MARK@ISIS.MIT.EDU (Mark Kriegsman / VAX Resource Center) (02/28/88)
The bitplanes/graphics coproc. ideas are good, but the thing that has let the
][ family stick around for more than ten year is *flexibility*:
o No other machine was created as EXPANDABLE as the //. I suspect the Mac II
will live some kind of along and prosperous life due to it's
deliberately expandable design.
o No other machine has had such a diverse array of applications and techniques
developed for it. In a sense, the -lack- of a _specific_ graphics
copr. (with specific abilities and limitations) has caused the
development community to constantly explore new ideas **IN SOFTWARE**.
Software is pretty much cheaper than hardware, in some sense. Starting
with a generalized machine, like the //s, anything can be implimented.
o The //s have traditionally been like the "old style" Lego sets: many pieces
in a few general shapes that allowed more diverse creation. By
selecting new features to add in HARDWARE, the //s would become like
the "new style" Lego sets: a smaller number of each kind of piece, with
many specialized pieces (like a certain axle bearing, or a certain
bevel block that only fits in certain places).
I have always thought of the //s as generic Turing machines. They were
composed of simple parts that could be combined IN SOFTWARE to accomplish
any given task.
Yes, display hardware can be a limiting factor, and I would like a
palette-driven high resolution polychromatic display for the //s.
Yes, the dedicated sound RAM can limit applications. (look harder at the
word dedicated, and think about the Lego)
My dream // is not "this graphics copr, and this sound copr, and this
I/O copr" but more like:
a faster, bigger (RAM), BUT STILL GENERALIZED machine:
How about a 65816 (or 65832! Anyone heard more rumors about this, other than
the reference to it in the documentation for the '816?) running at some
large number of MHz, like 20. A nifty little pipelined processor with
a small instruction set (think Lego again) is a fine idea. Furthermore,
it could still run "old" '02 code!
And how about letting this machine have all it's RAM addressable, like
a true 24 or 32 bit bus? (this depends more on what the '832 wants.)
I'd like a big, generic bitmapped display, with a bunch of colors and a
color-lookup table.
Yes, it's horrible watching the GS refresh the desktop. But if it were
simply FASTER, that would be great. Adding hardware specific to the task
limits the tasks that can use it. (want to try and reprogam the "DESKTOP
GRAPHICS" chip to do shaded 3D z-buffered graphics?)
The flexibility MUST be left in if the // line is to live up to
"Apple II Forever!" I actually think that the GS might have strayed from
the philosophy that carried the //s to fame and glory, and I'd rather
see a "Super //" than a GS+.
Let the software figure out the details. Remember what the graphics
looked like on the //s in 1977? Notice how much better they are today
ON THE SAME HARDWARE? No had to upgrade their hardware to change from
playing "Star Cruiser" to playing "Zaxxon" or "Hadron" or "WayOut" or
"{name your favorite great-looking graphics program}". I'd still
rather market software that can say "Runs on any Apple // with 48K."
When the upper limit of what the hardware can support is ACTUALLY reached,
then it's time to come out with a new generation of hardware. But
that new hardware should have no ceiling in sight, because it comes up
faster than you think.
I cannot say how much I hope the next // machine is more generic, and more
powerful. But I do hope that. Any thoughts? Apple are you listening?
Let's not make the next Apple // an Amiga, ok?
-Mark Kriegsman
Mark%ISIS@Athena.MIT.EDU (internet)
Mark%ISIS%Athena.MIT.EDU@INTERBIT (bitnet)
PS. Anyone want me to post the program I wrote that gives TRUE 80 column
display (560 horiz points) on an unmodified ][ / ][+ ? How about the
one that lets you control the volume of the speaker-that-has-no-
volume-control on any ][? How about the "bouncing ball" demo (like the
Amiga's 'boing') that runs on any ][? How about the one that lets
you draw graphics in 16 colors with no orange-next-to-green style
restrictions, and supports a color lookup table that also runs on
and ][ that has had the "color-mod" (remember that one, folks?) ?
PPS. Get the idea?mdelama1@ORION.CF.UCI.EDU (Michael De La Maza) (02/29/88)
I would like you to post the program that gives true 80 column display,
the program that gives you speaker control, the bouncing ball demo, etc.
Better yet, is there any way I can pay you to mail them to me on
a disk???
-- Michael de la Mazajm7e+@ANDREW.CMU.EDU ("Jeremy G. Mereness") (02/29/88)
the way to improve the GS is simply to adjust the
limitations that people discover hampering its potential.
I feel it should have....
-512K RAM standard
-support for the full 16 Megs (not 8, c'mon! That limits the future!)
-256+K ROM standard (put those damned tools in ROM! not on disk)
-128+K sound chip RAM
-stereo outputs for the Ensoniq (no excuse; this takes no effort at all)
-a SCSI port built in (I mean, why in hell not?)
