Xorg@cup.portal.com (Peter Ted Szymonik) (11/13/89)
=========================================================================
(C) 1989 by Atari Corporation, GEnie, and the Atari Roundtables. May
be reprinted only with this notice intact. The Atari Roundtables on GEnie
are *official* information services of Atari Corporation.
To sign up for GEnie service, call (with modem) 800-638-8369. Upon
connection type HHH (RETURN after that). Wait for the U#= prompt.
Type XJM11877,GEnie and hit RETURN. The system will prompt you for your
information.
==========================================================================
<JEFF.W>
Thank you for attending this special RTC. Our guest is Dave Small,
of Gadgets by Small, makers of the Specter 128 and Spectre GCR Mac
emulators. During the RTC, you can queue up to ask questions by
typing in /RAI (raise your hand). Without further adieu, I'll ask
Dave to make some opening remarks and then we'll get to your questions
and comments. Welcome, Dave!
<DAVESMALL>
Thanks, Jeff! I appreciate you having us here on a special Tuesday
night conference, and missing Moonlighting just to see us. Quite a
sacrifice! Tonight I'd like to talk a little about the Spectre GCR,
and take any questions people have, then talk about a new project
Gadgets is considering. The Spectre GCR allows your Atari ST, simply
put, to become a plug and play Mac. It runs nearly all Mac software,
sometimes better than a Mac does, and has some very superior features.
It also works with Stacy, and provides, err, a "portable Mac platform",
shall we say. Jeff, would you like to split the CO into GCR and then
68030 project parts, or should I do the 68030 intro now as well?
<JEFF.W>
Dave, with that tease, how could I ask you to make these folks wait
for 030 info? <grin>
<DAVESMALL>
Okay; here's a 70-odd line intro. I'll "upload" it into the RT. If
you need to slow it down, just press CTRL-Q, then CTRL-S to restart.
(to pause this listing while reading it, press CTRL-S; to resume the
listing, press CTRL-Q).
Gadgets' 68030 Project
Last year in May, after leaving Data Pacific, there was a
GEnie conference on whether or not I should develop a 128K ROM
Mac emulator. Many people attended, said I should, and it was a key
reason we decided to do the Spectre 128 (and Spectre GCR) products.
Tonight I'd like to try the same thing again: propose a new
Gadgets project to you, and see if you're interested. We're already
working on this product for our own internal use; we want a hot,
fast ST that's competitive with the latest from NeXT and Apple. What
we need to decide is if we market it to the world.
Present ST Speed / Processor
Right now, your ST is designed around an 8 Mhz 68000. This is
a fine design, but, unfortunately, is starting to become dated.
Newer processors, which give you far more power, are out, and at
higher speeds; Apple's got the Mac II, a 68020 at 16 Mhz, and the
Mac IIci, a 68030 at 25 Mhz. The NeXT is at 25 Mhz. Ultimately all
that horsepower lets people write programs they couldn't write before.
68030
What we want to do is give present ST owners an *upgrade
path* to a very powerful new processor: the 68030. For many rather
technical reasons, a 68030 is much faster than the ST's present 68000,
and could give many new capabilities to the ST -- memory mapping,
memory protection for multi-tasking, better Spectre capabilities (our
best thing!), and so forth.
If we could get a 68030 board out for the ST, I'd bet the people
at Codehead wouldn't sleep for weeks, just thinking of new things
to do. The 68030 is very far ahead of the 68000 and opens up all
kinds of possibilities. But software compatibility ought to be very
high, since the 68030 is very compatible with 68000 software.
We will also "clock" the 68030 processor at a much higher speed
than the ST's 68000. The ST runs at 8 mhz (millions instructions /
second); we'll go for at least 16 Mhz and possibly more.
The primary intent of the 68030 board will be *speed*. While it
is difficult to estimate the speed improvement, we're confident of
at least 4 - 8 times speed increase.
Why not an exact figure? Well, it's the old drag-racing adage:
Speed costs money; how fast do you want to go?
For instance, we can use a 25 mhz 68030, the same processor as in the
new NeXT and Mac IIci, but it will cost a lot; a 16 mhz 68030 would be
much less expensive. We don't want to make a product so expensive that
ST owners can't afford it; we want this to be a *practical*,
*affordable* upgrade for ST owners. We want to keep the pricing as far
under $1000 as we can, probably with options to let a user expand later
as their budget can afford.
[Of course, I'll probably put a 25 Mhz 68030 in mine...]
Memory Limit
We also want to "break" the 4 megabyte memory limit. We see no reason
why the ST shouldn't be able to go to 12 megabytes with a bit of
work, and with a faster processor, there will be plenty of uses for
that memory. (If nothing else, the digitized music industry could
really use extra RAM to hold digital samples. But imagine a 10
megabyte Ramdisk...). This will also solve the primary problem of
why it's hard to accelerate the ST: video contention.
Compete with TT? Nope.
We aren't trying to compete with the Atari TT here, either. The
TT is a completely new machine, with multiple serial ports, VME
slots, new high res color modes, and so forth. For instance, the UNIX
option is definitely not intended for home users! What we are trying
to do is give your ST enough horsepower to extend its useful life some
years.
That's a brief, 83-line intro to the Gadgets 68030 project. Again,
we're certainly going to do one of these for us, to make code
development for Spectre/GCR much, much faster; we're interested if
anyone else out there wants a very fast, large memory ST.
-end-
Okay, that's the "brief" intro. Hope it came through ok!
<JEFF.W>
Here come the questions...
<[Bruce] B.HANSFORD>
Dave - this sounds great. Question - how will the speed of a 16mhz
68030 compare with an accelerator board like T16? ga
<DAVESMALL>
Well, it's a difficult question -- apples and oranges. Whups, I
shouldn't say "Apples". Anyway. The 030 has a different architecture
and generally runs faster, and does more, at 16 mhz than the 68000.
A real whiz like Jim Allen could tell you exactly why. Now the present
T-16 board ends up at around 12 Mhz in best case, because it's fighting
with video memory;
<J.ALLEN27>
Yes the 030 will kick butt!
<DAVESMALL>
(heh!)
<DAVESMALL>
The 68030 will run without fighting with memory, and is more efficient
to begin with. We are confident in the 4-8X speed improvement.
<J.ALLEN27>
But it isn't cheap...speed costs money.
<JEFF.W>
How much? GA
<DAVESMALL> Jeff -- it depends on the Mhz model of 68030. 16 mhz is ok,
25 is $$.
<J.ALLEN27>
And 33 is $$$$$.
<DAVESMALL>
So to try to answer the question, the T-16 is around 12 mhz, the 68030
in equivalent terms is probably a 20-25 mhz 68000 .. .Jim, think that's
fair?
<J.ALLEN27>
Yah, equal to a 25Mhz 68000.
<S.WHALEY>
Is it just a matter of simply putting in a new cpu or ...?
