NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/sys/amiga/programmer/920

Internet Message Format | 1996-08-05 | 4.9 KB

Path: informatik.tu-muenchen.de!fischerj
From: [e-mail] (Juergen "Rally" Fischer)
Newsgroups: comp.sys.amiga.programmer
Subject: Re: >>>>>>> The Future: Amiga goes PC (...aeh: "PowerMsPenti...?) <<<<<
Date: 12 Jan 1996 18:23:11 GMT
Organization: Technische Universitaet Muenchen, Germany
Distribution: world
Message-ID: <4d68uf$[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]>
NNTP-Posting-Host: hphalle8a.informatik.tu-muenchen.de
Originator: [e-mail]


In article <[e-mail]>, [e-mail] (Dave Haynie) writes:
|> Lines: 72
|> Reply-To: [e-mail] (Dave Haynie)
|> References: <[e-mail]> <[e-mail]>
|> Nntp-Posting-Host: gator
|> Organization: Scala Computer Television, US Research Center
|>
|> In <[e-mail]>, [e-mail] (Mats Magnusson) writes:
|> >>Yeah right. Akiko is not faster compared to modern c2p algorithms.
|>
|> >>> If there was an Akiko in every Amiga, this action could be done with
|> >>> 2 moves on a 040.
|>
|> Clearly, what you want in every Amiga is some form of corner-turn
|> memory, not an Akiko, which is a system glue chip, specific to the
|> CD32, that just happened to have a small chunk of corner turn memory
|> put in, although management didn't want it at the time. That was

They really better would have let the original CIAs, because clever
software can do c2p at about akiko speed with using blitter
(when fastmem present, akiko loses vs blitterc2p on 020!)

if CD32 had become a computer by just adding keyboard and disk drive,
maybe even with simm sockets for fastmem, WOW, CD32 would have
been a really atractive multimedia homecomputer! And you would have
maybe even saved money not having to design  new chips!!!

But I had to chose the A1200, I need a computer.

Adding CD to A1200 is expensive.
Making CD32 a computer is damn expensive.

argh, the cheapest way is buying them BOTH, something is really wrong!

I hope the CD version of the new A1200 comes true.

|> Hedley Davis' design; I had a different corner turn memory design in
|> the AAA system specs.

Well, how difficult ($$) would it be to have kind of "dma akiko",
which does both load from fast and store to chip independent
from cpu ? would give 256 color lores chunky in realtime.

|> That's not a "stupid thing", that was intentional. The pin count on
|> Akiko, as with most chips, is quite limited. Integrating the 8520 bits

mhm what is expensive about pins ? I never understood this...

|> that a CD32 required was the proper job -- that you can't as easily
|> turn the CD32 into a full Amiga may not make you happy, but the CD32
|> designers were determined to make the CD32 a viable system. Start
|> adding in stuff you don't need, the costs go up, and it's CDTV all
|> over again.

aaah CDTV, the most pretty C= computer. Who made CD32 look so ugly...
I deal with the thought to make the case of my A1200 black :)

|>
|> >which means
|> >that you still must have two standard 8520 also to get the paralell port,
|> >disk control signals and other such stuff...
|>
|> >Why didn't they implement c2p in denise,
|>
|> The corner turn conversion is CPU managed -- you don't want to have to
|> wait for chip bus register access for that. Not to mention that RGA

would corner turn in chipset be cheaper than a real chunky mode ? wondering...


|> It's not quite that simple; you still need some temporary
|> storage. Every chip bus graphics cycle, after all, fetches 8 bytes,

ah, BTW: would it have been very expensive to make the dma fetches
of 8 planes lore like 2 planes shres (which are better for cpu acess).

|> not one. But sure, I suspect some clumsy chunky mode could have been
|> managed with a few additions to Lisa. I really don't know if it had

what disadvantages ? chunky lores using dma of shres ? :)

|> been discussed; perhaps there was some feeling that the end result
|> would be just as bad as no chunky mode at all (ok, now, here's how
|> you draw a line....).
|>
|> You could play with this kind of thing in an AA system today, if you
|> really wanted to. Put Lisa in LUT-bypass mode, store up 8 planar
|> pixels, then shift 'em out in weird-chunky mode (basically, you're

would finescrolling still work ?

|> putting a registered corner-turn memory as a Lisa post-processor); you
|> would probably want a LUT or something after that, which would
|> complicate the experiment all the more.
|>
|> Dave Haynie          | ex-Commodore Engineering |   for DiskSalv 3 &
|> Sr. Systems Engineer |  Hardwired Media Company | "The Deathbed Vigil"
|> Scala Inc., US R&D   |    Ki No Kawa Aikido     |     [e-mail]
|>
|>          "Feeling ... Pretty ... Psyched" -R.E.M.
|>
------------------------------------------------------------------------
   [e-mail] (Juergen "Rally" Fischer)   =:)