According to the rules (which are hard to find), there must be nothing on the evaluation stack when localloc is called (except its own size parameter). The generated string interop code didn't respect that. This didn't cause any problems in existing code.
* move messageConfig to its own folder
* start MessageEdit user control
* more setup
* wire stuff up
* progress
* progress
* fixes and cleanup
* start mesage row component
* move method from MessageConfig to extension method
* MessageRow - progress
* progress
* Message row - progress
* progress
* fix initial load
* fix Restore btn
* use RadioButtonEx
* start ColorRow user control
* progress
* progress
* progress
* rename
* make color picker work
* cleanup
* cleanup
* Message config - don't immediate set config values on restore defaults, allow user to press ok or cancel to decide to actually save
* cleanup
* cleanup
Looks like a big diff, but isn't. It separates the zwinder buffer part (which does most of the work) with the rewinder impl on top of it. This is part of the effort for a zwinderstatemanager for tastudio, which isn't done yet, but these changes can come in just fine
- adds a static IsValid method to IMiniWatch implementations
- reverts .csproj changes
- 'remove out of range' button on ram search didn't seem to actually do anything. this was fixed
fixes two minor bugs relating to ram watch display type
1.binary display was broken on 32 bit watches
2.changing a watch's size to one with an incompatible display type resulted in an exception (e.g. 32 bit float -> 8 bit unsigned)
for reference, I converted it to .bmp and back w/ ImageMagick
the warning was to stderr and it read `libpng warning: sRGB: cHRM chunk does not
match sRGB`
* Handling for BufferLost
Fixes issue #1212 where adding/removing headphones would lead to a crash by restoring the sound device when it's detected and being defensive with exception handling.
* Internalize the handling of buffer lost and make it a bit more efficient.
Remove interface function for SoundLost and move the logic to soley live inside DirectSoundOutput. Additionally I discovered I did not need to tear down the entire device to restore sound, typically Restore() and Play() handle it. Still need to wrap every place that can throw an exception in a try/catch block and wait for WriteSamples to handle it.
* Update DirectSoundSoundOutput.cs
Logic is hard.
* Retry limiter added to recovery
Added a self reducing retry counter that will try to start sound 5 -> 4 -> 3 -> 2 -> 1 times each time it tries to recover until it succeeds, at which point the counter returns to 5. This allows for quicker attempts at recovery without the risk of an infinite loop or terrible performance from sleeping 10 ms.
YoshiRulz