22 jberger I did say "for the most part"
04:22 preaction deutscheland über alles!
04:23 jberger The Germany Portugal game was the only one i have actually watched all the way through
04:23 sri yea, we don't say that
04:24 preaction ah. third verse only. I learned something new today
04:26 jberger Does the US have a chance?
04:26 preaction the US coach already said no
04:27 sri well, you've got klinsmann
04:27 preaction apparently people got annoyed at the coach's frank appraisal of the situation
04:30 jberger Sounds like a stand-up guy
04:31 preaction we americans do love boundless, irrational exhuberance
04:33 jberger Until we lose and then we remember that we don't like this game anyway
04:34 jberger 😋
04:34 sri lets jus say the us team looks better than usual
04:34 preaction eh, in all things. the chicago cubs have not won the series in more than a century, and yet still the fans show up and cheer and talk about how much they'll win this season
04:35 jberger 107 years this year
04:36 jberger If we had relegation the Cubs would be playing in a parking lot next to a 7-11
04:36 jberger But we still love them
04:36 sri lol
04:37 jberger Hey, they almost got close in 2003
04:38 jberger That still counts for....
04:38 * jberger sobs
04:39 preaction was that right before or right after the white sox broke their century-long spell?
04:39 jberger Right before
04:40 preaction 88 years (1917-2005)
04:40 jberger If the cubs had won, all three 80+year droughts would have ended in a row
04:41 preaction that would have been amazing, but they're the cubs
04:41 jberger Cubs, red sox, white sox
04:41 jberger But yeah, cubs
04:41 jberger Ok well now I'm going to go cry myself to sleep
04:43 jberger Tomorrow = much mojo!
04:43 preaction yeah baby
04:43 purl get preaction's mojo on
04:43 preaction purl++
05:16 sujithm joined #mojo
05:37 disputin joined #mojo
05:45 sujithm joined #mojo
05:53 Kundun joined #mojo
06:22 Averna joined #mojo
06:56 sujithm joined #mojo
06:57 Vandal joined #mojo
07:00 Eke- joined #mojo
07:22 sh4 joined #mojo
07:27 sujithm joined #mojo
07:31 mire joined #mojo
07:31 kyshtynbai joined #mojo
07:38 rem_lex| joined #mojo
08:01 sh4 joined #mojo
08:27 dexteruk joined #mojo
08:39 arthas joined #mojo
08:45 denis_boyun joined #mojo
09:33 neyasov__ joined #mojo
09:41 arthas_ joined #mojo
09:42 basiliscos joined #mojo
09:53 abra joined #mojo
09:55 sujithm joined #mojo
09:57 dexteruk joined #mojo
10:29 trone joined #mojo
10:44 Eke- joined #mojo
11:04 d4rkie joined #mojo
11:17 meshl joined #mojo
11:23 punter joined #mojo
11:34 abra_ joined #mojo
11:36 d4rkie joined #mojo
11:40 inokenty joined #mojo
11:43 irq joined #mojo
11:52 rem_lex|pivo joined #mojo
11:56 sujithm joined #mojo
11:58 meshl joined #mojo
11:58 basic6_ joined #mojo
12:19 Eke- joined #mojo
12:36 Gedge joined #mojo
12:47 meshl joined #mojo
12:51 meshl joined #mojo
13:25 mire joined #mojo
13:29 nikita_ joined #mojo
13:32 nikita_ @marcus hello, how would you advice implementing master failover with redis sentinel in an application that uses mojo::redis?
13:35 nikita_ @marcus i mean an application must find out what current master is somehow
13:37 nikita_ @marcus i'm trying to do that just by maintaining a separate connection to sentinel and watching for +master +slave and +promoted-slave events
13:38 nikita_ @marcus and modifying env variables appropriately
13:41 nikita_ @marcus like this https://gist.github.com/nikita-d/bfe544fa9c98a22695e8
13:47 zivester joined #mojo
13:55 shmuel joined #mojo
14:34 denis_boyun joined #mojo
14:43 crab what if the sentinel dies?
