Re: Did Commodore cheat with the quad density floppies?

Den Wed, 09 Jan 2019 00:41:11 +0100 skrev André Fachat <afachat@gmx.de>:
> Am 8. Januar 2019 23:08:17 schrieb Mia Magnusson <mia@plea.se>:
> > Den Tue, 8 Jan 2019 21:58:39 +0000 skrev smf <smf@null.net>:
> >> On 08/01/2019 21:32, Thom Cherryhomes wrote:
> >>> I do want to seriously bitch-slap IBM for doing such a horrible
> >>> job on the CGA card (no dual ported memory... _FOR A CARD THAT
> >>> COST HUNDREDS OF DOLLARS BY ITSELF_ )
> >>
> >> I think you underestimate how much 16k of ram cost them in 1981.
> >>
> >> Was dual ported memory even available back then?
> >
> > Dual ported memory is an expensive solution to what's usually an
> > incorrectly formulated problem.
> >
> > The solution would be to use fast enough memories to be able to do
> > as in 8-bit computers.
> 
> "Fast enough" would probably have been too expensive. Banking i.e. 
> increasing memory bus width as with the 8032 could have been a
> solution.

The irony is that if IBM had opted for an 8086 instead of an 8088 and
had made the expansion bus 16 bits wide, it would had come more natural
to equip the CGA card with a 16-bit bus internally and then it would
have had double the bandwith and this problem wouldn't had existed :)

As a comparison, the first Commodore product fast enough to display 80
cols with an 8-bit bus (and no wait state / "snow" problem) were the
CBM-II/B series and later the PET 8296, and those came out later than
the initial IBM PC.

At the time, 80 cols text mode and 640x200 graphics mode were better
than most of the competition.

Also many third party CGA cards solved this problem.

-- 
(\_/) Copy the bunny to your mails to help
(O.o) him achieve world domination.
(> <) Come join the dark side.
/_|_\ We have cookies.