-either two SHR screens, that can be combined,
like double-hires is a combo of main and aux hires pages
(this is to enable the possible use of page flipping)
OR
Virtual Screen Memory, which means the SHR screen
can look at any 32-64K bank of memory as color graphics
-a faster 65816, like 7.5 MHz or so. Selectable speeds.
a graphics co-processor would be nice, or better yet
a Quick Draw processor, but that gets close to hardwiring
again, and flexibility is lost. If there was some way to make
such a thing software transparent, like shadowing is, then maybe.
I feel if the desktop refreshes 3 times faster, it will be no worse than
a Mac 512, which is what we compare these routines to. (and no, this does
NOT mean I am only looking for a cheap Mac //)
-a faster-loading, more efficient, Prodos 16 and Loader!!!!!
-GS Basic interpreter bundled STANDARD!!! (for God's sake, the Mac has
Hypercard bundled, and part of the ][ is the fact that you can access
everything right out of the box. We had to LEARN to be
dissatisfied with Applesoft, remember?
-an Appletalk Toolkit (like the old DOS toolkit) containing LQ and
Laser support routines, and Mac-Prodos/Prodos-Mac conversion
utilities (Apple put its foot in this one; if you're going to support
Appletalk, support it 100%!!)
None of this is really that difficult or too much to ask foror too expensive
to implement. None of this did not come simply from observations
of how the present GS operates. The capabilities
of the 65816 scream for more memory and more speed. It is a good chip;
let it do its job. Let the software gurus take it from there, like a Switcher
that can go between Prodos 8 and 16 would be a godsend.
I hope Apple does not make the mistake of incorporating all devices
as virtual memory like in the Mac. It is ridiculous that the Mac w/ a Meg
keeps looking at the disk for system routines and such (hard disks are
not a good solution to this, either).
Finally, although some people despise how the reduced instruction set
of the GS is a s.o.b. compared to that of the 68000, the 65816
is more flexible, and therefore has more surprises; more potential.
And as PS/2's and Mac //s move the technology toward hardwired,
institutional, $5000 BIG-business machines, a wider niche opens for the GS,
a capable, flexible computer that has firm roots but modern capabilities
(a.k.a. windowing) with a decent price for the small business, the school,
and the inventive home users, who may not know WHAT they want
a computer to do
But...
bet your bottom buck, they'll think of something.
It's up to Apple to deliver; either that or the Amiga, which is
cheap and powerful as sin.
=============================================
Capt Albatross
jm7e+@andrew.cmu.edu
disclaimer: my opinions are subject to change utterly
as I run head first into further enlightenment
on whatever subject I may be going on about.
any kind of feedback is not only appreciated,
it'll probably make me less stupid in the long run.gwyn@brl-smoke.ARPA (Doug Gwyn ) (03/02/88)
In article <8802271745.AA00234@ATHENA.MIT.EDU> MARK@ISIS.MIT.EDU (Mark Kriegsman / VAX Resource Center) writes: >Yes, it's horrible watching the GS refresh the desktop. In case anyone isn't aware of it, the basic problem is that there is a lot more work involved in copying a color-pixel bitmap than in copying the same number of black-and-white (1-bit) pixels. Many of our notions about how the desktop environment should look were based on experience with 1-bit pixel Macintoshes (or Suns, Blits, etc.), which even with equal processor speeds would therefore have appeared much faster.
aehl@csd4.milw.wisc.edu (Donald J Aehl) (03/02/88)
In article <8802271745.AA00234@ATHENA.MIT.EDU> MARK@ISIS.MIT.EDU (Mark Kriegsman / VAX Resource Center) writes: > >PS. Anyone want me to post the program I wrote that gives TRUE 80 column > display (560 horiz points) on an unmodified ][ / ][+ ? How about the > one that lets you control the volume of the speaker-that-has-no- > volume-control on any ][? How about the "bouncing ball" demo (like the > Amiga's 'boing') that runs on any ][? How about the one that lets > you draw graphics in 16 colors with no orange-next-to-green style > restrictions, and supports a color lookup table that also runs on > and ][ that has had the "color-mod" (remember that one, folks?) ? > >PPS. Get the idea? Yes, I would like to see these programs posted! If you can't do this try to mail them to me I would really appreciate it. Thanx. ============================================================================== ARPA: aehl@csd4.milw.wisc.edu USMAIL: Donald Aehl Box #204 3400 N. Maryland Ave. Milwaukee, WI 53201 UUCP: ihnp4!uwmcsd1!uwm-evax!aehl ICBM: 43 4 58 N / 87 55 52 W ==============================================================================
mdelama1@ORION.CF.UCI.EDU (Michael De La Maza) (03/04/88)
Yes, I would like to see these great programs posted too (just like Donald Aehl). If you can't post 'em mail 'em!