Be

This isn't at the top of my priorities now, but I may eventually get around to working on this. Easier to start in this small proof-of-concept repo than a real...

@freysteinn Did you ever resolve this issue? I'm considering replacing my RME Babyface Pro with the AudioFuse Studio for a bit more I/O and easier control on the hardware. I...

Thanks for the quick response. I am glad to hear that Arturia didn't just write you off and had the decency to let you return it. I have been working...

While this may be useful for local development and use, it would not be practical to distribute applications using midir with this feature until [dynamic loading is implemented upstream in...

> it would not be practical to distribute applications using midir with this feature until dynamic loading is implemented upstream in jack-sys. I suppose that doesn't really matter without midir...

Version 0.9 of the JACK bindings have been released together with version 0.3 of jack-sys. This brings major breaking changes for jack-sys because it now uses [dlib](https://docs.rs/dlib/latest/dlib/) instead of always...

> supporting the jack feature on Windows is a breaking change Yes, so bump the minor version to indicate that. > Also, I assume that https://github.com/Boddlnagg/midir/issues/102 would apply to Windows...

Is this fixable within midir's design? If not, I suggest to remove the JACK backend.

Zero chance of heap allocation http://www.rossbencina.com/code/real-time-audio-programming-101-time-waits-for-nothing

Are sysex messages actually arbitrarily sized or is there a defined upper limit? > We actually call the user input handler (data.callback) within this code. So any restrictions would apply...