next »

[InfiniteStateMachine]

hahahahah, I can certainly relate

[Boreal]

Yeah, it's ridiculous how often this happens.

I'm running my app through gDEBugger and getting an access violation on my glDrawArrays call.  Which is silly since I've loaded three vertices and I'm drawing three vertices.  It's an extremely simple test program, and I can't understand why it's running but not drawing when run directly and crashing when run through the debugger.

[tjcbs]

I hate it when people use the phrase "code smell". It is so fucking obnoxious. I always picture an awful, deeply pretentious programmer sticking his long nose in the air and sniffing. It's made worse by the fact that half the time these "smells" are perfectly fine, and its just some jackass pretending his stupid ideology is some kind of absolute.

[Layl]

I hate it when people use the phrase "code smell". It is so fucking obnoxious. I always picture an awful, deeply pretentious programmer sticking his long nose in the air and sniffing. It's made worse by the fact that half the time these "smells" are perfectly fine, and its just some jackass pretending his stupid ideology is some kind of absolute.

The entire point behind the concept of "code smell" is to investigate each to see if it's more than just a smell. That said, I've met plenty of people who skip the investigating and immediately assume the worst.

[tjcbs]

The entire point behind the concept of "code smell" is to investigate each to see if it's more than just a smell. That said, I've met plenty of people who skip the investigating and immediately assume the worst.

Hah, you're right, I I guess I missed the point. I took noun-form "smell" to be similar to noun-form "fail", which was irritating to me. But it does seem like people use it that way, as you say.

[ThemsAllTook]

The run loop code I've been using on Mac OS X for years is suddenly only capable of giving me 30 FPS when running in a window, even when running a game that's not taxing in the slightest, and even with vsync off. There's some major inefficiency or timing issue somewhere in here, but I've already eliminated several possible culprits with no change in behavior. I'm at work now so I can't keep poking at it, but it's going to be bugging me all day...

[oahda]

Literally.

[oodavid]

harumph - cross-platform data persistence is a cockhead

[shellbot]

Did anybody know that pygame only plays uncompressed sound files?! Because apparently only the people who MADE pygame know that pygame only plays uncompressed sound files. Nobody on StackOverflow even mentioned this. It took me 3 hours to learn that I have to convert all my files...

I guess it's my own damn fault for not reading the documentation more carefully

[oahda]

Did anybody know that pygame only plays uncompressed sound files?! Because apparently only the people who MADE pygame know that pygame only plays uncompressed sound files. Nobody on StackOverflow even mentioned this. It took me 3 hours to learn that I have to convert all my files...

I guess it's my own damn fault for not reading the documentation more carefully

Make sure you post your own solution on StackOverflow so that others can find it in the future!

[InfiniteStateMachine]

You mean like mp3's?

I think it's on the safe side to assume a sound library doesn't handle compressed files unless the docs say so.


That's a bit of an issue with Haxe. Windows targets need wav's and flash targets need mp3's. You have to do a little preprocessor stuff to get around it.

[shellbot]

You mean like mp3's?

I think it's on the safe side to assume a sound library doesn't handle compressed files unless the docs say so.


That's a bit of an issue with Haxe. Windows targets need wav's and flash targets need mp3's. You have to do a little preprocessor stuff to get around it.

Oh I wasn't aware of that. I haven't worked with sound much on any of my projects before. Next time I'll be sure to check the docs first.

Did anybody know that pygame only plays uncompressed sound files?! Because apparently only the people who MADE pygame know that pygame only plays uncompressed sound files. Nobody on StackOverflow even mentioned this. It took me 3 hours to learn that I have to convert all my files...

I guess it's my own damn fault for not reading the documentation more carefully

Make sure you post your own solution on StackOverflow so that others can find it in the future!

I would but all the threads are from years ago

[InfiniteStateMachine]

You mean like mp3's?

I think it's on the safe side to assume a sound library doesn't handle compressed files unless the docs say so.


That's a bit of an issue with Haxe. Windows targets need wav's and flash targets need mp3's. You have to do a little preprocessor stuff to get around it.

Oh I wasn't aware of that. I haven't worked with sound much on any of my projects before. Next time I'll be sure to check the docs first.

