Camelia, the Perl 6 bug

IRC log for #mojo, 2013-07-22

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

All times shown according to UTC.

Time Nick Message
00:38 shmuel joined #mojo
01:29 KindTwo joined #mojo
01:56 Meiermann joined #mojo
01:59 d4rkie joined #mojo
02:01 abra joined #mojo
03:19 basic6_ joined #mojo
03:30 zivester joined #mojo
05:08 yakudza joined #mojo
05:40 Britzel joined #mojo
05:46 asarch joined #mojo
05:56 Meiermann joined #mojo
06:08 dod joined #mojo
06:12 dpetrov_ joined #mojo
06:29 Mike-PerlRecruiter_ joined #mojo
06:33 hrupp_ https://www.evernote.com/shard/s22/sh/dbb6​5955-4afb-4619-b782-6423d0548c08/7b6041f12​123e4ec29d995ccfa9dc554/deep/0/wirc.pl%20-​%20Chat%20and%201.%20wirc.pl:0.1.0%20su%20​wirc.pl%20(ssh)%20and%20%C2%B5Torrent.png
06:33 hrupp_ oops
06:34 dod joined #mojo
06:53 maxhq joined #mojo
06:55 Vandal joined #mojo
07:05 yakudza joined #mojo
07:05 fhelmber_ joined #mojo
07:13 trone joined #mojo
07:13 dod joined #mojo
07:40 nicolaas joined #mojo
07:56 Britzel joined #mojo
07:57 arthas joined #mojo
08:04 rem_lex joined #mojo
08:23 cosmincx joined #mojo
08:25 preflex joined #mojo
08:46 jzawodn joined #mojo
09:12 abra_ joined #mojo
09:26 punter joined #mojo
09:44 mrphilov joined #mojo
09:55 batman https://metacpan.org/module/Promises # nice :)
09:57 batman sri: haha! fixing the typo now (colleagues)
09:57 marcus joined #mojo
09:58 wk joined #mojo
09:59 fhelmbe__ joined #mojo
10:11 cosmincx joined #mojo
11:06 avkhozov joined #mojo
11:18 iankent joined #mojo
11:26 bowtie_ joined #mojo
11:31 KindTwo joined #mojo
11:59 rem_lex|pivo joined #mojo
12:05 dotan joined #mojo
12:05 SmokeMachine joined #mojo
12:13 dod joined #mojo
12:17 dotan joined #mojo
12:20 sri hmmmmm
12:21 * sri wonders if ensure_index in mango should be split into create_index and ensure_index
12:22 sri every mongodb driver handles it differently
12:23 sri python has a 300 second timeout on indexes created with both, node.js actually checks the index list on the server for ensure_index, and java keeps a list of created indexes unconditionally
12:29 dod joined #mojo
12:33 rem_lex| joined #mojo
12:35 diegok sri: it's very common for an index to take much more than 300s to complete :-/
12:36 sri diegok: i don't think that's relevant
12:37 diegok you're right! :-/
12:40 sri indexes are created by an insert into a magical collection called system.indexes, i don't think it matters how many inserts you perform for the same index
12:40 sri it's magical because you need to run a command to actually drop an index again, remove will not work
12:41 sri i'm rather puzzled by how the node.js driver works
12:41 sri performing a query to check if an index exists is so much more expensive than just inserting into the magical collection again
12:42 sri starting to get the feeling that all the create_index/ensure_index hacks might just be legacy shenenigans
12:42 diegok sure... iirc mongodb docs suggest to just call ensure every time...
12:42 diegok s/ensure/ensure_index/
12:42 sri which is ambiguous
12:43 sri every driver implements ensure_index differently
12:43 sri i have it implemented to always perform the insert unconditionally atm, which works just fine
12:45 diegok it looks like the thing I can see on the console js function :-o (never looked at it)
12:46 diegok ensure_index just call create_index that just insert the thing.
12:47 diegok :)
12:59 hrupp_ joined #mojo
13:09 crab insure_index!
13:12 punter joined #mojo
13:16 fhelmber_ joined #mojo
13:16 ver joined #mojo
13:33 bluescreen joined #mojo
13:42 punter Anyone able to answer this? https://groups.google.com/forum/​#!topic/mojolicious/cERZa_LFTMM
13:42 btyler joined #mojo
13:42 crab punter: give me a subject or message-id?
