Marcel Kilgus writes: > > If its so much bother implementing SMSQ/E on a faster (but different) > > processor, isnt it worthwhile revisiting the multi-processor option again? > > If you cant get faster processors, use more of them! > > Well, that's only worthwhile if you have work that can be done in > parallel. For 99% of all SMSQ/E installations I guess this is not the > case. > It might make sense if we had a display driver that can update buried > windows, in that case one CPU could do the screen updating and > probably other minor jobs. But the user jobs won't be much faster this > way either.
Existing programs arent designed to take advantage of multi-processor processing, and they probably dont need to either (Quill is fast enough ;) but they should be able to run in a multiprocessor environment without change. Im thinking of possible future applications. One of them could be what you suggest, but that is probably a major undertaking. Not less of an undertaking, but perhaps more pressing are things like a web browser (rendering graphics in parallel), media applications, networking and the like. Depending on available tools, or future ones, program granularity could be reduced allowing for more parallelism which would open up vistas for scientific and technical applications - and users. There is virtually no hope of proselysing among your average Home PC user. The technical and scientific community might find the QL environment more congenial (and we them). Of course, I can see that if one could muscle in on the embeded devices thing wed have a mass market, and this is something that should be explored by those few who have the potential to make a major commercial success of it. Thats great! Go for it! But there is a risk here that the thing would take off in a completely different direction. That large commercial entities might take control and develop in ways that dont benefit the QL community at all. So dont lets wreck what we have in the process by introducing years of bugs and instability in what now is a relatively stable, yet still evolving, system. I see the "embedded" and the multi-processor proposals as two separate issues. The former is a commercial venture which, of necessity, requires a separate forum. The latter is about the future of the QL. Please discuss! Per _______________________________________________ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
