Camelia, the Perl 6 bug

IRC log for #mojo, 2013-06-13

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

All times shown according to UTC.

Time Nick Message
00:03 btyler joined #mojo
00:21 mrphilov Thanks again for Mojo. Really well-thought-out framework.
00:22 mrphilov Was just looking at a template source and saw something I'd done and said to myself: "Whoah. That is the sort of weird thing you'd expect to work, but a lot of frameworks would probably choke on." Mojo handled just fine.
00:25 btyler joined #mojo
00:36 jberger <3 Mojo::Template
00:37 sri \o/
00:38 jberger ūüé©
00:39 whitebook joined #mojo
01:11 moltar joined #mojo
01:22 whitebook joined #mojo
01:25 Adurah joined #mojo
01:26 bc547_ joined #mojo
01:27 plu_ joined #mojo
01:27 mattp_ joined #mojo
01:27 xdbr joined #mojo
01:27 dabudabu_ joined #mojo
01:27 moritz_ joined #mojo
01:27 Phil21_ joined #mojo
01:28 gabiruh- joined #mojo
01:28 tardisx` joined #mojo
01:28 _mkrull joined #mojo
01:28 xxtjaxx_ joined #mojo
01:28 coff_ joined #mojo
01:28 Kov|sumika joined #mojo
01:28 charsbar___ joined #mojo
01:28 rafl_ joined #mojo
01:29 buu joined #mojo
01:29 preflex joined #mojo
01:29 Andreas joined #mojo
01:29 jeff joined #mojo
01:29 zivester joined #mojo
01:29 robins joined #mojo
01:29 andrefs joined #mojo
01:29 Averna joined #mojo
01:29 gtodd joined #mojo
01:29 tempire joined #mojo
01:29 Caelum joined #mojo
01:30 preflex joined #mojo
01:31 Bender joined #mojo
01:32 mst joined #mojo
01:32 mfontani joined #mojo
01:33 garu joined #mojo
01:40 tempire ūüé©
01:42 gtodd joined #mojo
01:59 d4rkie joined #mojo
02:10 whitebook joined #mojo
02:25 tba joined #mojo
02:50 good_news_everyone joined #mojo
02:50 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/DHC6Cw
02:50 good_news_everyone mojo/master a265d7f Sebastian Riedel: added Mojo::JSON tests for empty keys
02:50 good_news_everyone left #mojo
02:50 inokenty joined #mojo
03:03 btyler joined #mojo
03:04 sri hahaha... good old windows http://www.cpantesters.org/cpan/report‚Äč/16bf04da-6bf9-1014-8755-943d3130705e
03:05 tba joined #mojo
03:06 _inokenty joined #mojo
03:11 sri it seems that test fails quite a bit on widnows though, would be nice if someone could investigate
03:21 tba joined #mojo
03:23 phillipadsmith joined #mojo
03:40 btyler joined #mojo
03:48 tba joined #mojo
03:56 tba joined #mojo
03:59 preflex_ joined #mojo
04:04 tba joined #mojo
04:21 btyler joined #mojo
04:23 rem_lex|pivo joined #mojo
04:24 good_news_everyone joined #mojo
04:24 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/gg34CA
04:24 good_news_everyone mojo/master 2ad071e Sebastian Riedel: fixed typos in test descriptions
04:24 good_news_everyone left #mojo
04:52 delias joined #mojo
04:54 amirite joined #mojo
05:07 gtodd joined #mojo
05:09 tomi-ru joined #mojo
05:41 nicomen joined #mojo
05:44 basiliscos joined #mojo
05:49 Britzel_ joined #mojo
06:24 dpetrov_ joined #mojo
06:24 tba joined #mojo
06:31 Mike-PerlRecruiter_ joined #mojo
06:31 suy joined #mojo
06:43 batman phillipadsmith: btw...do you know what 400 means? i often use http://en.wikipedia.org/wik‚Äči/List_of_HTTP_status_codes when i forget the actual meaning...
06:44 batman the server should have responded with 401/403 imo though...
