- Added support for GoldSrc voice mode, including initialization and shutdown functions.
- Implemented data size verification for the GoldSrc voice package.
- Creation and processing of voice packets, CRC verification of data integrity.
- Added a Steam ID field to the client_static_t structure
- Automatic gain control has been introduced for voice samples.
- The voice status structure has been updated.
- Refactored voice data handling for better type consistency and maintainability.
- Fixed ISO C90 compliance issues for better portability.
- Fixed decoder index usage.
* Despite Opus Custom have strict requirements, it's more barebones, allowing us to use maximum
frame size and custom sample rate, without resampling
* Encode each frame size to network buffer, allowing smooth voice chat even in 10 FPS
* Fix possible buffer overruns, underruns and races with platform side
* Revise all usages of offset variables, samples vs bytes
