Hi all,
I'm absoluitely delighted by all the support Yoshimi is getting these days.
it's probably just as well I don't have a webcam, so you're spared the sight of
my pathetic attempt at break dancing :)
While out today getting some fresh air I had a good long think.
First of all for all the known internal short-comings Yoshimi is in fact
remarkably stable (although I've just been informed of a slightly obscure crash
situation which I'll be looking into myself).
The things I know need sorting are:
1 Midi Learn
2 Gui Separation
3 Command Line Interface
4 Common Interface / API
5 Alsa Audio
The top 4 are quite closely interlinked, and 4 itself it pretty funadmantal.
Number 5 is pretty much a side issue as most people use jack for audio.
Something I've mentioned privately to a few people, but now think should be
aired publicly is my reason for wanting command line access. It is because I
am very keen to make Yoshimi accessible to blind musicians. I've known one or
two in the past and they struggle mightily with most music software currently
available.
The proposals we have on the table all seem to overlap to varying degrees,
However;
Andrew's work is mostly focussed on LV2.
Rob's ideas are a root & branch operation - rather exciting but very invasive.
Louis' Midi Learn upgrade looks like it could be implemented relatively
quickly.
The problem is, how to make best use of peoples time, without wasted
duplicated effort. So here's some ideas to knock about.
Midi Learn is becoming almost an embarrassment, we've had people knock on the
door for over 4 years, and most other systems have it implemented - although
not as nicely as Alessandro's original (from a user's point of view) so I'd
like to ask Louis to carry on with that against the current master. As well as
filling an important hole, what we learn from that can also be applied to a
later 'gold' version.
Meanwhile, if Rob tackles the deeper structural work on a different branch we
can do a switchero later, while always maintaining a stable version for our
users. I don't know if this can be done in modular phases. It would be nice if
it could. That would reduce the divergence while work was going on.
With this low-level work I would be inclined to suggest some discussion between
Rob and Andrew.
At the same time I can continue to tackle odd-ball issues. Things like vector
control are just about fully baked now, but there's that alsa problem and
various gui/usability stuff - it seems to be what I'm best at :o
How does that all sound?
1 Practical
2 Possible
3 Difficult
4 Not Likely
5 You're 'avin a larf mate!
P.S.
Talking about vector control, Harry has been looking at the idea of a gui
section for setting it up as an alternative to NRPNs. I don't know how far he's
been able to get with this.
P.P.S
I now think putting all the NRPN stuff in MusicIO was a mistake and am
considering moving it to SynthEngine. That would take out a lot of
indirection/pointers and put it in the same logical area as all the other
control stuff.
--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.