07:04 sri wow, elixir already has an oreilly book too :o
07:06 dod joined #mojo
07:24 denisboyun_ joined #mojo
07:26 dod joined #mojo
07:31 trone joined #mojo
07:36 maxhq joined #mojo
07:43 Vandal joined #mojo
07:43 sri i guess the upside of the whole dots pragma thing is that the trolls have something else than signatures to be angry about, maybe it can still serve as a smoke screen
07:51 Vandal why trolls don't like dots pragma?
07:58 jast who knows
08:06 jzawodn joined #mojo
08:13 fhelmber_ joined #mojo
08:31 SmokeMac_ joined #mojo
08:47 SmokeMachine joined #mojo
08:54 abstract joined #mojo
09:02 tba joined #mojo
09:20 dotan joined #mojo
10:17 rem_lex| joined #mojo
10:19 rem_lex joined #mojo
10:41 nic The only thing I read about the dots pragma was Ricardo pointing out it creates surprising bugs
11:04 tba joined #mojo
11:16 Kulag joined #mojo
11:16 TitanOfOld joined #mojo
11:19 cooper joined #mojo
11:38 libsysguy joined #mojo
11:47 denisboyun_ joined #mojo
12:03 btyler joined #mojo
12:07 d4rkie joined #mojo
12:08 whitebook joined #mojo
12:28 mire joined #mojo
12:31 ryozi joined #mojo
12:51 amirite joined #mojo
13:11 moltar joined #mojo
13:11 whitebook joined #mojo
13:12 dotan joined #mojo
13:31 inokenty joined #mojo
13:33 phillipadsmith batman: thanks. :)
13:39 btyler joined #mojo
13:41 btyler_ joined #mojo
13:46 bluescreen joined #mojo
13:59 whitebook joined #mojo
14:16 marty o/
14:21 libsysguy \0
14:22 r0b3rt strawberry perl 5.16.3 / Mojolicious 4.14 / Windows Server 2008 - all tests passing for me
14:23 jberger IIRC that's been the pattern with this failing windows test, its a classic heisenbug
14:24 depesz joined #mojo
14:24 sh4 joined #mojo
14:33 whitebook joined #mojo
14:37 nic One of the problems with perl on windows is people trying to use it seriously over cygwin
14:37 * nic <3 teh strawberries
14:37 whitebook joined #mojo
14:41 jberger <3 teh penguins
14:42 whitebook joined #mojo
14:42 jberger also my irc client worked really hard to fix my 'teh'
14:42 jberger :-P
14:43 libsysguy there should be a / flag to turn off spell checking
14:43 libsysguy like /s teh awesomez
14:43 jberger hehe
14:46 whitebook1 joined #mojo
14:51 nic I just get my spellchecker to auto-install-to-dictionary any words it thinks are wrong
14:51 nic means the spellchecker works twice as hard
14:51 nic but it's worth it
15:33 fhelmbe__ joined #mojo
15:37 btyler joined #mojo
15:37 btyler_ joined #mojo
15:48 whitebook joined #mojo
16:23 sri nic: we have more reports for failing tests on cygwin than on strawberry
16:23 sri r0b3rt: from what i hear it only fails randomly
16:27 nic sri: exactly my point
16:40 basiliscos joined #mojo
16:45 libsysguy sri, in mango, how does one use the $gt or $lt operations?
16:45 libsysguy I tried { column => { lt => $value}}
16:45 libsysguy and it didn't seem to work
16:49 mikegrb joined #mojo
16:49 tianon libsysguy: { column => { '$lt' => $value } }
16:50 libsysguy tianon: that is what the perl tutorial said as well, but perl picks up $lt as a variable
16:50 tianon because otherwise you're looking for a field whose value is literally the document { lt => $value }
16:50 tianon note the single quotes
16:50 tianon very important
16:51 libsysguy ah
16:51 tianon double quotes or no quotes will interpolate ;)
16:51 libsysguy I bet I was using double
16:51 libsysguy newb fail on my part
16:51 tianon :)
16:52 libsysguy hey that worked, tianon++
16:52 tianon lol
16:52 tianon been using Mango since the day sri first posted on Twitter that he had created it <3  the "official" driver is pretty much just garbage
16:53 tianon been using that "official" one much longer, unfortunately
16:54 libsysguy sri keeps saying its not prod ready but it has been pretty stable thus far
16:54 sri IT'S NOT PRODUCTION READY!
