Perl 6 - the future is here, just unevenly distributed

IRC log for #sdl, 2015-03-17

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

All times shown according to UTC.

Time Nick Message
01:23 perigrin joined #sdl
04:02 FROGGS_ joined #sdl
06:57 dod joined #sdl
07:02 dod joined #sdl
08:33 fuzzix joined #sdl
08:59 basiliscos joined #sdl
09:12 basiliscos joined #sdl
10:24 jq joined #sdl
10:33 Mithaldu FROGGS_: am i reading this right that anyone using pollevent has no need to call pumpevent?
11:08 SDL [SDL] wchristian created remove_pump_events (+1 new commit): http://git.io/hmSg
11:08 SDL SDL/remove_pump_events 39885ac Christian Walde: remove apparently unnecessary pump_events to smooth out performance profile...
11:13 Mithaldu FROGGS_: do you know what happened to kthakore?
11:26 travis-ci joined #sdl
11:26 travis-ci PerlGameDev/SDL#70 (remove_pump_events - 39885ac : Christian Walde): The build passed.
11:26 travis-ci Change view : https://github.com/PerlGameDev/SDL/commit/39885ac837f0
11:26 travis-ci Build details : http://travis-ci.org/PerlGameDev/SDL/builds/54702703
11:26 travis-ci left #sdl
12:42 FROGGS_ Mithaldu: kthakore was busy with his RL (one or two startups etc)
12:42 Mithaldu well, at least nothing happened to him then
12:43 Mithaldu i tried looking online after pinging him in that PR and saw almost no activity
12:54 FROGGS_ "As this function implicitly calls SDL_PumpEvents(), you can only call this function in the thread that set the video mode. "
12:54 FROGGS_ where "this" is SDL_PollEvent
12:54 FROGGS_ so yes, you are reading correctly
13:23 Mithaldu FROGGS_: baller
13:23 Mithaldu would be good to have that in the next release then :)
13:30 FROGGS_ sure
13:34 Mithaldu btw, FROGGS_: re Changes, it is good practice to have Changes entries recorded only in the commit that commemorates a release
13:34 Mithaldu make commit history less churny
13:38 FROGGS_ I dont mind adding Changes every now and then
13:39 FROGGS_ it is more important to have an accurate Changes file instead of a nice git history
13:40 Mithaldu FROGGS_: no that's the point
13:40 Mithaldu you only get an accurate changes file by collating the changes *when* you release
13:40 FROGGS_ that certainly also depends on how many collabs you have
13:40 Mithaldu otherwise you need to double-check everything anyways in case someone put stuff into changes and it didn't get properly removed
13:41 FROGGS_ and how these write their commit messages
13:41 FROGGS_ that is always your duty
13:41 Mithaldu that's why you work with feature branches and have someone browbeat contributors into writing useful messages :P
13:43 FROGGS_ yeah, sure :P
13:43 Mithaldu i mean, you've seen my last commit message
13:43 FROGGS_ yes, which was a bit too much :o)
13:43 Mithaldu nope
13:43 Mithaldu it was exactly the required level
13:44 Mithaldu in case someone finds in 3 years that removing that *did* break something subtle, they'll be able to trace back why we broke it
13:46 FROGGS_ you dont have to explain trice why you remove an unneeded call to something
13:46 Mithaldu i have to when there are previous notices of someone observing issues with it
13:46 Mithaldu heck, that's my entire point
13:46 FROGGS_ and embedding screenshots in markdown which are not visible in the commit message does not help either
13:47 Mithaldu that's the kind of quality a project like this needs
13:47 Mithaldu i know, and the commit message doesn't really need it, and that's also why i described it ;)
13:47 FROGGS_ you can state a performance penalty in shorter words
13:48 Mithaldu i just had to laugh
13:48 FROGGS_ next time you could include your source code as well :o)
13:48 Mithaldu i spent the past 2 weeks analyzing a myriad combination of timing issues
13:49 Mithaldu it's easy to say "performance is better", but 2 years later it's clear as mud exactly what that means
13:50 Mithaldu also, my code is public :v
13:50 Mithaldu https://github.com/wchristian/Microidium
13:50 Mithaldu i should replace that screenshot at some point
13:50 FROGGS_ yes, I am not saying that a commit message shall not contain the "why", but you could have described it with less words without loosing content
13:52 Mithaldu i disagree
13:53 FROGGS_ well, you are free to do so
13:53 Mithaldu i wrote it to address the questions and info requests i'd have myself when looking at it in 2 years and having forgotten 90% of it
13:53 Mithaldu i mean
13:53 Mithaldu heck
13:53 Mithaldu i'm still not sure whether that'll blow something up
13:54 FROGGS_ sure but who cares about red sparks? talking about a performance loss due to that call is enough, isnt it?
13:54 FROGGS_ referencing the commit that introduced and the text surrounding it is very good of course
