well maybe... I've been stricken by a serious case of yoshimi block. Stricken
by a
few things actually. The subject of my health came up momentarily - it's true,
I'm
really not very well. But mostly I am having a good time :-). Life in the new
flat is
nice - excellent light, quiet enough (the occasional drum practicing neighbour,
all
the usual St Kilda stuff), a nice place to be.
Kind people have wished me a speedy return to health, but that isn't what's
happening.
Pancreatic cancer (5 years) plus breast cancer (two years) simply don't work
that way.
The progression of the disease is well documented, so the question is how fast.
So far
for me it's been remarkably slow, and I'm comfortable with that. The
researchers just
love me :-).
Over the past few months I'd been experimenting with various aspects of yoshimi,
including some serious exploration of a dssi implementation. I pretty much gave
up
on that shortly before the move, when I started looking at what it would take to
provide sensible program bank management. Dssi is not for me, nor for too many
others
as best I can tell.
I'd also been chasing down ways to improve the midi latency, and maybe (just
maybe)
it's better, certainly the code's better organized. A couple of points still
need
attention, but I quite like it.
Alsa audio ... now separates the synth buffer size from the alsa buffer size,
improving latency (allegedly).
A side issue of the dssi program bank thing was some exploration of rdf, and
that has
genuine potential for replacing the existing xml. Now that would be a _brave_
move!
Thinking about it, maybe it's not quite the utter shambles I've been imagining.
Which
brings me back to the current state of yoshimi block. Everything's been plugged
in for
at least a week, yet I haven't been able to bring myself to play a note, let
alone a
chord. I've ventured into JEdit a couple of times, but very quickly found other
things
to do. All very weird, but then the whole of life as I knew it has been turned
on it's
head of late (very much for the better!).
There's a couple of fixes that deserve to be out there - Nedko's lazy interrupt
thing,
and sse detection to appease the package builders. I'll try to put together at
least
that sometime soon, based on 0.056, not the wild experimental stuff. I've no
idea if
or when I'll try to make some sense of the experimental stuff.
cheers folks!