16:54 libsysguy ahh
16:54 * libsysguy ducks
16:55 sri i keep saying it because i didn't say it during mojolicious development... and that hurt our reputation quite a bit during the first year
16:55 mikegrb joined #mojo
16:55 libsysguy okay I won't tell anybody its production ready if I can keep using it (not in production :p)
16:58 mire_ joined #mojo
16:58 batman is it possible to use Server::Daemon as an "anything" tcp daemon?
17:05 batman maybe i can replace the "req" object inside after_build_tx
17:08 basiliscos joined #mojo
17:11 dotan joined #mojo
17:12 libsysguy tianon whats your strategy for storing datetimes in mongo?
17:12 libsysguy it looks like there are a few approaches you can take
17:13 jargan joined #mojo
17:14 tianon we decided to go with the "Mongo" approach and store them as actual Date() objects (which with the MongoDB driver takes a hit in performance since it converts to DateTime objects automatically, but is worthwhile when viewing through the mongo console)
17:14 libsysguy it seems that Mango::BSON converted all my dates to Timestamp objects
17:14 tianon we played with using unix timestamps (and still have several products that do that), but just didn't like manipulating them and having to look up an epoch converter every time we needed a quick "what date did X happen" piece of info directly from the database
17:14 tianon to Mango::BSON-specific timestamp objects :)
17:15 tianon very lightweight # sri++
17:17 tianon but Mango really is "unstable" and sometime in the future sri is going to finally commit his core rewrites to solidify that in our minds again :)
17:17 * libsysguy passes sri cookies as a bribe
17:31 sri interesting to see the tokudb folks actually advertising their modifications in the mongodb channel
17:41 sri the $lt question reminds me that i wanted to add -lt support as an alternative
17:42 dod joined #mojo
17:50 batman marcus: is this something you would use: https://github.com/jhthorsen/mojo-apns # referring to the "feedback" part
17:51 bpmedley joined #mojo
17:51 libsysguy yayy my newb question generated a silent feature request
17:51 batman will i always receive a "close" after/before i receive "timeout" events on a stream object?
17:52 duncanthrax joined #mojo
17:53 * sri wonders if an operator shortcut is really a good idea or if it might be too confusing
17:54 libsysguy well if for nothing else it would be documented and prevent someone like me from embarrassing themselves with a stupid quote question :p
17:55 sri that's not a very good reason
17:57 libsysguy well I didn't say it was *good*
17:57 libsysguy :p
17:57 marty sri: shortcuts for all '$' operators or just $lt?
17:58 sri all $...
17:58 libsysguy code readability maybe
17:58 libsysguy I feel like DBIx::Class uses the - replacement
17:58 sri the patch is basically "$name =~ s/^-/\$/;" :)
17:59 sri the point of - is that it doesn't have to be quoted
18:00 sri {foo => {-lt => 23}}
18:00 marty Hmmm.  interesting.  yes, it would decrease typing and might increase readability.  Especially with some syntax highlighters.
18:00 denisboyun joined #mojo
18:03 marty Yep, the sublime text 2 highlighter works pretty good with it.
18:22 gtodd is making the time calculation example on the mojolicio.us home page (the bit in @@ clock.html.ep   that uses: % my $now = localtime; ) be "live" (updated every second like a real clock)  a job for javascript or a job for perl/mojo ?
18:22 sri WHY NOT BOTH!?
18:23 libsysguy http://1-media-cdn.foolz.us/ffuuka/boa‚Äčrd/vp/image/1364/61/1364614434169.jpg
18:25 delias joined #mojo
18:27 gtodd a little func and sub :)
18:33 Mike-PerlRecruiter_ joined #mojo
18:57 abstract joined #mojo
19:01 basic6 joined #mojo
19:05 trone joined #mojo
19:14 marty weeeee... ¬†http://nordaaker.com/docs/skitch‚Äč/Mojolicious-20101130-230417.png
19:16 jberger the wrath of lucas will rain upon marty
19:17 jast s/lucas/disney/
19:17 marty actually, it was created by Jedi master marcus.
