Re: It cannot be done under OS X !!
Re: It cannot be done under OS X !!
- Subject: Re: It cannot be done under OS X !!
- From: Neil Johnson <email@hidden>
- Date: Fri, 26 Mar 2004 20:54:52 +0000 (GMT)
Hi,
Kurt Revis wrote:
>
Their statement that "the data is split up into tiny sysex packets" is
>
very odd. There is no notion of "packets" at the MIDI level, just a
>
stream of bytes.
Well, they may be referring to MIDI File Dump, which is a bit like FTP,
supporting the transfer of arbitrary data, as a stream of packets, across
a MIDI connection.
>
So even if things are packaged up into smaller packets in the computer,
>
or in the communication between the computer and MIDI interface, it
>
shouldn't make any difference at all when it finally gets to the device.
True, the protocol stack should hide the presence of packets completely,
so that it all looks like a continuous stream of bytes.
>
> Because of the way MIDI is implemented on the Mac -- in Windows it is
>
> different -- the data is split up into tiny sysex packets during
>
> transfer, thus interrupting the upload.
>
It sounds like they are saying that the Mac puts longer or shorter
>
pauses between individual bytes of sysex, in a different way than the
>
PC does.
No, I think they are saying that the Mac does the right thing and splits a
file into packets (112 bytes maximum per packet). I have no idea what
Windows does, as there is (as far as I can see) no standard protocol for
streaming arbitrary bytes, so that all data is eventually sent as packets
in one form or another. Remember there's the finnagling around to ensure
that all packet bytes (other than the start and stop bytes) are in the
range 00h-7Fh.
>
But as far as I know, the MIDI spec does not specify absolute timing for
>
individual sysex bytes. As long as the pauses are within reasonable
>
bounds, their hardware should be able to deal with it.
Well, if doing a sample dump there are two timings described in the
standard:
2 secs to wait for an initial ACKnowledge, or abort;
20ms to wait for an ACK between packets, after which the transmitter
switches to open-loop mode and streams packets without waiting for ACKs.
>
But apparently their hardware can't do that, even though hundreds of
>
other MIDI devices (some more than fifteen years old!) can.
Sounds to me a lttle like "We got it running on Windows, and don't want
the hassle of supporting Macs." It even seems, on the info given, that
they *might* have a broken implementation that kinda works with a
*possibly* broken MIDI implementation in Windows. But this is pure
speculation on my part :-)
Cheers,
Neil
--
Neil Johnson :: Computer Laboratory :: University of Cambridge ::
http://www.njohnson.co.uk http://www.cl.cam.ac.uk/~nej22
---- IEE Cambridge Branch:
http://www.iee-cambridge.org.uk ----
_______________________________________________
coreaudio-api mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/coreaudio-api
Do not post admin requests to the list. They will be ignored.