<DAVESMALL>
Well, it's a whole circuit board. It will take over the function of
the 68000 and delegate the 68000 to things like disk I/O. There's
several ways to do it: chop out the old 68K and replace it; plug into
Mega slot; piggyback 68000 -- and so forth. This isn't that major a
design consideration, actually. (Not exactly an answer, but it's the
truth right now.) .ga.
<[Jim] J.HARRISON4>
So it isn't a new motherboard right...
<DAVESMALL>
That's correct. We want to use the present ST keyboard, display, etc.
as is; we want to accelerate the main ST with the software you use.
<[Jim] J.HARRISON4>
it will be a plug in so we still have reverse compatibility right?
<DAVESMALL>
(Oh, yes -- you can always "go back" to 8 mhz.)
<J.ALLEN27>
REAL 8Mhz!!! 68000 8Mhz!!!
<DAVESMALL>
[By the way, if you're asking a question, do feel free to let me know
if you're interested in such a board -- more or less the point of the
CO. And suggestions are welcome.]
<[Jim] J.HARRISON4>
Sounds good, something will always be incompatible and I hate to lose
the ability to run some stuff. I have everything you ever built and I
don't see any reason to stop now, good show. That's all for me .ga.
<DAVESMALL>
Me too. It's important to stay compatible -- hence the emphasis on that
with GCR.
Thanks, Jim! Okay, 1 yes, 0 no. We're winning.
<K.BAD>
2 parter: first, I think this'd be a great hack *if* you could pull it
off. But I see a lot of problems. First of all, what TOS is it gonna
run? Do you plan on licensing TOS from Atari?
<DAVESMALL>
next question? (grin)
Suppose we could just leave it in Spectre mode.
No, seriously, when there's a will, there's a way. The old 192 K ROMs
may not appreciate the 68030, but the Mac ROMs didn't appreciate the
ST, either, and golly, they work.
<K.BAD>
Heh.. that'd work <grin> Okay, second question: My understanding
right now is that fast cache memory would be more limiting than the
price of the actual CPU, and you're gonna want a lot of cache memory
on this thing for it to work effectively with ST RAM right? ga
<DAVESMALL>
Well, ST RAM is effectively video shifter limited. So we either kick
the shifter onto a dedicated 32K static RAM, and free it up for 16 Mhz
use, or just blow it off. With RAM prices toppling faster than Ken's
UNIX hairdo, we feel that leaving RAM up to the user isn't a big
problem. Buy as much as you want.
<K.BAD>
Limiting the video memory will cause you a _lot_ of compatibility
problems, but I think you should be able to figure out a way
to work it out. Well, best of luck!
<J.ALLEN27>
You mean like the Moniterm Ken?
<K.BAD>
Exactly, Jim.
<DAVESMALL>
Thanks, Ken. We got your resume, and will look at it tomorrow.
<J.ALLEN27>
Not if you keep a 640X400 res!!!! Just point at the static ram!!!
<DAVESMALL>
Righto -- the static can be mapped anywhere.
<JEFF.W>
Before we continue, all of you may have noticed that Dave isn't alone
here. Jim Allen is the maker of Turbo 16 accelerator. Dave, would you
care to explain Jim's presence here?
<DAVESMALL>
Oh, Jim's the leading 68030 guru I know of, and is a good tech advisor
so I thought it might be prudent to have him along.
<J.ALLEN27>
Dave is the SW guru!!!
<DAVESMALL>
Also, Doug Wheeler is online, too.
<DOUG.W>
Hi.
<J.ALLEN27>
The HW is not a big deal...this is a SW project!!!
<DAVESMALL>
Anywho, it's entirely possible that Jim might have something to do
with this board .. <hint, hint, grin>
<[John] JJKENNEDY>
Jim, will there be an upgrade path with the T-16
<DAVESMALL> Too early to tell...
<J.ALLEN27>
I'm sure we can work something out!!!
<DAVESMALL>
Agreed. Next!
<DOUG.W>
$10 trade-in?
<DAVESMALL>
(isn't this nice.. not ONE "where's my GCR" question!!)
<J.ALLEN27>
A zero missing I think...
<T.REYES>
HOW ABOUT DOING SOMETHING WITH A 68881 WHICH WOULD BE CHEAPER
AS YOUR START ARTICLE STATES 10-30X SPEED UP IS POSSIBLE
<DAVESMALL>
Very few ST applications use the 68881 floating point unit. The
software must be written with the 68881 in mind, and as far as I
know, DynaCadd is the only one that does it. Otherwise the 68881 is
just a lump.. does nothing.
<J.ALLEN27> Chicken and egg.
<DAVESMALL> Does that ansdwer your question?
<T.REYES>
COULDN'T A TSR REDIRECT SOME FP TO THE 68881 AND AWAY FROM THE 68
<DOUG.W>
68881 is only useful for floating point math, not general operations.
<T.REYES>
YES BUT MUCH OF WHAT ONE WANTS FROM A 030 FOR SPEED INVOLVES
GRAPHICS MANIPULATION INCLUDING THE MATH
<DAVESMALL>
Not really ... mostly programs do the math inside themselves, and
don't even bother to ask if a 68881 is there to help them.
<J.ALLEN27>
The ST doesn't have a "system" math library that can be FIXED like
the Mac.
<DAVESMALL>
I agree. So we give *present* software a major acceleration, and
let the software writers fix it for 68881 whenever they get around to
it.
<J.ALLEN27> Mac SW will imediately use the 881/882 in Spectre mode on the 030
board <grin>!!!
<[fred beckman] FB>
I Want one and I know where my GCR is!
<DAVESMALL>
Okay, that makes it 2:0 with Ken abstaining. (grin)
<J.ALLEN27>
He'll bite...eventually <grin>
<[fred beckman] FB>
The 030 sounds like a great enhancement to the ST line.
<DAVESMALL>
Thanks, Fred. We want to give people a good upgrade path for as
little as we can afford to.
<K.FORD1>
Any ideas what effects the 68030 would have on PC-Ditto II or GCR?
<DAVESMALL>
I don't know on PCD-II, since I haven't seen one. The GCR, it would
make scream -- very very fast Mac emulation. Excuse me, I'm
slobbering on the keyboard.
<J.ALLEN27>
HW emulators will be uneffected.
<[Min] MINSYEE>
Dave, I want one... (Hello, Businessland: cancel that order for the
IIci :-)
<DAVESMALL>
hehehehehe!
<[Min] MINSYEE>
Couple of things, Dave. First, there's no doubt in my mind that
you're the person to do it if it can be done...
<DAVESMALL>
(oh, we're already doing it. Not to worry.)
<J.ALLEN27>
Secrets...
<[Min] MINSYEE>
...just wonder about fitting that in with the other demands of
Spectre development. Other thing is: the installation?
<DAVESMALL>
Second question first.
