Info-Atari8@SCORE.STANFORD.EDU (Info-Atari8 Digest) (11/27/86)
Info-Atari8 Digest Wednesday, November 26, 1986 Volume 86 : Issue 8
This weeks Editor: Bill Westfield
Today's Topics:
Future of Atari
Print Shop Companion for sale
----------------------------------------------------------------------
Date: Wed, 26 Nov 86 02:25:41 EST
From: Terry Conklin <conklin%cps.mich-state.csnet@RELAY.CS.NET>
To: info-atari8@SU-SCORE.ARPA
Subject: Future of Atari
This is to all 8 bit fans, but especially Atari corp., who have the
power of standardization.
The Atari 8-bit machine, now look upon in it's golden years, is a
technological masterpiece. Consider the ideas of co-processing, device
independence, human engineering and like that went into it, and it's
terribly impressive.
Though it was ahead of it's time, perhaps, we the users weren't, and
over the years technology as a whole has made much progress. A great
deal of this progress has been made with the IBM compatibles. Given
enough monetary impetus, they proved what miracles could be wrought
out of a backward 8088. Now that we have that technology, though, we
seem to be missing the opportunity to apply that our 8 bit hardware,
which is still elegant, if not up to date. With this in mind, I submit
the following list of directions which, together or seperate, would be
feasible to introduce, could long extend the lives of 8 bit machines,
revitalize markets and ease the transition from the home computer
environment to the multiprocessing environment, where the potential
consumer has his 8-bit, AND his ST both running productive tasks.
Key to introduction of any proposals at this point in the age of
the machine is support, either verbal, developmental, or preferably
end market items from Atari itself. While it is clear that the current
management feels no loyalty to the "long" history of Atari buyers,
perhaps there is enough engineering pride left to help. Without
further ado, the following ideas are public domain, free for the
taking for the greater life of Atari:
A -standard- large memory model. If Atari would consolidate the
extended 130 XE architecture with Claus popular 256k upgrade (and then
the 512k, 1 meg, and more) to provide a stabilized base for paged
program development. While the 6502 can be infuriating, a possible way
to implment this painlessly would be to define an M: memory managment
device driver. Because of the popularity of ramdisks, some form of
memory management would be needed in order to allow more than one
program access to an unknown and unspecified amount of memory. A set
of options and requests for memory would be something like malloc(),
but could be much more sophisticated. For example, a sample table of
M: requests,
M: arg function
1 Return number of free pages in system
2 Return total number of memory pages
3 Request N pages, returns page index
4 Return pages -> to free space map
5 Move page from loc. to loc.
6 Return allocation map
Further, processes requesting memory could allow for a 'priority'
level with the request. A process that prefers to have free memory but
can give it up to an application that demands can request all free
pages with a 'low priority' (For example, a RAM disk that can, if
necessarry, go to real disk space) and then the user's main program (a
word processor) can demand that memory. In a memory conflict, in this
situation, the process that requested the memory would be asked to
reliquish the memory to the higher priority request. The ramdisk would
then either "do something" with it's data (write to phys. disk) or
deny the hand over request, upon which that extended memory is not
available.
The most important function of accessing memory through a device
would be that any memory extention at all would be compatible. Exact
hardware descriptions of how to access extended memory would be
removed from the application program, and thus other devices could in
fact simulate M:, even a hard drive could "act" as memory. This
implies that M: would act as more than a mere manager, but would also
be called to access the differemt memory pages. Whether or not M:
would simply map memory (in arbitrary size chunks, say 1k) or be the
actual control program for alternate pages depends on the success of
Atari promoting a unified memory interface, though it is clear the
latter approach allows more flexibility and is closer to the machine's
design.
Given large memory models, the machines second most natural growth
path would be an adaptation of the Western Digital 16bit version of the 6502
or simply the ability to run the central processor at higher speeds.
Processing power could be tripled without causing undue engineering hurdles.
Another area only Atari engineering would know the true
feasibility of would be direct extentions of the Atari's custom chip
set. A great deal of expansion room was left on various memory
locations. For example, 5 bits were originally left free at GRACTL. 1
bit could be used to differentiate between old and extended-GTIA
(GTIA+) modes. Given the chip size of GTIA in comparison with today's
manufacturing sizes, a simple extention (ha) such as simply doubling
the data acquisition rate, using exactly twice as much memory to get
twice to output (such as a new 320 X 400 high res mode, 4 color 320 X
200, and 256 color GTIA modes.) could be compatibly added to the OS
with a bit extention. Since the popular memory model now allows for
16k paging, video memory would no longer have to reside in the memory
map proper, giving programs even more room to run. A 'page' register
added to ANTIC could make displaying the page automatic, or
conversely, a simple modification to vblank to handle the swap.
As yet another request, a card cage. Bring back the 1090 (was it?)
Hard drives, memory, 80 column boards, and ICD's -great- releases as
of late would have been so much simpler if people were only allowed to
buy 1090's. What would it cost to manufacture a card edge XM301, I
wonder? I would also add to that request that the card cage be made XL
bus compatible, since I have yet to see new Atari come out with
anything along the lines of the 1450XLD, which a card cage XL bus
could support, and that XE's most probably could as well access with
an adapter.
And last but not least, private progamming groups, ICD, OSS, and
others have soundly proved the feasibility of the bank select cart.
But that could just as well paved the way for a bank select OS,
extended, but still compatible because of the defined entry points.
With but a few of these extentions, the Atari XS (Xtended Super) could
for less than say $200 easily command the market of the Apple II GS,
and be more compatible with it's family line to boot.
Atari has 3 generations of programmers now with a world of
experience. All they need is but a few simple tools like those their
PC cousins have. Perhaps extending the depth of the 8 bit line (which
could be a 16 bit as well) is not the highest profit line Atari could
touch, but it builds corporate character, which is what TT sales are
built upon.
For anyone interested in working on a public domain specification
for these extentions, I have created a conference to work on them
without cluttering the net on the Club. US (517) 372-3131 3/12.
Terry Conklin
Club Net Coordinator.
A last request, (shotgun approach, ask Santa for the world, you
might at least get the cookie...) would Atari please update and sell
De Re Atari again? Or the operating system manuals? Or the lovely in
depth documentation of yesteryear?
------------------------------
Date: Wed, 26 Nov 86 8:17:55 EST
From: Alex McKenzie <mckenzie@j.bbn.com>
To: info-atari8@su-score.ARPA
Subject: Print Shop Companion for sale
I just bought "Print Shop Companion" for my Atari 800. After opening the
shrink wrap (thus making it unreturnable) I discovered it requires 64K, and I
only have 48K. So if you want a copy of Atari Print Shop Companion in brand
new condition, make me an offer. It requires:
disk
64K Atari (XL, or perhaps an expanded 800?)
Print Shop (from Broderbund)
It provides:
additional fonts
additional graphics
additional borders
additional editing tools for everything
I paid:
$23
Alex McKenzie
mckenzie@j.bbn.com
------------------------------
End of Info-Atari8 Digest
**************************
-------