Added the ability for the user to provide a DataSourceFactory, which cAudio will use to get audio data. This interface can be overridden to provide cAudio the ability to read from zip archives, TCP streams, potentially any location. DataSourceFactories can also be prioritized.
Made the existing FileSource a factory and registered it by default with the manager. If the user wishes to prevent cAudio from reading from disk, unRegister the "FileSystem" data source.
Made tutorials conform with the changes to the API.
Added a convenience function (move()) to the cAudio source, making it easier to move a source and do velocity for doppler effects with a single call.
Internally, audio sources are now stored in an std::vector for better performance. This change also makes it so that it is no longer required for the user to name a source. The user can pass NULL to name to avoid naming it.
Some documentation fixes.
Added a stress test app, allowing users and developers test how well cAudio performs under pressure.
Removed an unnecessary include in cAudioManager.h
Added a define to set the number of buffers used by an audio source
Added the last of the previously unsupported OpenAL source parameters (Cones, attenuation tweaking, max/min volume, and rolloff factor)
Cleaned up the implementation of cAudio source. It no longer keeps a redundant copy of the state information and functions are now in order that they are defined. Some functions have been renamed to be clearer.
Other changes throughout the code to support the changes to the cAudio source.
Updated logger to allow a user to create callbacks to receive log messages
Logger also gives more information, time the message was sent, the source, the importance, and the message itself.
User can now query playback devices and specify the device for the audio manager to use. The audio manager is no longer a singleton, although creating multiple managers is not advised.
Clean up and optimization of manager initialization code.
IAudioCapture can now be told which device to capture audio from. The user can also create multiple IAudioCaptures to capture audio from multiple sources at the same time.
Added additional error checking.
A seperate thread is used to capture audio now.
Defines added to allow a user to remove support for default codecs.
Made cAudio, cAudioCapture, cAudioManager and other classes thread safe.
Made cAudioManager use a seperate update thread by default.
Added cAudioSleep, a cross-platform sleep() function
Moved global defines to cAudioDefines
Added defines to disable thread safety or the internal update thread if the user wishes it
Updated tutorials to reflect the changes made
Updated tutorials to fix minor bugs and pathing issues. Also fixed crash bug on failure to create the audio object.
Added msvc projects for all tutorials.
Updated the listener class to be self contained and only handle stuff related to the OpenAL listener. It does NOT init OpenAL anymore, that has been moved to cAudioManager.
Extended the listener class to support all settings that native OpenAL supports.
Cleaned up cFileSource and fixed a crash bug on NULL file handle
Fixed returning bad audio buffer chunk sizes from the cOggDecoder on errors in the ogg stream
Seeking can now be down in fractions of a second now, changed the seconds field from int to float
Fixed various odd formatting.
Fixed potential crash bug in cMemorySource if memory could not be allocated.
cMemorySource will no longer clear the buffer you give to it before filling it with data. This prevents an overwrite from happening in case of error but the user should provide a zeroed buffer to cMemorySource anyway for safety.
Relative seeking is now supported by cOggDecoder.
Joshua Jones