I think easiest "installation" of course is a Mega-plug-in, but
that locks out 520/1040 owners. That's just a push-in-board. A
T-16'ish replace 68000 installation isn't all that impractical. But
really, there are so many ways to hook into the 68K bus, it isn't
that important.
<DAVESMALL>
First question: It fits in nicely with Spectre development. We
started out witrh the idea to get a faster machine to speed up
assemblies, and other Spectre development (assembling Spectre takes
*forever*!). It's a natural ouytgrowth of that.
<[Min] MINSYEE>
A big part of the Spectre success is that it's a plug-in.
<DAVESMALL>
Remember, we're more discussing the idea here, not what pin connects
to what pin -- whether or not a high horsepower ST is something
that's wanted.
<[Eric] MAJOR-HAVOC>
Could this board also have a SCSI port? My vote: YES, do it!
<DAVESMALL>
Hmmmm. The NCR 8530 supports SCSI pretty directly. But with ACSI,
is there a real need for it? Atari's DMA is mighty, mighty fast.
Jim --opinion?
<J.ALLEN27>
I think the ASCI is good enough!
<[Eric] MAJOR-HAVOC>
ok thanks, hope it all goes well! ga
<DAVESMALL>
Certainly an interesting idea, though -- would help us in Mac
emulation in GCR.
<J.ALLEN27>
Some things are tougher than others
<DAVESMALL>
(Is that a vote? It's 3:0, 1 undecided, 1 unrepentent)
<J.ALLEN27>
Mac compatibility would drive it.
<M.SAVU1>
dave would this 030 board be linked to spectre so you could run
programs that would require the 030 like soft PC and some versions of
excel and I vote positively YES!! to sell it!
<DAVESMALL>
Yes, it would, and thank you! The 128K ROMs support the 030 directly,
with a little hacking ...softPC should run without any trouble, but
you never know until you try it. Besides, pc-ditto-II should be
pretty cool.
<J.ALLEN27>
And you could use things like Virtual memory programs!!
<DAVESMALL>
You bet. Jim brings up an interesting point:
The 68030 is so powerful it can support many different machines *at
once* in the same memory architectyure (if you have enough memory).
So, with 68030 helping us in hardware, it would be easy to bring up
an ibm/Mac/ST that would be a keyflip away between the machines ..
or, multiple STs, where if one ST crashes, the others keep right on
going. The 68030 is really an amazing chip. .ga.
<J.ALLEN27>
That's for sure!!!
<DAVESMALL>
That makes it 4:0:1
<D.HAEFNER>
OK. The 68030 resides with the 68000. How about running Dave, how
about MacII like graphics for DTP, CAD
<DAVESMALL>
Oh, I see. The 256K ROMs are required to support color as in Mac II,
and they're not available on the market for sale, so I can't "do"
them. (Or did I misguess the question?) .ga.
<D.HAEFNER>
No, you got the question. Just looking for better graphics that the
68030 might support...
<DAVESMALL>
Primarily, the speed improvement will make a big difference.
However, I will say we're thinking about just-for-fun tossing in one of thos
e ultra-high-powered TI graphics chips ... I mean, if we have 5 meg
of RAM, a 68030, and so on, why not add a video output? That's very
undecided, though.
<[Dave Troy] TOAD-SERV.>
Dave: three things - 1) YES I want one - can we trade for Toadfile 44
carts? ;-) 2) I got all my GCRs today! ;-) (again) and 3) This isn't
my area, but how are you going to implement memory, in terms of
installation & managment, how will it fit (physically) and what will
be the fate of the old MMU in light of the built in MMU in the 030?
<DAVESMALL>
The new RAM will be on the 68030 board. There will be an interface,
some way, into the ST's old memory, provided we don't decide to shut
down the shifter and use it directly at 16 mhz. The 68030 handles
most of its own memory management. For right off, I'd give it a
straight memory map, but that's all programmable, for all the hackers
out there, to do all sorts of fun stuff! Does that answer the
question?
<[Dave Troy] TOAD-SERV.>
Yes - Dave it sounds funky! I am sure you can do it. Count me in for
a few dozen.
<DAVESMALL>
Okay, 4+a few dozen:0:unrepentanty holdout.
<D.HODSON>
Any connection with HyperWeb here?
<DAVESMALL>
Doug: Noper. HyperWeb is a DOS project -- software.
<D.HODSON>
Lots go'in on?
<DAVESMALL>
oh yes .. plenty going on..
<D.HODSON>
Availability?
<DAVESMALL> HyperWeb -- no availability announcement for some time, I
would expect.
<R.GHEESLING>
Dave, how about "Spectre GCR030, The Premier Macintosh IIci Emulator"?
<DAVESMALL>
heheheheheheh! You've just rolled Sandy on the floor. Mind if we
steal that?
<R.GHEESLING>
Nope, you can have it. Any Idea of time frame? ga
<DAVESMALL>
Time frame...hmmm. Well, it's a bitch to get something to market.
However, I'm told that the elves at Santaland are busy with CAD
programs laying out PCB's, so, if people are interested, we'll put
more resources into the project.
<D.BECKEMEYER>
OK, I have a few Ques. I'll try to be quick.
<DAVESMALL>
Go for it, Dave. I'll try to interrupt <grin>
<D.BECKEMEYER>
Back to which TOS? You said a hac of old TOS?
<DAVESMALL>
There's several ways there, dependent partly (but not completely) on
Atari.
<D.BECKEMEYER>
It would have to handle the problem of different type of stack frames
on traps.
<DAVESMALL>
That's easy enough..(Apple does it in their MacTrap handler)
<D.BECKEMEYER>
Are you thinking about writing a TOS "emulator" to handle "bad"
programs (i.e. the real cheaters)?
<J.ALLEN27>
Atari should do that for the TT!!!
<DAVESMALL>
Oh, heck, we might just kick the old 68000 back on and let it run that
way, if necessary. I presume we'll break a few programs, but a vast
number will work, and that's the point. In the Mac world, we broke a
few badly written programs, but all the important stuff worked.
<D.BECKEMEYER>
You said that there was RAM on the 030 board, 32-bit right?
<J.ALLEN27>
The ram will be 32 bit and might even use a second MMU chip <grin>
<DAVESMALL>
I believe otherwise the 68030 loses speed.
<D.BECKEMEYER>
How about a VM RTX? Can you say Demand paged RTX. AHH OOOO NOOO!
<DAVESMALL>
"Demand Paged RTX". ?
<D.BECKEMEYER>
What do you think of the idea?
<DAVESMALL>
Very interesting. Have to admit I hadn't thought of it. Let's talk
later.
<DAVESMALL>
("VERY interesting idea.)
<J.ALLEN27>
Yep.
<DAVESMALL>
Kinda makes your finger tingle, doesn't it Jim?
<J.ALLEN27>
A pretty good reception.