19:21 * jberger doesn't suspect that disney will end lucas' reign of terror
19:21 * jberger wishes marty and marcus best of luck :-P
19:36 sri \o\
19:36 sri /o/
19:36 marty |o|
19:37 jpn joined #mojo
19:37 * sri hears the tie fighter sound in his head
19:39 libsysguy \o/   ^o^   o[    /o\
19:40 sri it did sound better in my head :o http://www.youtube.com/watch?v=S5_svSADYks
19:56 jberger libsysguy: hahahaha
19:57 jberger sri: having just read about Sub::Infix your dancing guys take on a new meaning: https://metacpan.org/module/Sub::Infix
20:05 tba joined #mojo
20:09 sri btw. what worries me a bit about {foo => {-lt => 23}} is that '-whatever' wouldn't be a valid key anymore
20:09 sri and i'd rather avoid adding config options for serialization
20:10 sri it's already a rich format, no need to make it even more complicated
20:10 * sri shrugs
20:12 mattastrophe joined #mojo
20:12 marty agreed.  I was thinking of whether or not I would use it and I don't think I would if for no other reason than to keep constant with mongodb syntax.
20:12 * sri nods
20:13 jberger sri: that would be my same concern. also atm mango can basically just say: read the mongodb docs
20:13 sri that's a good point too
20:13 * jberger ninja-ed sri's praise :-)
20:13 marty jberger++   sneaky
20:13 sri ALL GLORY TO THE HYPNOTOAD!
20:14 marty what are you talking about sri, you nee... ALL GLORY TO THE HYPNOTOAD!
20:14 jberger marty, why can't you just ... ALL GLORY TO THE HYPNOTOAD!
20:15 marty http://buffalobeast.com/wp-conte‚Äčnt/uploads/2012/07/hypnotoad.gif
20:15 marty my appologies. I did not think before posting that
20:16 marty animated gifs in a channel are probably worse than blinking html text
20:17 marty afk
20:17 denisboyun joined #mojo
20:18 sri you should r... ALL GLORY TO THE HYPNOTOAD!
20:23 perlite joined #mojo
20:24 sri hmmm
20:24 * sri wonders how portable the inf/nan handling in Mango::BSON actually us
20:24 sri s/u/i/
20:24 sri they get packed as doubles
20:24 sri and somehow roundtrip ok
20:26 sri perl -MMango::BSON=bson_encode,bson_decode -MTest::More -E 'is_deeply bson_decode(bson_encode {test => [9**9**9]}), {test => [9**9**9]} ;done_testing()'
20:27 * sri wonders if that's actually portable
20:27 jberger works here: linux 32bit (with use64bitint)
20:28 sri :o
20:28 * jberger could install strawberry quickly enough
20:28 jberger on a virtual windows
20:29 sri that would be nice
20:29 jberger I needed that vbox windows for watching #yapc last week
20:29 * jberger changes computer
20:30 jberger_ joined #mojo
20:30 jberger_ o/
20:30 sri \o
20:32 jberger_ hmmmm, does strawberry build with use64bitint?
20:32 jberger_ guess we'll find out
20:33 abstract joined #mojo
20:34 sri i've committed the tests to github for now https://github.com/kraih/mango/commit/6‚Äč9da3dca858a6b2312983f9a7657eb261034d6dd
20:34 sri so you can just prove -l t/bson.t
20:35 jberger_ k
20:35 sri but considering that doubles are IEEE754 floats... i think nan and inf should actually be portable
20:35 jberger_ of course that would mean having git4win or whatever the heck that is
20:36 jberger_ oh or I think I can get a zipball right?
20:36 * jberger_ checks
20:36 TitanOfOld joined #mojo
20:36 sri http://latest.mojolicio.us
20:36 sri :)
20:36 jberger_ not for mango?
20:36 sri oops
20:36 sri right ;p
20:37 sri there's a zip button
20:38 jberger_ interesting: windows.github.com
20:38 jberger_ ok, maybe not interesting. possibly sufficient >
20:38 jberger_ ?
