[Nosc] toy workstation (composite reply)
- Subject: [Nosc] toy workstation (composite reply)
- From: Myron Plichota <myron.plichota@xxxxxxxxxxxx>
- Date: Sun, 9 Dec 2001 11:51:23 -0500
On Sunday 09 December 2001 09:47 am, roelf wrote:
> I'm using and have been using such a thing since 1988, it's called an
> Atari (in those days a very popular 68k based computer in Europe).
> For B M B con. , the performance group I work in, it's the main workhorse.
> No reason to replace it with modern powerbooks and such as far realtime
> sound processing is concerned (Video is something entirely different).
> Allthough...
>
> Specially their last model the Atari Falcon, in which the company had
> put a lot of their experience gained with their TransputerWorkstation.
> CPU, FPU, DSP, Videochip, MIDI, SoundDMA to unburden the CPU, etc.etc.
> Allthough running at a slow 16mhz rate, the integration of all these
> seperate processors and peripherals into some kind of multiprocessor
> machine, made they could easily compete with others having a lot faster
> cpu's.
From everything I've heard they got it right with the Atari. The
multiprocessor architecture approach served them well, and my notion is that
the same thing can be done with fewer and faster general-purpose Forth CPUs
talking to each other.
> I have allways programmed it in Forth only. A Forth from 1985, build
> around the 68k, taking the 68k as point of departure, rather than im-
> posing the prescribed VM from the 83-standard. Depending how one looks
> upon Forth, it was a match well made. So far I never felt the urge to put
> a machine- or colorForth on it.
I still have the Fig-Forth 68K listing in my library. It is Indirect Theaded
Code (ITC) but still quite efficient compared to any other Forth Virtual
Machine that I have seen (NEXT = 2 instructions). It would be a real screamer
if it was recoded as a Subroutine Threaded Code (STC aka MachineForth)
implementation.
> Continuety? None, Atari doesn't excist anymore. We pick up second
> hands, to be set for the future. Because they start to fall apart after
> the harsh treatment they get from us.
> For upgrading the cpu, I think Motorola has discontinued the 68k
> development, the Coldfire doesn't fit.
>
> The Atari is still remebered with a lot of fondness for its concept
> around here, it even has gained a hype status in the electronic (dance)
> music scene.
> A lot of my collegues in the electro-accoustic music niche would be
> very enthusiastic for a new dedicated machine which they could trust
> and program as of old, there're quite some (ex-)Forth users out there!
Precisely why I am so fired up about controlling the longterm availability of
our favourite silicon, mothererboard, and peripheral core technology.
> As for the word 'Toy', Ataris were associated in certain circles as toys,
> and belittled cause of that.
> It never withheld professionals, from Noble price winners in Chemistry
> to obscure artists in the electronic arts world, to use them, not at all!
>
> I do hope I presented evidence that such a concept/idea actually has
> worked and still works, and that a dormant demand might be satisfied
> (This was _not_ an ad for Atari, but rather for the concept, please!).
>
> You probably need more practical support than lip services, I'm sorry.
> I might be usefull in some way though.
Thanks for the encouragement. Please stay tuned, keep offering your views and
the benefit of your experience using the Atari. Digital audio is my own
particular hobbyhorse, but I hope that other end applications will be
addressed, or it is not likely to happen. I see this as an attempt to come up
with a truly general-purpose approach, certainly up to the demands of an
audio workstation, and more. Are there any realtime video animation gurus out
there?
BTW the interprocessor serial links could be just as easily used to connect
main CPUs in a workstation farm as to connect to lowly I/O coprocessors. The
concept has been around for a long time and still persists.
- Inmos Transputer
- Analog Devices Shark
- Ultratechnology F21
- Chuck Moore's 25X (on-chip!)
Myron Plichota
------------------------
To Unsubscribe from this list, send mail to Mdaemon@xxxxxxxxxxxxxxxxxx with:
unsubscribe NOSC
as the first and only line within the message body
Problems - List-Admin@xxxxxxxxxxxxxxxxxx
Main 4th site - http://www.