<P.ATKOCIUS>
Will mac 030 software compatibility be a problem since apple at 512K
ROMS?
<DAVESMALL>
No. The 128K ROMs are 030 compatible.
<J.ALLEN27>
No
<DAVESMALL>
I'm sure a few dumb PD programs will break, but big deal ... all the
major stuff, and most of the minor stuff, will work.
<J.ALLEN27>
Lots of 030 accelerators for Macs
<P.ATKOCIUS>
on another topic has GCR Problem with GTS-100 drives been solved
<DAVESMALL>
We've found that many GTS-100 drives have problems reading from track
65-79. We have a GTS-100 that we bought here in town and have it
spread out here in front of me, with a 'scope hooked up to it, probing
its innards. Doug is handling anesthedia, I believe. Anywho, nothing
solid yet -- we did get an error-free session, but it didn't repeat
nicely. We do have a good idea what's wrong, though.
<R.PERRY>
Dave, I want one, too! And since you mentioned it a few minutes ago,
where IS my GCR (didn't want to dissappoint you!)?..
<DAVESMALL> What's your zipcode, please?
<R.PERRY>
Seriously, I'm told it's on its way, but I'm getting impatient
<DAVESMALL>
Need your zipcode..
<R.PERRY>
02889
<DAVESMALL>
Okay, Sandy is looking it up. Next question while she does?
<R.PERRY>
nothing else from me, thanks.
<DAVESMALL>
Ronald Perry, right? It weas shipped last Tuesday, the
31st. Heck, you should have it or darn close..
<DAVESMALL>
okay, next question?
<B.MITTELMAN>
Dave, can you do anything to get output to a projection TV so I can
demo the GCR at Mac meetings ?
<DAVESMALL>
Hmmmmm. That's beyond my knowledge. I know at Comdex last year
they had a big screen hooked to monochrome; you could ask Frank Foster at Atari
how they did it. That's all I know, though.
<J.ALLEN27>
An overhead project hookup is around but costs $$$.
<B.MITTELMAN>
OK save a 68030 for me.
<DAVESMALL>
Of course, there's always the Moniterm. Okay, that makes it several
dozen:0:Ken (grin). Thanks! .ga.
<J.ALLEN27>
Hahahaha
<B.HILL2>
I've heard that the Spectre won't work with programs that use Midi/SCSI
to talk to musical instruments (i'm thinking of Alchemy for ex.).
Will the new board allow this? g.a.
<DAVESMALL>
It is true that Spectre won't work with only the MIDI Input-Output
portions of various Mac MIDI programs (like Opcode Systems, etc,
et all). It is possible that the mapper function of the 68030 could
fix that. Interesting, interesting idea! HmmMMMMmmmm. Thank you.
<B.HILL2>
Follow up - Consider using a Stacy with an 030 running this s/w.
Portable MAC for music purposed s but much more reasonable. ga.
<DAVESMALL>
Sure be nice. I dunno, though; many musicians tell me the ST is far
superior in music software (but those discussions are always sort of
like religious discussions). Still, a darn-fast-Stacy wouldn't hurt,
either -- particularly if the faster MIDI implementations get
standardized.
<DAVESMALL>
Speaking of Spectre, we had a mild event this week --
Microsoft Word 3.02, my old old enemy, came up this week and worked.
.. I fought that $%^#@$%^ thing for 1 1/2 years, beat it this week.
<J.ALLEN27>
Mr. Wizard strikes again!!!
<DAVESMALL>
The program was reading "from" 0, not writing to it. This also
appears to have greatly styabilized Multifinder. Jim's Zax found Yet
Another Bug... anyway, back to the "0-3-0-CO".
<[Norm] N.RECHTMAN1>
Dave, back to the drive question
<DAVESMALL>
Hi Norm!
<[Norm] N.RECHTMAN1>
Are there other manufacturers drives having problems? I know John
Wyncott is having a strange problem with a diverse data drive, MFD 720.
I was wondering if the TEAC mechanism was stable?
<DAVESMALL>
Not consistently. There are always onesy-twosy problems -- bad drive,
alignment problems -- but the two consistent ones we see are RFI
problems inside the Mega case, curable by shielding, and the GTS-100
drive (an NEC drive), which we haven't solved -- but we've only had
the drive a few days, give us a little time. I've been repeatedly
assured that these drives absolutely *can* read/write Mac data if
something is not interfering with them.
<[Norm] N.RECHTMAN1>
One other thing, it seems that a few of us are having problems copying
disks from the Spectre menu like Spectre to mac or even the same type.
<DAVESMALL>
Hmmmm. I'll look in the Gadgets RT. Always possible there's a bug in
the code. (Does work okay here in testing, so it might be a bad copy
of Spectre, or might be some glitch we didn't figure on. Please post a
note in the Gadgets RT summarizing what happens, okay?)
<STEPHANIE.A>
What is '**CAUTION RESTORE failed in XRESTORE**' ??? '030 board, YES!!!
<DAVESMALL>
Well, that's a goof of mine. Inside the Spectre GCR software, there
is an internal self-test routine, only meant for us to use. At the
last second, we decided to let users look at it. It had some bugs,
though --- for instance, if you run it with one drive, it hangs,
repeating your message. (And other cases). So we released a New, Much
Better tester for the GCR, that Doug wrote; it's in the Gadgets
Library. (Please use it instead!!)
<STEPHANIE.A>
I am having problems ejecting disks, am I having trouble WRITING to
them??
<DAVESMALL>
Are you using Finder 5.3?
<STEPHANIE.A>
yes I am using single 1040 drive system/color. I believe that it is
5.4, I just got this 3 hours ago!!
<DAVESMALL>
Try "dragging" the disk icon to the trash to do ejects. Those finders
always were twitchy.
<STEPHANIE.A>
I also have 6.0.3 but so far it has not been able to boot/ I did the
drag, no luck
<DAVESMALL>
That should always work; we do that here. 6.0.3. -- you probably need
boot sectors.
<STEPHANIE.A>
6.0.3 came right off mac II
<DAVESMALL>
Tell ya what, this is best left for the Gadgets RT -- lots of people
there that are new owners, too. (Right -- did you run Installer to get
it?) See, a straight Mac II System isn't right for us -- it's
configged wrong.
<STEPHANIE.A>
Ok, I just stopped here first, I was surprised to find you here
tonight, THANKS
<DAVESMALL>
Anywho, drop by the Gadgets RT and see if they can't help -- you're
welcome!
<R.GHEESLING>
Dave/Sandy, FYI, my GCR came today! Up and running in 5 min, using
Toshiba 3.5" drive. One more vote for new project!
<DAVESMALL> good to hear! thanks!
<[John] JJKENNEDY>
If this project is a go, who will market it? GBS, Fast-Tech,
both or.....ga.
<DAVESMALL>
Probably some sort of co-operative venture. It's more important to us
hardware/software hackers to define the project first, then set up the
legalisms.
