March 07, 2004 Re: Mathew, here ,more detailed, | ||||
---|---|---|---|---|
| ||||
Posted in reply to Charles Hixson | On Sat, 06 Mar 2004 10:30:49 -0800, Charles Hixson <charleshixsn@earthlink.net> wrote: > Dataflow seems to be something that it's quite difficult to make work properly, though when it does work it gives you an automatic parallelism that no other technique matches. It's as if, e.g., a for loop had no implicit order of operations, so separate iterations could be swapped out to every available processor. (Well, it only ran on a single processor system, so who knows how that would have worked out really, but it was quite impressive...but the [user] code was so bulky that it was quite difficult to use.) Hi, I know, I'm an inventor of a reconfigurable parallel processor that uses dataflow concepts. That's why I'm interested in this stuff. I have gone thru some ups & downs while developing a progamming system for this thing. > Sorry. No links. I may have an old set of Mac floppies for system 7.5 though...somewhere...and they might still be readable. What was it about? Was it a code generator, a compiler, a graphical editor? -- Robert M. Münch Management & IT Freelancer http://www.robertmuench.de |
March 07, 2004 Re: Mathew, here ,more detailed, | ||||
---|---|---|---|---|
| ||||
Posted in reply to Robert M. Münch | Robert M. Münch wrote:
> On Sat, 06 Mar 2004 10:30:49 -0800, Charles Hixson <charleshixsn@earthlink.net> wrote:
>
>> Dataflow seems to be something that it's quite difficult to make work properly, though when it does work it gives you an automatic parallelism that no other technique matches. It's as if, e.g., a for loop had no implicit order of operations, so separate iterations could be swapped out to every available processor. (Well, it only ran on a single processor system, so who knows how that would have worked out really, but it was quite impressive...but the [user] code was so bulky that it was quite difficult to use.)
>
>
> Hi, I know, I'm an inventor of a reconfigurable parallel processor that uses dataflow concepts. That's why I'm interested in this stuff. I have gone thru some ups & downs while developing a progamming system for this thing.
>
>> Sorry. No links. I may have an old set of Mac floppies for system 7.5 though...somewhere...and they might still be readable.
>
>
> What was it about? Was it a code generator, a compiler, a graphical editor?
>
It was a program development system on the Mac. It compiled, probably to virtual machine code. It included a graphic editor, in fact no non-graphic way was devised that could display the code. This probably contributes to its eventual failure. It came out of Canada (Labrador? Newfoundland?). I first encountered it in Datamation, either an article, a product anouncement, or an ad...unless I first encountered it at a show (ComputerWorld?) in San Francisco.
|
March 08, 2004 Re: Mathew, here ,more detailed, | ||||
---|---|---|---|---|
| ||||
Posted in reply to Charles Hixson | Try Google for 'dataflow mac' (no quotes). It will bring up some old stuff, but you will have to look at cached versions, because most of the original websites aren't there anymore or don't still have the stuff. That's a start - probably other 'dataflow + xxxxx' searches will find more. In article <c2fvr9$2vs1$1@digitaldaemon.com>, Charles Hixson says... > >Robert M. Münch wrote: >> On Sat, 06 Mar 2004 10:30:49 -0800, Charles Hixson <charleshixsn@earthlink.net> wrote: .. >> What was it about? Was it a code generator, a compiler, a graphical editor? >> >It was a program development system on the Mac. It compiled, probably to virtual machine code. It included a graphic editor, in fact no non-graphic way was devised that could display the code. This probably contributes to its eventual failure. It came out of Canada (Labrador? Newfoundland?). I first encountered it in Datamation, either an article, a product anouncement, or an ad...unless I first encountered it at a show (ComputerWorld?) in San Francisco. |
Copyright © 1999-2021 by the D Language Foundation