Did anybody know that pygame only plays uncompressed sound files?! Because apparently only the people who MADE pygame know that pygame only plays uncompressed sound files. Nobody on StackOverflow even mentioned this. It took me 3 hours to learn that I have to convert all my files...

I guess it's my own damn fault for not reading the documentation more carefully

Make sure you post your own solution on StackOverflow so that others can find it in the future!

I would but all the threads are from years ago

It's ok to still post your findings.

This comic is relevant

[ThemsAllTook]

The run loop code I've been using on Mac OS X for years is suddenly only capable of giving me 30 FPS when running in a window, even when running a game that's not taxing in the slightest, and even with vsync off. There's some major inefficiency or timing issue somewhere in here, but I've already eliminated several possible culprits with no change in behavior. I'm at work now so I can't keep poking at it, but it's going to be bugging me all day...

Update: I spent a whole week on this problem before half solving it, and eventually giving up on a complete solution. I wrote this diagnostic test app in case anyone wants to try: http://ludobloom.com/temp/framegraph_stem.zip

Each sample in the graph is the length of time it took to draw the previous frame. V to toggle vsync, alt-enter or cmd-ctrl-F to toggle fullscreen, 0-9 to change the graph scale (0 is auto-adjusting, 1-9 cover ranges of n+1 ideal 60hz frame intervals).

I updated from OS X 10.9 to 10.10, and it improved windowed performance with vsync on quite a bit (40 -> 60 FPS with the window maximized), but an issue that persists is input events causing frame time spikes. The root of this problem runs pretty deep: When I ran the app under Instruments, I caught the event polling API talking to WindowServer while WindowServer was blocked on the GL driver, seemingly waiting for a buffer swap. Once it actually returned the event to my code, after processing it and drawing, I'd of course missed vblank and had to wait for *another* buffer swap. Modifier keypresses and mouse movement under some circumstances seem to trigger it in particular. The problems I observed seemed more severe on my computer than on anyone else's who tested it, so hopefully I'm the odd one out and most people's performance will be OK.

[SuperDisk]

Did anybody know that pygame only plays uncompressed sound files?! Because apparently only the people who MADE pygame know that pygame only plays uncompressed sound files. Nobody on StackOverflow even mentioned this. It took me 3 hours to learn that I have to convert all my files...

I guess it's my own damn fault for not reading the documentation more carefully

This isn't true; MP3 support is flaky, but it has full OGG support, and that's better than MP3 in every way (Including filesize).

[shellbot]

Did anybody know that pygame only plays uncompressed sound files?! Because apparently only the people who MADE pygame know that pygame only plays uncompressed sound files. Nobody on StackOverflow even mentioned this. It took me 3 hours to learn that I have to convert all my files...

I guess it's my own damn fault for not reading the documentation more carefully

This isn't true; MP3 support is flaky, but it has full OGG support, and that's better than MP3 in every way (Including filesize).

Sorry I should have been more specific. Pygame only plays OGG and uncompressed WAV files. All my sound files are WAV

[Geti]

Convert them to oggs then. ogg vorbis is compressed.

@superdisk - "every way" leaves you open to issues like lacking hardware decode support fwiw I'm very pro-ogg but superiority is a relative thing.

[shellbot]

Convert them to oggs then. ogg vorbis is compressed.

@superdisk - "every way" leaves you open to issues like lacking hardware decode support fwiw I'm very pro-ogg but superiority is a relative thing.

But I've heard that ogg has a worse quality than wav. At this point I don't need the sound files to be compressed, so I would rather the higher quality that wavs provide. Unless I was misinformed and oggs are the higher quality option. I've never worked that much with sound before so this is all new to me.

[BorisTheBrave]

But I've heard that ogg has a worse quality than wav

Not true.

wav's are lossless so what you say is literally true - mp3 and ogg are worse quality than wavs. However, this is not a practical concern. Quality level is a setting when encoding oggs and mp3s, and both can be set to have indistinguishably good quality while still being many times smaller than wavs.

[Sik]

Pretty much every compressed format except FLAC has worse quality than WAV, it's inherent to lossy formats =P In practice you won't notice any difference when listening unless 1) you're an audiophile or 2) you overcompressed the file (and how much you can afford to compress depends on the sound). On the flipside, the filesize will thank you.