<[John] JJKENNEDY>
Thanks.....And YES!!
<DAVESMALL>
Gadgets certainly wants it, though, and with some of the income from
the GCR, I want to plow it back into the ST community and do good
things.
<J.ALLEN27>
Exactly, we need the combined expertise to pull it off.
<DAVESMALL>
Not to mention customers (grin)
<[Wilton] W.VARGAS2>
Night everybody, I just got here so excuse me if I am asking something
that somebody asked before but, any plans for a MAC II emulator for
the TT or the ATW? .ga.
<DAVESMALL>
Nope. The ROMS for it aren't freely being sold -- first requirement.
<[Norm] N.RECHTMAN1>
Why doesn't someone make a Memory upgrade for the ST that uses SIMMS,
???
<DAVESMALL>
Probably space considerations. The ST isn't very tall.
<J.ALLEN27>
Too big a form factor.
<DAVESMALL>
And it's too big a form factor.
<J.ALLEN27>
Hehehe
<[Norm] N.RECHTMAN1>
Oh well
<DAVESMALL>
We'll probably use sockets for RAM as well, although that is not
certain certain.
<[Eric] MAJOR-HAVOC>
How long before Gadget Labs have a working 68030 board? .ga.
<DAVESMALL>
Remember, all, this is NOT a promise to do this project and market it;
this is market research. We don't know yet. (I shudder when I hear
the word "working". Getting boards -- the process has started. Getting
them WORKING -- that's a whole 'nuther kettle of wax.)
<W.CHOW>
1. where's my gcr? area code 02135
<DAVESMALL>
okay, it's being looked up
<W.CHOW>
2. what happens to the current ram in the computer if you plan to have
ram on the 030 board?
<J.ALLEN27>
Keeps on trukin
<DAVESMALL>
Depends on if we decide to use it for anything. There is a possibility
we have to investigate of using it in 16 mhz mode, or just using it for
68000 local functions.
Your GCR was shipped Halloween.
(What, is UPS running ultra-slow or something?!?)
<W.CHOW>
3. i would buy one, if you leave a socket for the 030, so i can buy a
16Mhz or 25 or 33 Mhz depending on $$$ i have available
<DAVESMALL>
That's how I want it, too. Remember, a big Mhz 030 costs BIG dollars ..
<J.ALLEN27>
It ain't that easy...the Mhz impacts the design.
<DAVESMALL>
Sure does. But I'd like a faster CPU option as well. However, even
16 would smoke right along.
<[Double Click] M.VEDERMAN2>
Hi dave and jim... I came in late too (sorry :-)
<DAVESMALL>
Hi Mike .. our Jamie is enjoying your Double-Clicker. A new love
interest will soon enter your life, and, you will soon receive a raise.
next?
<[Double Click] M.VEDERMAN2>
Let me ask a couple questions, which may be premature considering this
is market research, but will TOS 1.6 (030 TOS) be compatible, or plans
for it to work?
<J.ALLEN27>
NO COMMENT!!!!
<DAVESMALL>
hehehe. No, seriously, it's one thing we're looking into, but we don't
want to bias Atari on this issue.
<[Double Click] M.VEDERMAN2>
another Q... Could higher than 16 mhz be acheived, or would y'all
have to use an onboard clock?
<DAVESMALL>
Anything can be done, but it can cost more. 16 seems most approachable,
but I'd be willing to pay for 25 capability. Question is if anyone
else would.
<J.ALLEN27>
We would use the 32 Mhz clock in the ST!!!
<DAVESMALL>
This thing will be a rocket even at 16, though. 32 would run socks
off of several 386 systems I know.
<[Double Click] M.VEDERMAN2>
Three... Ready to open the 'compatibility' can of worms that this
will surely open?
<J.ALLEN27>
Someone has too.
<DAVESMALL>
I don't really think it's that bad. Sure, there's stack frame problems
-- but with the TT, they'll all be raised anyway, so we might as well
get started. It's nearly an identical issue.
Lots of stuff will work, lots more will work with a minor patch, a few
things will break horribly. This is why we have Codehead software.
Someone's gotta fix these things.
<[Wilton] W.VARGAS2>
Exactly what are the projected benefits, speed increase and
compatibilty of the 68030 board?
<DAVESMALL>
It's unanswerable at this point. We're asking more marketing questions
than giving exact numbers, because the amount people will pay will
affect the product a great deal.
<[Wilton] W.VARGAS2>
If it speeds up CAD-3D, then you have my support!
<DAVESMALL>
So, I can't project exact speed, because I don't know the exact speed
yet, and so forth. We're asking if people are interested.
<[Wilton] W.VARGAS2>
Any defined price range?
<DAVESMALL>
I hate to seem like I'm dodging the subject -- ask me when I get a
wirewrap board.. (grin)
<P.ATKOCIUS>
by the way has my gcr been shippred yrt? 02192
<DAVESMALL>
Sandy's shut down the Mac II, alas ... sorry .. can you email me?
<P.ATKOCIUS>
okay
<DAVESMALL>
(it probably has --)
<[Wilton] W.VARGAS2>
Is the GCR a cartridge or it goes inside the computer? Also does it
work with the STacy?
<DAVESMALL>
GCR is a cartridge, and it works fine with Stacy. In fact, Doug is on
tonight on a Stacy, so we know.
<[Wilton] W.VARGAS2>
will it work concurrently with the PC ditto II?
<DAVESMALL>
Incidentally, if you make it to Comdex, please drop by the Atari booth
and say hello. It's nice to meet the faces behind the names.
<[Wilton] W.VARGAS2>
I WISH I COULD!!!!!!!!!!!!!!!
<DAVESMALL>
Not concurrently (at same instant) but they don't fight each other.
Ah, with a 68030, they might be able to... (no kiddin'..)
<[Wilton] W.VARGAS2>
so its possible to have both things at the same time and that would
mean you have three computers on one?
<DAVESMALL>
Yup -- well, it works that way now -- you just select the computer
type you need and boot up into it. Switching is just a reboot.
<[Wilton] W.VARGAS2>
That 030 will be a great thing to wait for!
<[John] JJKENNEDY>
How compatible will the 030 board be with the TT (if at all)? Will
some/many/most prg written for the TT work on the 030 upgrade?
<DAVESMALL>
Very compatible. Very. However, the TT has features the ST does not
in the screen, etc, and in sound, that it couldn't be backwards
compatible with.
<J.ALLEN27>
Yes unless they are using the TT specific video wonders.
<DAVESMALL>
Give me a video chip and a star to steer it by -- we'll see what we
can do.
<[John] JJKENNEDY>
OK Many thanks!
<JEFF.W>
So you're really looking at a very, very, very, fast ST basically?
<DAVESMALL>
Jeff > Yes indeed. An ST that'll be fast enough to compete in the 90's.
