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

IRC log for #mojo, 2017-06-10

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

All times shown according to UTC.

Time Nick Message
01:08 aborazmeh joined #mojo
01:43 jberger sri: I think I was thinking of Mojo::DOM
01:58 maschine joined #mojo
02:25 tencendur joined #mojo
02:38 noganex joined #mojo
02:54 stryx` joined #mojo
04:04 dboehmer joined #mojo
04:45 irqq_ joined #mojo
06:43 dod joined #mojo
06:49 dod joined #mojo
07:33 karjala_ batman, in Mojo::mysql, is it possible to pass $password as a separate parameter in the new constructor? I'm afraid it might contain : and @ characters which might mess up the connection string
07:36 karjala_ I guess I should uri-encode it
07:39 karjala_ or uri-escape it, even better
08:22 kiwiroy joined #mojo
09:07 someguy joined #mojo
09:18 kiwiroy joined #mojo
10:41 kiwiroy joined #mojo
10:48 salv0 joined #mojo
11:04 trone joined #mojo
11:30 htaunay joined #mojo
11:52 maschine joined #mojo
11:59 sri looks like there is not much interest in this https://github.com/kraih/minion/issues/50
12:00 stryx` joined #mojo
12:12 aborazmeh joined #mojo
12:31 maschine found a broken link at http://mojolicious.org/perldoc/Mojolicious/Guides/Contributing#CODE-OF-CONDUCT
12:32 maschine at the bottom of the section - http://speakup.io/coc.html
12:42 sri it's been broken for a long time
12:42 karjala_ someone asked me to do a marketing mailing list server for them. this "issue 50" suddenly became relevant again, sri. Shouldn't I create one minion job per mail recipient? Isn't that the way to go?
12:42 karjala_ So need a way to add 1000 jobs quickly.
12:43 karjala_ Hm... Maybe I should do all recipients in one job
12:44 sri there is no one right answer, it all depends on what exactly you're doing
12:53 rshadow joined #mojo
13:09 someguy joined #mojo
13:29 irqq joined #mojo
13:55 PryMar56 joined #mojo
14:11 kes joined #mojo
14:20 mat joined #mojo
14:23 kes Hi.
14:24 kes It there in documentation the description
14:24 kes about resolution order for 'content'?
14:26 kes * Topic for #mojo is: ???? nom nom | http://mojolicious.org | http://irclog.mojolicious.org | http://code-of-conduct.mojolicious.org
14:26 kes Both templates generate 'content' buffer, but they works differently
14:26 kes http://paste.scsys.co.uk/564410
14:34 kes and this difference appears only if I use layouts
14:35 pink_mist kes: it's a global stash value afaik - the last content you set is what will be used
14:35 pink_mist global for the request that is
14:36 kes I do not expect any difference when I work with it implicitly or explicit
14:36 pink_mist what
14:36 pink_mist why would there be a difference?
14:37 kes Where may I post app to show this?
14:37 kes the .tar for this app
14:38 pink_mist just use a gist with the different files
14:46 aborazmeh joined #mojo
14:53 kes pink_mist: https://gist.github.com/KES777/451abd9817449d8230de529a32da0e0c
14:53 kes the app to reproduce
14:53 kes https://github.com/KES777/unexpected_rendering
14:53 kes I expect same output for `script/my_app get /2/0`
14:54 kes as for `script/my_app get /2/1`
14:54 pink_mist uhm, you *do* get the same output for those two
14:54 pink_mist oh no, wait
14:54 pink_mist misread
15:01 kes I suppose 'content' buffer in case of implicit usage is created BEFORE template code is run, in case of explicit usage it is created AFTER template is run
15:02 kaare joined #mojo
15:04 kes So in case of implicit use when 'base.html.ep' is processed its 'content' is ignored (New content will be ignored if the named buffer is already in use.) because 'content' is already in use
15:15 ramo joined #mojo
15:38 * sri opened https://github.com/kraih/minion/issues/52 for planning Minion backend optimizations
15:38 sri i think a test case with about a million jobs with dependencies would be a good start
15:45 CandyAngel How long should I run a benchmark between 2 implementation for?
15:46 CandyAngel I have it on 10 seconds which gives me ~1.5 million iterations
15:46 CandyAngel I may have found the smallest optimisation ever :P
15:47 sri up to you to find the right scenario to prove your optimization ;p
15:48 CandyAngel There is absolutely no reason to do this, other than the fact that it is "faster".. for a tiny definition of faster
15:48 sri i would expect EXPLAIN ALALYZE output to show the difference very well though
15:48 CandyAngel Though I am surprised that it is always faster, even when it should be the same
15:49 CandyAngel Oh never mind, it isn't
15:50 CandyAngel In the common case, it is slightly slower
15:50 CandyAngel At least, I presume 1 frame per read is the most common case for a websocket
15:50 sri oh, you were not talking to me :(
15:50 * sri vanishes
15:51 CandyAngel Well I kinda was, just not about the Minion thing
15:51 CandyAngel Just-sorta-related-and-useful-to-know :)
15:52 CandyAngel Maybe for fast things, it is better to run it for X time and for slow things, for X iterations :)
15:55 CandyAngel http://codepad.org/jyBASnmE
15:57 CandyAngel Unless that doesn't even work, which is possible
15:58 CandyAngel It works in my benchmark
15:59 CandyAngel http://codepad.org/Li7e5FGj
16:02 CandyAngel I'm actually curious why 'assign' is slower when you only add one character to $BUFFER
16:03 CandyAngel Oh.. oops, let me fix the test
16:03 CandyAngel Helps if they are actually doing the same thing
16:15 CandyAngel http://codepad.org/Qwq4XafE
16:16 cheako joined #mojo
16:35 tencendur joined #mojo
16:42 kes pink_mist: Did you reproduce that?
16:44 dod joined #mojo
16:45 pink_mist kes: sorry, I had to step away and literally just got back
16:45 pink_mist am about to give it a shot now
16:45 kes ok. Thank you.
16:47 pink_mist just reproduced ... but I'm still not convinced it's a bug ... using layouts, extends, and content like you're doing is a bit on the complicated side, and the behaviour you're showing might be correct
16:49 kes maybe correct, but I did not found documentation for that. so confused.
16:59 tencendur joined #mojo
17:02 cheako joined #mojo
17:09 pink_mist kes: perhaps the content_for and/or content_with helpers may be what you need
17:10 pink_mist https://metacpan.org/pod/Mojolicious::Plugin::DefaultHelpers#content_for and https://metacpan.org/pod/Mojolicious::Plugin::DefaultHelpers#content_with
17:11 irqq joined #mojo
17:13 kes I think, that will be workaround for the problem
17:40 gryphon joined #mojo
18:15 stryx` joined #mojo
18:24 morton joined #mojo
18:32 cheako joined #mojo
18:47 jnbek joined #mojo
19:21 cheako joined #mojo
21:07 stryx` joined #mojo
21:45 cheako joined #mojo
22:22 genioweb joined #mojo
23:26 sri it's funny how google has pushed so much for http/2, and now most of their chrome traffic runs over the quic protocol

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