Re: JiffyDOS, SpeedDOS and IEEE

From: silverdr_at_wfmh.org.pl
Date: Sat, 24 Jan 2015 22:32:01 +0100
Message-ID: <54C40F51.7040404@wfmh.org.pl>
On 2015-01-24 22:17, Ruud@Baltissen.org wrote:
> Hallo allemaal,
>
>
> I could be completely wrong but IIRC JiffyDOS is faster than
> SpeedDOS. If that indeed is the case, I wonder why? My only
> explanation: bad programming.

JD came later, when things were better explored. SpeedDOS in a 
simplified view (there is much more) is only a parallel extension for 
transmission. It speeds-up one of several components, which can be 
improved. Three major other being:

- track seek
- sector seek
- GCR decoding

JD is a different generation and most probably touches some of the other 
areas than transmission only. I wouldn't call it bad programming. It's 
probably just like early and late demos, SID tunes, etc. Some of the 
early ones were still great. For its time.

> If this is indeed the case, one could examine the code of SpeedDOS
> and try to improve it. Then the thought crossed my mind: what about
> replacing the part that handles the serial transmission of bytes in
> JD by a parallel transmission? IMHO JD could become even faster!
>
> OK, and what has IEEE to do with this? The idea was NOT using the
> userport for the parallel part but an IEEE cartridge.

Interesting. But it is probably more efficient to patch e. g. 
DolphinDOS, which is quite well optimised and uses parallel transmission 
already. Or Prologic, ...

> Nice touch: this can be realised by taking the JD Kernal (which on
> itself is also an overlay (but to be done yet)) as base and creating
> a overlay that contains the parallel routines.

I think the same applies for Dolphin. But JD is already there. It still 
would require probably much bigger patch/overlay to be used effectively 
as parallel speeder. And remember that you have to patch also the 
drive-side firmware.

-- 
SD!

       Message was sent through the cbm-hackers mailing list
Received on 2015-01-24 22:00:42

Archive generated by hypermail 2.2.0.