Remember, the ST is basically a 1985 design ... it's time it was
brought up to date with the NeXT and Mac IIci speed.
<[Norm] N.RECHTMAN1>
How did Doug get a STacy, did he borrow Donny Osmonds?????? When are
they gonna be released?
<DAVESMALL>
We haven't been told. No one knows. (Doug is currently dating Marie
Osmond, is how)
<DOUG.W>
oh? :-)
<DAVESMALL>
errrrrrrrr I'm sorry, I can't answer that one -- legal agreements and
nondisclosures.
<DOUG.W>
So *that's* who she is!
<[Wilton] W.VARGAS2>
what exactly do you mean with "barely' seen a TT"?
also any graphics upgrades on that fabulous mind?
<DAVESMALL>
they're a definite possibility. I'm a big fan of TI graphics chips.
<[Wilton] W.VARGAS2>
AAH! I wish I would be an insider!
<DAVESMALL>
me too!
<[Wilton] W.VARGAS2>
that's it! thanks and go ahead! (but you know more than I!
<DAVESMALL>
thank you!
<JEFF.W>
That's about it for the formal part....
<J.ALLEN27>
Thanks folks!!!
<JEFF.W>
Thanks all for coming. Thanks to Dave, Jim, and Doug. Before we
open this up, any closing remarks?
<DOUG.W>
I'd like to thank Dave for typing all the replies I was feeding him!
(Just kidding *grin*)
<DAVESMALL>
Let me add my thanks for all of you for attending the conference and
for the interest shown! While I'm not absolutely positively promising
this as a product, we're certainly talking about this CO and the
interest generated, as we did the Spectre 128 and its conference. The
on-line networks bring a new way to gauge consumer interest if
companies just will take the time to ask, and listen to comments.
Thanks again for coming.
<DAVESMALL>
Doug: Good luck with Marie.
<JEFF.W>
Again...thanks to all for coming.achowe@tiger.waterloo.edu (anthony howe) (11/14/89)
After reading the conference, I WANT ONE!!! Ohhh Honey sweet, tastes
really neat!! Heavy panting. A nice solid 030 board with scads of RAM
sounds good.
Q: While many applications should work like skiing speed demons, will
there be some facilty, like a switch (keyboard or physical), to flip
between the orginal 68K and the 030. I ask this because, while I love
programming and exploring the ST, I love my games too. Now I feel that
many games will work with no problem, I think some games will be
impossible to play because they'll be so fast!!! I mean I still haven't
finished << Time Bandits >> and up around those 3rd and 4th level,\
greebles move pretty frigging quick as is.
- ant
achowe@tiger.waterloo.edu | "It is hard to make the world go away
_ -|-|_ _ | when it has decided to notice you."
(_\ |\| | | | (_) |\| \/ | - Spock's World
___/ | disclaimer...gl8f@astsun9.astro.Virginia.EDU (Greg Lindahl) (11/15/89)
In article <23970@cup.portal.com> Xorg@cup.portal.com (Peter Ted Szymonik [ quoting dave small ]) writes: [ Dave Small's blurb about this maybe-030 upgrade for the ST... ] > We will also "clock" the 68030 processor at a much higher speed > than the ST's 68000. The ST runs at 8 mhz (millions instructions / > second); we'll go for at least 16 Mhz and possibly more. This is wrong. The ST has a clock rate of 8 mhz. But it takes 4 cycles for the processor to fetch a word from memory, and so memory is really available only 2 million times a second. Then access is interleaved between video/refresh and the CPU, so the CPU gets 1 million per second. Gosh, I don't really remember anymore. Anyway, it isn't 8 million instructions per second. ------ Greg Lindahl gl8f@virginia.edu Astrophysicists for Choice.
dnewton@carroll1.UUCP (Dave 'Post No Nicknames' Newton) (11/15/89)
In article <2278@hudson.acc.virginia.edu> gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) writes: >In article <23970@cup.portal.com> Xorg@cup.portal.com >(Peter Ted Szymonik [ quoting dave small ]) writes: >> than the ST's 68000. The ST runs at 8 mhz (millions instructions / >> second); we'll go for at least 16 Mhz and possibly more. >Gosh, I don't really remember anymore. Anyway, it isn't 8 million >instructions per second. You mean... my ST isn't an 8-MIPS machine? Oh no. Sadness. -- David L. Newton | uunet!marque!carroll1!dnewton | The Raging Apostle-- (414) 524-7343 (work) | dnewton@carroll1.cc.edu | for the future-- (414) 524-6809 (home) | 100 NE Ave, Waukesha WI 53186 | for the world. "Isn't it fun to take two unrelated sentences and mix the batter lightly?" -me
pa1323@sdcc13.ucsd.edu (Some call me...Tim) (11/16/89)
In article <2278@hudson.acc.virginia.edu> gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) writes: >In article <23970@cup.portal.com> Xorg@cup.portal.com >(Peter Ted Szymonik [ quoting dave small ]) writes: > >[ Dave Small's blurb about this maybe-030 upgrade for the ST... ] > >> We will also "clock" the 68030 processor at a much higher speed >> than the ST's 68000. The ST runs at 8 mhz (millions instructions / >> second); we'll go for at least 16 Mhz and possibly more. > >This is wrong. ... [Rationalization deleted] >Gosh, I don't really remember anymore. Anyway, it isn't 8 million >instructions per second. The 68000 processor takes about 8-16 cycles per instruction, on the average. The minimum instruction time is 4 cycles, up to a worst case of 170 cycles for one instruction (a signed divide on an absolute memory address). The ST therefore operates at about .75 million instructions per second, or MIPS. With a 68030, however, there are three speed improvements without even changing the clock--the instructions take fewer cycles, memory is accessed in 32-bit chunks instead of 16, and the 68030 has a cache, meaning that it reads data and instructions before they are needed, while it's doing other things. Result: You can get a 4-8 times speed improvement without even changing the clock rate, and with twice the clock, 8-16 times... These numbers are approximate, and if you want to argue them with me, don't waste net bandwidth, please! Respond via email and we can hash it out there... ------------- Tim Mensch. Internet: tmensch@ucsd.edu Bitnet: UCSD!TMENSCH P.S. I agree that the Atari vs. the world bickering has gone on too long, and should be rerouted to either email, alt.flames, or /dev/null. I'm not picky.
