FoFiX 4.0.0 Development Thread

[raynebc]

If this project is going to be put down indefinitely, it would be nice to have the Sysex parsing bug fixed.

[blazingamer]

None of us really understand the MIDI parser so I'm sorry, but I can't help you there :/ We really do need to get out there asking for help.

[NewCreature]

What is the problem with the Sysex parsing? It should be easy to fix as long as FoFiX isn't supposed to be doing anything with these events. The code looks ok but it doesn't handle divided Sysex events at all:

Code: Select all

            # Is it a sysex_event ??
            elif status == SYSTEM_EXCLUSIVE:
                # ignore sysex events
                sysex_length = raw_in.readVarLen()
                # don't read sysex terminator
                sysex_data = raw_in.nextSlice(sysex_length-1)
                # only read last data byte if it is a sysex terminator
                # It should allways be there, but better safe than sorry
                if raw_in.readBew(move_cursor=0) == END_OFF_EXCLUSIVE:
                    eo_sysex = raw_in.readBew()
                dispatch.sysex_event(sysex_data)
                # the sysex code has not been properly tested, and might be fishy!

The code falsely assumes that the event will not be divided but still checks for the terminator byte. If the event is divided, the parser will choke because it will be one byte behind. If the sysex events aren't being used, you should be able to just do this to fix it:

Code: Select all

            # Is it a sysex_event ??
            elif status == SYSTEM_EXCLUSIVE:
                # ignore sysex events
                sysex_length = raw_in.readVarLen()
                sysex_data = raw_in.nextSlice(sysex_length)
                dispatch.sysex_event(sysex_data)
                # the sysex code has not been properly tested, and might be fishy!
            elif status == SYSTEM_EXCLUSIVE_CONTINUED: #needs to be defined in constants.py
                # ignore sysex events
                sysex_length = raw_in.readVarLen()
                sysex_data = raw_in.nextSlice(sysex_length)
                dispatch.sysex_event(sysex_data)


The MIDI spec I am looking at is confusing, though. It suggests the terminator should be included in the data that sysex_length covers. If that is true, the above code should work. If the terminator is separate then this should work:

Code: Select all

            # Is it a sysex_event ??
            elif status == SYSTEM_EXCLUSIVE:
                # ignore sysex events
                sysex_length = raw_in.readVarLen()
                sysex_data = raw_in.nextSlice(sysex_length)
                if raw_in.readBew(move_cursor=0) == END_OFF_EXCLUSIVE:
                    eo_sysex = raw_in.readBew()
                dispatch.sysex_event(sysex_data)
                # the sysex code has not been properly tested, and might be fishy!
            elif status == SYSTEM_EXCLUSIVE_CONTINUED: #needs to be defined in constants.py
                # ignore sysex events
                sysex_length = raw_in.readVarLen()
                sysex_data = raw_in.nextSlice(sysex_length)
                if raw_in.readBew(move_cursor=0) == END_OFF_EXCLUSIVE:
                    eo_sysex = raw_in.readBew()
                dispatch.sysex_event(sysex_data)

[raynebc]

It should be sufficient for FoFiX to ignore the event by just parsing past it, like FoFLC does. The bug opened on this problem is here:
http://code.google.com/p/fofix/issues/detail?id=1268

Without knowing more about Python, I'm assuming that the problem is that there's no sysex_event() method defined, so the "dispatch.sysex_event()" that occurs fails. Perhaps that part can be commented out, since nothing is expected to be done with the data?

[VeTRooTpat]

I consider, that you are not right. I am assured. Let's discuss. Write to me in PM.

[thekiwimaddog]

Or you could just talk about it here, where it's currently being discussed.

[NewCreature]

Looking at the code I can see that 'dispatch.sysex_event' should be 'dispatch.system_exclusive'. That would at least let the program continue to run. Still not sure if the sysex parsing itself is correct.

Edit: The dispatcher doesn't do anything with sysex events so it would probably be better to just eat up the sysex events and not try to dispatch them. If nobody else wants to try to fix this, I can do it. I just need a chart with sysex events to test with.

[raynebc]

Any notes.mid file written by EOF that has slider/hi hat/rimshot/open strum notation should be fine for testing.

[NewCreature]

Removing the line 'dispatch.sysex_event(sysex_data)' does indeed fix the problem. Here is a patch:

FoFiX MIDI Sysex Patch

[raynebc]

Nice! Hopefully they'll make a new release with this fix.

[weirdpeople]

NewCreature i have have applied your patch.

[italianstal1ion]

This is great news. Thanks developers!

[weirdpeople]

Just fixed vocals, there is just some rockmeter issues but they work so far. There is also this crash that randomly started happening after i posted this originally.....

[lastsaves]

Hi guys,
is there any chance for FoFiX 4.0 being actively developed?
Its quite awesome stuff and makes me sad that is on the halt since November 2010.
Thumbs up for anyone willing to continue.

[AustinG]

fofix 4.0 dont work on my comp, it keeps saying stuff bout log problems