13:43 punter crab: Oh, I thought I deep-linked, sorry. "Problems holding websocket-related data in memory"
13:43 crab you might've, i just don't want to look at google groups ever again
13:45 * crab thinks
13:46 crab isn't there a perl module that opens an in-memory sqlite db that persists only for the life of the process?
13:46 punter ! If there's such a module, I'm save
13:46 punter d
13:46 punter I'll check
13:48 punter Can't find it
13:49 crab btw, if you're storing this stuff in a relational database anyway, i'd just use postgres and not worry about performance (i.e. in-memory or not). just let the os cache stuff.
13:49 jast depending on the backend, there might be multiple processes involved, so having a per-process sqlite db might cause desyncs
13:50 crab the likelihood that you're going to want to look up data for something that's stored by another backend process increases over time, and having one database is better than having a bunch of small databases comepting for resources.
13:50 crab s/ept/inept/
13:50 Zx3 If the filename $dbfile is ":memory:", then a private, temporary in-memory database is created for the connection. - from DBD::SQLite
13:51 jast cominepting?
13:51 crab Zx3: that's it!
13:51 jast sounds good to me
13:51 punter Zx3++
13:51 punter Thanks a lot EVERYONE
13:51 crab don't thank me, all i did was to discourage you from doing what you wanted.
13:52 punter hehe
13:52 punter If I need to share between processes, I have a Redis for that already up and running
13:53 punter I'll write a followup email about :memory:
13:53 d4rkie joined #mojo
13:53 crab why on earth do you want to use redis AND sqlite?
13:54 jast wouldn't you have exactly the same problem with :memory: that you would have with a normal perl data structure once it goes out of scope?
13:54 jast (or stops being referenced)
13:55 crab why?
13:55 punter crab: Right now I use Redis just as a pub/sub mechanism
13:56 crab jast: presumably it lives as long as the handle to it is in scope, so it sounds ideal in this case.
13:56 punter jast: I'll make sure it doesn't go out of scope then
13:57 jast then why not just use a perl hash or something?
13:57 jast it didn't sound like you *needed* a relational data store
13:57 punter jast: I need it to hold complicated relations with many tables, indexed on many fields.
13:57 jast okay, makes sense then
13:58 punter jast: If it were simple, I'd hold everything in 5 hashes
13:58 jast my bad
14:15 gryphon joined #mojo
14:23 sh4 joined #mojo
14:44 cosmincx joined #mojo
15:21 sh4 joined #mojo
15:22 rem_lex|pivo joined #mojo
16:12 Britzel_ joined #mojo
16:37 dod joined #mojo
16:51 asarch joined #mojo
17:47 bowtie_ joined #mojo
18:01 trone joined #mojo
18:09 delias joined #mojo
18:16 btyler joined #mojo
18:20 asarch joined #mojo
18:21 fhelmber_ joined #mojo
18:30 Mike-PerlRecruiter_ joined #mojo
18:32 dod joined #mojo
19:04 doby joined #mojo
19:06 dwierenga joined #mojo
20:00 thaljef joined #mojo
20:02 thaljef Anyone ever used the content() helper with the Haml renderer?  I'm starting to really like Haml, but I can't figure out how to stash little bits of markup the way I did with ep.
20:32 Adura joined #mojo
20:44 tm joined #mojo
20:45 tm left #mojo
21:20 * sri doubts there are that many folks using haml
21:24 jamesw there needs to be a scss for haml
21:25 thaljef *is bummed*  haml is much more enjoyable to write.  except when it's not.
21:25 thaljef time for a `get reset --hard`
21:25 thaljef s/get/git/
21:26 preflex_ joined #mojo
21:26 sri a perl port of slim would be nice http://slim-lang.com
21:34 marcusie joined #mojo
21:38 lukep joined #mojo
21:47 maxhq joined #mojo
22:43 gtodd joined #mojo
23:27 Mikey joined #mojo
23:31 gautier joined #mojo

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