covertr@force.UUCP (Richard E. Covert) (11/16/89)
In article <2278@hudson.acc.virginia.edu>, gl8f@astsun9.astro.Virginia.EDU (Greg Lindahl) writes: > In article <23970@cup.portal.com> Xorg@cup.portal.com > (Peter Ted Szymonik [ quoting dave small ]) writes: > > [ Dave Small's blurb about this maybe-030 upgrade for the ST... ] > > > We will also "clock" the 68030 processor at a much higher speed > > than the ST's 68000. The ST runs at 8 mhz (millions instructions / > > second); we'll go for at least 16 Mhz and possibly more. > Actually, I read DaveSmall's articel with great interest. I hope that between Dave Small and Jim Allen they can produce this 68030 addon. One interesting note was that, for the $$, you could get their addon with a 33 MHz 68030. Not that's CHOICE!!! Rich Covert
rehrauer@apollo.HP.COM (Steve Rehrauer) (11/16/89)
In article <46dba878.14a1f@force.UUCP> covertr@force.UUCP (Richard E. Covert) writes: >Actually, I read DaveSmall's articel with great interest. I hope that >between Dave Small and Jim Allen they can produce this 68030 addon. >One interesting note was that, for the $$, you could get their addon >with a 33 MHz 68030. Not that's CHOICE!!! Whoa, time-out! I read Dave Small saying he wasn't promising What, Where, When, How Much or even IF. Let's allow Mr. Small to establish what we will or won't get for our $$, or come this time next year we'll all be screaming "Vapor!" at the poor guy. (I also hope he does sell the beast, though I'd certainly want to know the What and How Much before saying I'd buy one.) -- >>"Aaiiyeeee! Death from above!"<< | Steve Rehrauer, rehrauer@apollo.hp.com "Flee, lest we be trod upon!" | The Apollo Computer Division of H.P.
nemeth@gpu.utcs.utoronto.ca (Gabe Nemeth) (11/17/89)
Whats so great about Dave Small's vapourware product? There is a german board out called PAK68 - full documentation, full schematics, full ROM modifications that has been out for well over a year now and is relatively cheap. Its a 68020 with space for floating point processor. Sounds pretty good to me and is available now!
ins_bac@jhunix.HCF.JHU.EDU (Ajay Choudhri) (11/17/89)
In article <1989Nov16.180311.23182@gpu.utcs.utoronto.ca> nemeth@gpu.utcs.UUCP (Gabe Nemeth) writes: >Whats so great about Dave Small's vapourware product? There is a german >board out called PAK68 - full documentation, full schematics, full ROM >modifications that has been out for well over a year now and is >relatively cheap. Its a 68020 with space for floating point processor. >Sounds pretty good to me and is available now! Available now...well could you post more info for us non-europeans. I'd be interested in seeing some specifics on this.. compatibility, speed increase, etc thanks -Ajay Choudhri ins_bac@jhunix.HCF.JHU.EDU ins_bac@jhunix.UUCP
mike@system.Cambridge.NCR.COM (mike reiss) (11/18/89)
In article <1989Nov16.180311.23182@gpu.utcs.utoronto.ca> nemeth@gpu.utcs.UUCP (Gabe Nemeth) writes: >Whats so great about Dave Small's vapourware product? There is a german >board out called PAK68 - full documentation, full schematics, full ROM >modifications that has been out for well over a year now and is >relatively cheap. Its a 68020 with space for floating point processor. >Sounds pretty good to me and is available now! Is this a crock of S*** or what? Somebody starts talking about making a board that will help the ST and what happens ... somebody starts jumping on him for having vaporware !!! Gees, the guy has just said is is going to consider do something ... he hasn't even said for sure that he will do it and he gets jumped on. Be real, at least give the guy a chance ... if you want another product, go for it - no problem, but don't jump on Dave. He's trying to do something positive for the ST. He has a proven record of succeeding at what he tries also. Dave --- keep up the good work, the ST needs more people like you !!! mike
hase@netmbx.UUCP (Hartmut Semken) (11/20/89)
In article <3322@jhunix.HCF.JHU.EDU> ins_bac@jhunix.UUCP (Ajay Choudhri) writes: >In article <1989Nov16.180311.23182@gpu.utcs.utoronto.ca> nemeth@gpu.utcs.UUCP (Gabe Nemeth) writes: >[PAK68 68000-proc replacement with 020/881 availeble now] >Available now...well could you post more info for us non-europeans. >I'd be interested in seeing some specifics on this.. >compatibility, speed increase, etc So, here t goes: The PAK68 is an c't project. c't is a german computer magazine. Instuctions to build it were published some time ago, together with full circurit description. compatibility: none. TOS doesn't run; a TOS version hacked to run is not very compatible (why, why did Atari use line-F for anything but floating point?) speed increase: well, the 020 runs at 8 MHz. The circurit was not designed specifically for the ST... circurits like Turbo-16, CMI speed or the like were designed for the St, so they can fall back to 8 MHz wherever necessary and run at higher clocks when the async BUS is used. The PAK68 has support for high speed SRAM on a daugterboard. But nobody made this availeble to TOS; the custom chips (DMA and SHIFTER) cannot acess it (their adress registers are inside the MCU, so they can only acess the MCU-supported DRAM). And most programs do not like to run in anything but the standard RAM... overall the speed increase is compareble to any 16-MHz 68000 accelerator for the ST. With ther german RTOS (nothing to do with TOS!), the PAK helps a lot more, cause RTOS can use the cache, the 020 specific instructions and the 881. I exspect the Gadgets board to help a lot with Spectre and do almost nothing for TOS. The ST is a closed machine; a compatible, faster redesign is pretty exspensive. TOS not only does not support processors other than 68000, it fights them :-( hase -- Hartmut Semken, Lupsteiner Weg 67, 1000 Berlin 37 hase@netmbx.UUCP Dennis had stepped up into the top seat whet its founder had died of a lethal overdose of brick wall, taken while under the influence of a Ferrari and a bottle of tequila. (Douglas Adams; the long dark teatime...)
gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) (11/20/89)
In article <3795@netmbx.UUCP> hase@netmbx.UUCP (Hartmut Semken) writes: >I exspect the Gadgets board to help a lot with Spectre and do almost >nothing for TOS. Hm. I thought the FaST Tech board was a 16mhz 68000 with a write-through 32k cache. So if you do lots of reading in a row, you get a big win. Now if you essentially mounted a 16mhz 68030 with a 16 bit data path onto the same board, and somehow get around the Fline problem, then you'd have something that would speed things up more than the Turbo16. Right? It wouldn't be as good as a 32-bit data path with no speed decrease on writes, but still better than a 68000 at 16mhz. Of course, I don't know if the '030 can handle different bus widths. My Moto books aren't that recent. The way that it wins over the PAK68 is that Jim Allen has partially decoupled the processor speed from the video speed. > The ST is a closed machine; a compatible, faster >redesign is pretty exspensive. But the TT *is* a compatible redesign with 32-bit everything, right? It will be interesting to compare the TT to other similar '030 boxes, when/if it arrives. ------ Greg Lindahl gl8f@virginia.edu Astrophysicists for Choice.