14:45 nikita_ @crab i think i should get all sentinels and subscribe to all of them
14:45 nikita_ @crab +switch-master fires when a new master already elected
14:46 nikita_ @crab and also check if newly received master is not the same received previously
14:53 nikita_ @crab also theres +sentinel event
14:55 nikita_ @crab and it works well, i am able to keep session alive master was switched
14:55 nikita_ *when master was switched
14:57 neyasov__ joined #mojo
15:10 tianon joined #mojo
15:14 meshl joined #mojo
15:15 crab sounds good
15:16 ivan joined #mojo
15:16 basiliscos joined #mojo
15:19 tempire joined #mojo
15:19 mire joined #mojo
15:36 dexteruk joined #mojo
15:56 dexteruk joined #mojo
15:58 irq joined #mojo
15:59 disputin joined #mojo
16:00 meshl joined #mojo
16:10 Eke- joined #mojo
16:33 sri o/
16:33 sri \o
16:40 mire joined #mojo
16:45 tianon joined #mojo
16:50 btyler joined #mojo
17:11 thowe joined #mojo
17:12 Eke- joined #mojo
17:26 denis_boyun joined #mojo
17:33 jberger joined #mojo
17:33 jberger o/ from Convos 0.6
17:48 cpan_mojo joined #mojo
17:53 jberger sri: how does this look for rate limiting messages to irc? https://github.com/jberger/Mojo-FriendFeed/blob/queue/ex/cpan.pl#L103-L105
17:54 dexteruk joined #mojo
18:11 sri jberger: i guess it works
18:11 jberger OHohohOHOHohH
18:11 dexteruk joined #mojo
18:12 jberger new Mojo::Collection method idea
18:12 sri ohohohohoh
18:12 jberger sub Mojo::Collection::tabilify { Mojo::ByteStream->new(Mojo::Util::tablify $_[0]) };
18:13 sri heh, i'm not such a big fan of tablify ;p
18:13 sri one of those "this sucks the least" type of choices
18:14 jberger ah, but look how cool this is: http://pastie.org/9311982
18:14 sri but that reminds me... you also need to finish the $c/$self and Minion::Backend::DBMDeep patches :o
18:14 jberger :o
18:15 jberger $c/$self was just a question, but I take it from your comments that its something you would like?
18:15 sri s!patches!patches/modules!
18:15 sri jberger: i need to see it, think it's something that needs to be decided by vote
18:15 * jberger would have liked to see the leaning toothpicks version of ^^
18:16 sri which also reminds me... marcus wanted to work on an FAQ answer for module namespaces
18:17 jberger that is probably something worth doing, do we have a good answer for it
18:17 jberger ?
18:17 sri and i'm sure tempire wanted to work on something as well... but i don't remember atm :o
18:17 mire joined #mojo
18:17 sri jberger: i don't have a good answer, which is why i was happy marcus wanted to do it :)
18:18 jberger hehe
18:18 jberger I guess the easy thing would be to un-deprecate the MojoX namespace and encourage all non-cutesy non-specific modules to use it
18:18 sri re tablify, i think the fact that it only works on an array of arrays kinda kills it
18:19 jberger and by non-specific I mean, I think that Mojo::FriendFeed doesn't have much potential to step on core toes
18:19 sri i'm happy with cutesy names like Minion standing on their own
18:20 sri would never rename it
18:20 jberger no, I'm not suggesting that, quite the opposite
18:20 jberger I'm saying, if you want a cutesy name, go ahead,
18:20 sri oh, you said *non*-cutesy
18:21 jberger if I wanted to make something called Mojo::CacheParams or something, that probably is best left to a MojoX namespace
18:22 sri i suppose so
18:22 * sri is a little worried about people stealing Mojo::Transaction::HTTP2 and the like
18:22 jberger yeah, that should really be discouraged
18:22 dexteruk joined #mojo
18:27 jberger batman/marcus, why is cpan only providing Convos 0.6
18:27 jberger it seems to have 0.7 available but not listed as latest
18:33 meshl joined #mojo
18:34 dexteruk joined #mojo
18:50 mire joined #mojo
18:51 lb i think batman must've confused pause
18:52 lb i've seen it before, but can't remember what caused it, besides uploader not being a maintainer
18:52 lb i'm sure mst knows what it is
19:22 mire joined #mojo
19:58 sri this stuff makes me feel very very uncomfortable about relying on EV http://code.activestate.com/lists/perl5-porters/209005/
19:59 avenj he is pretty poisonous, definitely a shame
20:12 basiliscos joined #mojo
20:17 cpan_mojo Mojolicious-Plugin-I18N 1.3 by Anatoly Sharifulin - http://metacpan.org/release/SHARIFULN/Mojolicious-Plugin-I18N-1.3
20:26 sh4|2 joined #mojo
20:39 mire joined #mojo
20:40 batman I'm on vacation. Sort of. Back tomorrow :)
20:49 dexteruk joined #mojo
20:53 Ptolemarch joined #mojo
20:59 jberger batman: have fun!
21:16 Ptolemarch joined #mojo
21:50 hesperaux joined #mojo
22:04 Ptolemarch joined #mojo
22:12 mire joined #mojo
22:34 tempire What are the other options?
22:34 purl the other options are KFC
22:34 tempire libuv?
22:34 purl libuv is probably bsd/mit from joyent ...
22:38 sri not yet
22:38 sri don't think there are other options currently
22:39 sri performance is only one problem, the other is sharing a backend with multiple event loops (AnyEvent, IO::Async...)
23:07 hesperaux_ joined #mojo
23:15 meshl joined #mojo

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