More Song Bugs? (C++ on Windows)

By jlnr (dev) Date 2011-10-18 20:11

Is there any debug message in the output? Because it calls _wassert and then abort(), as if there was an error condition that Gosu/libogg is checking for.

By jlnr (dev) Date 2011-10-18 21:00

Oh, there is actually an assert() in the OggFile method mentioned in your backtrace. I assert that this input value from the file is always 0:

> bitstream
> A pointer to the number of the current logical bitstream.

I have no idea what that even means. There can be multiple bit streams in a single ogg file? Hmmm...

By Maverick Date 2011-10-18 21:01

By jlnr (dev) Date 2011-10-19 06:49

How did you generate the ogg files? You could try exporting them again from Audacity (aah quality loss!!), at least for now. Maybe some bizarro tool uses the index. (Speculation: They might use the additional streams to implement segments for looping?)

By Maverick Date 2011-10-19 11:50

I am using audacity for creating the files. I'll look into anything else that may be the problem. (This doesn't happen on mac side though.)

By Maverick Date 2011-10-19 06:28

If you find anything else out let me know. I'll find out with the output said tomorrow. I'd like to not crash user's computers. :P

By Maverick Date 2011-10-19 23:51

Interestingly, it runs fine now under a Release build. Even tested it on a few friend's PCs. *shrugs* Terrible "bug solving" but hey.

By jlnr (dev) Date 2011-10-20 07:29

Yeah, because Release builds don't have assert() in them. If the sample/song sounds correct as well, maybe I should just remove this assert()... I was being a bit defensive there because I have no clue what this stream thing even means. :)