13:54 Mithaldu red sparks in this case could mean ignored or delayed inputs
13:55 Mithaldu which is worse than 30 wasted nanoseconds
13:55 FROGGS_ yes, you know that since you know what these red sparks are
13:55 FROGGS_ nobody else does though
13:56 Mithaldu so you're saying i didn't put enough in the commit message! :D
13:56 FROGGS_ IMO everything up to the url to pullevent is enough in here: https://github.com/PerlGameDev/SDL/commit/39885ac837f0
13:56 FROGGS_ >.<
13:56 Mithaldu hahaha
13:56 FROGGS_ no, you put unrelated stuff in there
13:56 Mithaldu i dropped the entire bug report and resolution in there
13:56 Mithaldu which is the optimum
13:57 Mithaldu do you dislike big commit messages for some reason?
13:57 FROGGS_ well, if it starts to be distracting...
13:57 Mithaldu how?
13:57 Mithaldu trying to understand you here :)
13:58 Mithaldu maybe your workflow is different than mine or something
13:58 FROGGS_ a commit message that explains the how and why in six lines should be enough usually
13:58 Mithaldu usually, yes
16:57 dod joined #sdl
17:18 SDL [SDL] KohaAloha pushed 1 new commit to master: http://git.io/h3Bd
17:18 SDL SDL/master 983d9a0 Mason James: add travis build status icon to README.md
17:20 SDL [SDL] wchristian force-pushed remove_pump_events from 39885ac to 0834b0a: http://git.io/h30U
17:20 SDL SDL/remove_pump_events 0834b0a Christian Walde: remove apparently unnecessary pump_events to smooth out performance profile...
17:27 SDL [SDL] KohaAloha pushed 1 new commit to master: http://git.io/h3zD
17:27 SDL SDL/master 3c93c9a Mason James: tweak build alerts
17:29 SDL [SDL] wchristian force-pushed remove_pump_events from 0834b0a to cc9b323: http://git.io/h30U
17:29 SDL SDL/remove_pump_events cc9b323 Christian Walde: remove apparently unnecessary pump_events to smooth out performance profile...
17:57 perigrin left #sdl
18:00 travis-ci joined #sdl
18:01 travis-ci PerlGameDev/SDL#72 (master - 983d9a0 : Mason James): The build passed.
18:01 travis-ci Change view : https://github.com/PerlGameDev/SDL/compare/9ad43f4f7238...983d9a0bab33
18:01 travis-ci Build details : http://travis-ci.org/PerlGameDev/SDL/builds/54750385
18:01 travis-ci left #sdl
18:10 travis-ci joined #sdl
18:10 travis-ci PerlGameDev/SDL#73 (remove_pump_events - 0834b0a : Christian Walde): The build has errored.
18:10 travis-ci Change view : https://github.com/PerlGameDev/SDL/compare/39885ac837f0...0834b0a21c4a
18:10 travis-ci Build details : http://travis-ci.org/PerlGameDev/SDL/builds/54750755
18:10 travis-ci left #sdl
20:51 mtj heya Mithaldu, FROGGS_
20:52 mtj Mithaldu, interesting performance stuff... ^
20:53 FROGGS_ hi mtj
20:53 mtj ..your luftrauser code is looking nice too  :)
20:53 mtj hey FROGGS_, long time, no chat buddy
20:53 FROGGS_ :o)
20:53 mtj life is busy huh
20:53 FROGGS_ here is usually nobody to talk to
20:54 FROGGS_ well yes, busy in #perl6 :o)
20:54 mtj yep, ive been tracking #perl6 irc, for some time now..
20:54 mtj im just a lurker there
20:55 FROGGS_ yeah, spotted you :P
20:55 mtj tee hee
20:55 FROGGS_ I'm currently fighting a nasty MOP bug :o(
20:57 mtj hmm, MOP?
20:57 mtj aah googled
20:58 FROGGS_ meta object protocol
20:58 FROGGS_ we have a grammar profiler that changes what the grammar keyword means
20:59 FROGGS_ and something goes nuts on jvm
21:00 mtj yow, sounds deep
21:01 mtj curious... have you learnt any java from your perl6 hacking, FROGGS_?
21:01 mtj p5-mop project is all new to me
21:01 FROGGS_ mtj: I've learned enough to not like Java's syntax :o)
21:02 FROGGS_ but I do like parts of JNI/JNA
21:02 mtj ha, no surprises there.. ^
21:02 FROGGS_ mostly the fact that it works probably
21:02 mtj nice
21:02 FROGGS_ though, it is, as Java, too verbose
21:04 mtj yeah, its always been my problem with it too :/
21:05 mtj you really need an IDE with java, to help autoselect code for you
21:05 mtj else... cobol hands :/
21:05 FROGGS_ yes, but I usually dont like auto completion either :/
21:06 mtj yep
21:07 mtj FROGGS_:  have you peeked at  http://nim-lang.org...
21:08 FROGGS_ it looks okayish
21:10 kaare joined #sdl
21:13 mtj just reading its wiki now..
21:14 mtj i didnt know it was python/pascal..ish
21:14 mtj hmm... anyhoo :)
21:17 mtj this bit i found interesting tho..
21:17 mtj "The compiler generates optimized C code and defers compilation to an external compiler[12] (a large range of compilers including clang and GCC are supported) to leverage their optimization and portability capabilities. "
21:17 mtj http://en.wikipedia.org/wiki/Nim_(programming_language)
21:23 FROGGS_ like Rust
21:40 Mithaldu mtj: cheers, and thanks :)

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