ge@kunivv1.sci.kun.nl (Ge' Weijers) (11/22/89)
dnewton@carroll1.UUCP (Dave 'Post No Nicknames' Newton) writes: > You mean... my ST isn't an 8-MIPS machine? Oh no. Sadness. It's not even a 1-MIPS (VAX/780-MIPS!) machine. It's on the same level as a VAX/750, at 0.7 MIPS or so. Ge' Weijers Internet/UUCP: ge@cs.kun.nl Faculty of Mathematics and Computer Science, (uunet.uu.net!cs.kun.nl!ge) University of Nijmegen, Toernooiveld 1 6525 ED Nijmegen, the Netherlands tel. +3180612483 (UTC-2)
ge@kunivv1.sci.kun.nl (Ge' Weijers) (11/22/89)
nemeth@gpu.utcs.utoronto.ca (Gabe Nemeth) writes: >Whats so great about Dave Small's vapourware product? There is a german >board out called PAK68 - full documentation, full schematics, full ROM >modifications that has been out for well over a year now and is >relatively cheap. Its a 68020 with space for floating point processor. >Sounds pretty good to me and is available now! The PAK (Prozessor Austausch Karte) is a 68020 running with a 16-bit bus and a slow clock. It speeds up things, but not by a factor 4-8. No competition, and it's compatibility is low. Ge' Weijers Internet/UUCP: ge@cs.kun.nl Faculty of Mathematics and Computer Science, (uunet.uu.net!cs.kun.nl!ge) University of Nijmegen, Toernooiveld 1 6525 ED Nijmegen, the Netherlands tel. +3180612483 (UTC-2)
kbad@atari.UUCP (Ken Badertscher) (11/22/89)
gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) writes: | Now if you essentially mounted a 16mhz 68030 with a 16 | bit data path onto the same board [as T16] | then you'd have something that would speed things up more | than the Turbo16. Right? Right. And something else to keep in mind: the TOS used in the STE does not use the line F compression that previous TOS versions used. | But the TT *is* a compatible redesign with 32-bit everything, right? Right. -- ||| Ken Badertscher (ames!atari!kbad) ||| Atari R&D System Software Engine / | \ #include <disclaimer>
saj@chinet.chi.il.us (Stephen Jacobs) (11/22/89)
In article <529@kunivv1.sci.kun.nl> ge@kunivv1.sci.kun.nl (Ge' Weijers) writes: >dnewton@carroll1.UUCP (Dave 'Post No Nicknames' Newton) writes: > >> You mean... my ST isn't an 8-MIPS machine? Oh no. Sadness. > >It's not even a 1-MIPS (VAX/780-MIPS!) machine. It's on the same level >as a VAX/750, at 0.7 MIPS or so. On the other hand, the main computing resource at the laboratory where I work (food R&D, but what the hey) was a VAX 11/750 for several years. And I got a LOT of mileage out of casually mentioning to the real computer people that my ST acted a lot like their monster, except that since it was single user... Steve J.
hase@netmbx.UUCP (Hartmut Semken) (11/22/89)
In article <2307@hudson.acc.virginia.edu> gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) writes: >In article <3795@netmbx.UUCP> hase@netmbx.UUCP (Hartmut Semken) writes: >>I exspect the Gadgets board to help a lot with Spectre and do almost >>nothing for TOS. >Hm. I thought the FaST Tech board was a 16mhz 68000 with a >write-through 32k cache. >The way that it wins over the PAK68 is that Jim Allen has partially >decoupled the processor speed from the video speed. True. The PAK68 was designed to be swapped against any 68000 proc. It's not ST specific and does not fit in the ST enviroment nicely. A ST specific board can do pretty much more. The main problem with the 020 cache is self modifying code. The external, write trough cache of the 16 MHz 68000 boards works well, the 020 and 030 internal chaches won't. As Alan said, some programs will loose on the TT; the reason is the cache. >> The ST is a closed machine; a compatible, faster >>redesign is pretty exspensive. >But the TT *is* a compatible redesign with 32-bit everything, right? >It will be interesting to compare the TT to other similar '030 boxes, >when/if it arrives. For me, the "if" is true. I'll believe in this machine, if I could buy one at my dealer's. Mr Hartman said, it would ship Sep to Okt 1989 (march, CeBIT). This statement was corrected at the Atari fair in Duesseldorf to Nov 89. Now it's "early 1990". Too bad. Now, Nov, is the time, everybody discovers, he has something left in the budget. He thinks about what to spend it on. He will not spend it on a non-existing machine... hase -- Hartmut Semken, Lupsteiner Weg 67, 1000 Berlin 37 hase@netmbx.UUCP Dennis had stepped up into the top seat whet its founder had died of a lethal overdose of brick wall, taken while under the influence of a Ferrari and a bottle of tequila. (Douglas Adams; the long dark teatime...)
Xorg@cup.portal.com (Peter Ted Szymonik) (11/23/89)
You guys are amazing (for lack of a better word), Dave and Jim hold a CO to judge the market for their *proposed* new ST product and already its vaporware!! Some of you needs HEAVY doses of valium my friends! :-] Happy Thanksgiving Peter Szymonik Xorg@cup.portal.com
walkerb@tramp.Colorado.EDU (Brian Walker) (11/24/89)
In article <1814@atari.UUCP> kbad@atari.UUCP (Ken Badertscher) writes: >gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) writes: >| [comments to previous conversation deleted] >Right. And something else to keep in mind: the TOS used in the >STE does not use the line F compression that previous TOS versions used. >| But the TT *is* a compatible redesign with 32-bit everything, right? >Right. Sounds great. I am curious. What did you do with the line F compression? Did you move it over to line A, convert them into procedure calls or was it something else? Would you care to indulge my (and perhaps other's) curiosity? Brian Walker, University of Colorado at Boulder walkerb@tramp.colorado.edu ...!{ncar,nbires}!boulder!tramp!walkerb A person who claims that absolute zero is impossible to obtain has not taken a quiz in thermo yet.
rwa@cs.AthabascaU.CA (Ross Alexander) (11/25/89)
gl8f@astsun.astro.Virginia.EDU (Greg Lindahl) writes: >Of course, I don't know if the '030 can handle different bus widths. >My Moto books aren't that recent. It can, and does a very nice job of it; it can resize the bus on a cycle-by-cyle basis. So you could have local 32 bit wide memory on the 68030 daughter card plus a 16-bit wide path to the ST mem section. This would work very nicely if you were careful to keep this like PhysBase and the DMA pointer pointing into ST memory - it would be easy to hack ST-Minix to do this. Ross
kbad@atari.UUCP (Ken Badertscher) (11/28/89)
walkerb@tramp.Colorado.EDU (Brian Walker) writes: | What did you do with the line F compression? It's gone. Because of the larger ROM space in the STE, (and TT), the line F compression is no longer required to make the TOS ROM image fit. With the exception overhead gone, the AES moves along noticeably faster, too. -- ||| Ken Badertscher (ames!atari!kbad) ||| Atari R&D System Software Engine / | \ #include <disclaimer>