20:38 sri there's also http://mac.github.com
20:38 jberger_ I've been waiting for real libgit2 bindings for Perl
20:39 * sri has not used those clients yet though
20:40 * jberger_ probably should have given the vbox more memory :-)
20:40 jberger_ kinda SLOW
20:41 jberger_ fyi: strawberry 5.18 for 32 bit has use64bitint
20:42 tianon jberger_: might also want to test against the windows builds of MongoDB, just in case
20:42 tianon while you're there testing :P
20:42 jberger_ ok bson tests pass
20:43 jberger_ I don't have mongodb installed
20:43 jberger_ I'm happy to do it
20:43 tianon that's what I'm suggesting would probably also be a prudent test, in case they do something funky on their compile for the server-side storage/bson-parsing :P
20:43 jberger_ Mango does all its own bson :-)
20:43 sri \o/
20:43 sri jberger++
20:44 sri tianon: bson has a rather strict spec
20:44 sri they can't mess with endianess or the like
20:44 jberger_ did mongo define bson or did bson predate mongo?
20:45 sri http://bsonspec.org/#/specification
20:45 sri 10gen did make the spec too
20:45 tianon sri: indeed, but might as well rule it out now, right?  it wouldn't be the first time their windows builds had odd issues :P
20:45 sri it has a few very mongodb specific types
20:45 tianon ObjectId ftw :)
20:45 tianon and the wtf of their internal Timestamps :)
20:45 sri tianon: not like i would add workarounds for such screwups anyway ;p
20:46 tianon sri: haha, too true :)
20:47 sri i've not had much trouble with the wire protocol, some things are a bit weird, but it's all manageable
20:48 * jberger_ types ls rather than dir at EVERY level :-/
20:49 sri BSON in general is a bit odd, you notice quickly it's made for use inside a database
20:49 sri values have length prefixes that really don't need to
20:50 sri small optimizations to jump around big chunks of data faster
20:52 jnbek joined #mojo
20:52 sri strings end with a null byte *and* have a length prefix ;)
20:56 jberger_ how the blazes do you set env vars in windows??
21:01 jberger_ hmmmm mango non-bson tests not passing on windows
21:09 jberger_ http://pastie.org/8040453
21:13 mattastrophe joined #mojo
21:15 Akron joined #mojo
21:21 jberger_ sri: if you want me to do anything else with the windows instance, let me know
21:21 jberger_ I'm shutting it down for now otherwise
21:24 sinkovsky joined #mojo
21:30 sri jberger: doesn't have to be a windows bug, there is a known bug preventing mango from working with some mongodb setups
21:34 lukep joined #mojo
21:34 SmokeMac_ joined #mojo
21:39 jberger understood
21:39 jberger this was 2.4.4 btw
21:40 jberger and I could see from the mongod window that it was connecting etc
21:55 sri jberger: i think it might be authentication going haywire or so
21:55 sri jberger++
21:58 jberger I can tinker with it more tomorrow
21:58 * jberger is leaving for chicago.pm soon
21:58 sri have fun
21:59 jberger I will
21:59 jberger though sadly project nights aren't very well attended :-/
21:59 sri i wouldn't complain if you fix the bug btw ;)
22:00 jberger if I had the vbox on my netbook I would bring it with as a project, but I don't think I could manage that in the time I have
22:00 jberger something for tomorrow :-)
22:52 KindOne joined #mojo
23:10 mire__ joined #mojo
23:10 btyler joined #mojo
23:10 btyler_ joined #mojo
23:13 marty I'd rather work with the mongodb wire protocol over the rethink protocol.  That spec was scary.   :)
23:19 ryozi joined #mojo
23:42 btyler joined #mojo
23:48 sri my favorite is still the redis protocol
23:51 mspo joined #mojo
23:51 mspo I send this into the web frameworks benchmark list, is it accurate? https://groups.google.com/forum/?fromgroup‚Äčs#!topic/framework-benchmarks/_NfY7bF18NA
23:52 sri no
23:52 sri afaik those benchmarks are using psgi and starman or so
23:53 sri not even a good psgi setup
23:54 sri uWSGI is the best from what i hear
23:54 sri (in case you really care, because i don't)
23:55 mspo plackup -E production -s Starman --workers=" + str(args.max_threads) + " -l " + home + "/FrameworkBenchmarks/mojolici‚Äčous/frameworks-benchmark.sock -a ./app.pl", shell=True, cwd="mojolicious"
23:56 sri hahaha
23:56 sri so they ran mojolicious 3.x in development mode
23:57 sri glorious
23:57 mspo now is the time to say "please run like this instead"
23:57 sri knock yourself out
23:58 mspo well I would say "use mod perl" ;)

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