Re: [fluid-dev] State of my FluidSynth fork

[Josh Green]

Hey Miguel,

I'd be interested in seeing your port.  Do you have a location to
download it from?  As I've mentioned before, FluidSynth itself could use
a lot of cleanup for my own willingness to contribute to it.  If for
some reason your C++ binding looks like the right direction, it may make
sense to use it in FluidSynth itself and provide an older C
compatibility binding.  Of course there may still be differences in
direction of your project versus FluidSynth, in which case they would
remain separate, but it would at least make it easier for the two
projects to share changes.

Best regards,
        Josh


On Wed, 2008-01-09 at 21:17 +0000, Miguel Lobo wrote:
> All,
> 
> It's great to see more activity on the list, and I haven't said
> anything for a few months now, so I thought I'd give people an update
> on how my FluidSynth fork is going.  The objective of this fork is to
> create a synthesizer based on the FluidSynth codebase but written in
> C++ so the code is hopefully as readable and contribution-friendly as
> possible.
> 
> First thing to say is that due to a variety of reasons I have been
> unable to spend as much time on this project as I had hoped.  Also,
> one of the main features I want to complete before releasing the code,
> namely rewriting the synthesis loop so it doesn't need to run on
> fixed-size chunks of data or "frames", is proving more difficult and
> bug-causing than I had thought.
> 
> Keeping in mind these caveats, the current state of the project is as follows:
> 
> * The rewrite is complete.
> 
> * The code builds under Linux and Windows; under the latter it can be
> compiled using mingw32 or MSVC.
> 
> * The ALSA, DirectSound and output-to-file audio drivers are working.
> 
> * The build infrastructure correctly generates additional source code
> from the project's headers.  At the moment this functionality is only
> used to generate setters and getters for public classes, but in the
> future I plan to generate C and Python bindings using the same
> approach.  The C binding generation should be fairly easy, the Python
> one not so much but definitely doable.
> 
> Once I get the code to the state where it can produce reasonable audio
> output for a simple MIDI file and play the result using ALSA and
> DirectSound, or send the audio to a file, I will create a SourceForge
> project, release the code to a public SVN repository and make an
> announcement here (as long as Josh doesn't mind, of course :-).  I'd
> better not attempt to estimate when that will happen.
> 
> Any questions, suggestions or comments are always welcome.
> 
> P.S. As usual, I would like to thank past and present FluidSynth
> contributors for their great volunteer work that makes projects such
> as mine possible.
> 
> Regards,
> Miguel
> 
> 
> _______________________________________________
> fluid-dev mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/fluid-dev
>