rjc@geech.ai.mit.edu (Ray Cromwell) (03/12/91)
I received the following in Email. No disrepected intended for the author of this message, but the described product seems bogus, or a hyperbole. ----begin included text---- Received: by minyos.xx.rmit.oz.au Date: Tue, 12 Mar 91 11:03:56 +1000 From: s883334@minyos.xx.rmit.OZ.AU (James Alan Hall) Message-Id: <9103120103.28501@minyos.xx.rmit.oz.au> To: rjc@pogo.ai.mit.edu Subject: Emulations. Status: OR In comp.sys.amiga.emulations you posted the following : > Subject: Re: Atari Emulator? > Date: 6 Mar 91 02:47:33 GMT > > [ stuff deleted ] > > What is this Amiga emulator your talking about? Sounds like vaporware > to me. Just consider what's needed, OS Roms, Agnus, Denise and Paula > chip, along with the support circuitry and ram to support them. > This isn't an emulator, it's an Amiga. The only ST emulator I can > imagine is an ENTIRE Amiga motherboard on a card plugged into > what? Does the ST have an expansion slot? > > Nevertheless, this would cost more then the ST or the Amiga itself > to implement. > > The Amiga is one of the only computer that defies to be emulated. Without wishing to start a computer war, a friend of mine has an Amiga emulator for his Atari 1040ST. It is a commercial package, thats been around since 1987, released by Metacomco. The manual states that it runs around 7 times faster than a real amiga, disk drive i/o is much faster, ^^^ but the whole thing slows down when it comes to emulating the specialised chips (some of them not all). It requires the atari blitter to be installed, which runs approx 40 times faster than the amigas blitter which somewhat ^^^^ compensates for the lack of those amiga chips. It also requires an amiga drive, which has to be attached as the second drive on the atari. Keeping in mind his version is from 87, I'm sure there would be improved versions of it around today. James. s883334@minyos.xx.rmit.oz.au (I would have posted this, but I dont have such privileges, just e-mail) ----end included text---- An 8mhz ST running 7 times faster than an Amiga while emulating Amiga hardware? A blitter that runs 40 times faster than the Amiga blitter? Since the ST has no copper, to emulate it means polling video beam registers constantly. This sounds too unreal considering a 25mhz 68030 only runs 9 times faster than a 7mhz 68000. For the ST's blitter to run 40 times faster than the Amiga's it would need a clock speed of over 100mhz. The ST disk i/o being faster than the Amiga's DMA driven disk seems suspect too(especially when running the fast file system). Can anyone confirm or deny that this Metacomco product really exists? Either someone is exaggerating ALOT, or Metacomco is falsly advertising, because I don't believe this.
csbrod@immd4.informatik.uni-erlangen.de (Claus Brod) (03/12/91)
rjc@geech.ai.mit.edu (Ray Cromwell) writes: > Without wishing to start a computer war, a friend of mine has an >Amiga emulator for his Atari 1040ST. It is a commercial package, thats been >around since 1987, released by Metacomco. The manual states that it >runs around 7 times faster than a real amiga, disk drive i/o is much faster, > ^^^ >but the whole thing slows down when it comes to emulating the specialised >chips (some of them not all). It requires the atari blitter to be installed, >which runs approx 40 times faster than the amigas blitter which somewhat > ^^^^ >compensates for the lack of those amiga chips. It also requires an amiga >drive, which has to be attached as the second drive on the atari. The whole thing is a fake. I disassembled the startup program way back in 1987 when I tried to make it run on my ST. It only prints some kind of message (if I recall it right, something like 'You need a blitter to run this') and exits. It doesn't check for a blitter, or course, nor does it do anything else that might seem useful. Seems like someone trying to fool us under the hood of Metacomco's name. >This sounds too unreal considering a 25mhz 68030 only runs 9 times >faster than a 7mhz 68000. For the ST's blitter to run 40 times >faster than the Amiga's it would need a clock speed of over 100mhz. The ST blitter is faster in some operations than the Amiga's blitter, but certainly not 40 times faster. Likewise, the ST's CPU is slightly faster than those in standard Amigas, as we all know, but that's about 10%, not 700%. >The ST disk i/o being faster than the Amiga's DMA driven disk seems >suspect too(especially when running the fast file system). ST disk I/O really is faster, even if you're using FFS. At least this is true when comparing my Mega ST with a friend's A2000. ---------------------------------------------------------------------- Claus Brod, Am Felsenkeller 2, Things. Take. Time. D-8772 Marktheidenfeld, West Germany (Piet Hein) csbrod@medusa.informatik.uni-erlangen.de ----------------------------------------------------------------------
pochron@rt8.cs.wisc.edu (David Pochron) (03/13/91)
In article <1991Mar12.022622.25956@mintaka.lcs.mit.edu> rjc@geech.ai.mit.edu (Ray Cromwell) writes: >From: s883334@minyos.xx.rmit.OZ.AU (James Alan Hall) > Without wishing to start a computer war, a friend of mine has an >Amiga emulator for his Atari 1040ST. It is a commercial package, thats been >around since 1987, released by Metacomco. The manual states that it >Can anyone confirm or deny that this Metacomco product really >exists? Either someone is exaggerating ALOT, or Metacomco is >falsly advertising, because I don't believe this. Well, I don't know if this confirms it, but here is an interesting tidbit: Sometime during 1987, a supposed "Amiga emulator" showed up on on a local BBS that supported several brands of machines. Unlike many of the "fake" emulators that just put up a Guru screen on the ST, this one was 420K ARCed, and the docs said it required an ST with a blitter installed in order to run. I told one of the user's who had an ST with a blitter on the BBS to try this thing out, but they were strangely silent when asked whether they ever got it working or not... All I can say is the dates and requirements match, but since I never actually saw it run, that's about all I can say. I got the impression from the doc file that this was strictly an "OS-emulator" That is, if you tried to run an Amiga program that accessed the hardware in any way, it would fail. In any case, the reported speed of this emulator is definately bogus, for the same reasons the author of the article I am following up to has stated. And the ST blitter doesn't run 40 times faster than the Amiga blitter anyway, yet more proof there is a lot of BS in the message. -- -- David M. Pochron | "Life's a blit, | and then you VBI." pochron@garfield.cs.wisc.edu |
daveh@cbmvax.commodore.com (Dave Haynie) (03/13/91)
In article <1991Mar12.022622.25956@mintaka.lcs.mit.edu> rjc@geech.ai.mit.edu (Ray Cromwell) writes: >> What is this Amiga emulator your talking about? Sounds like vaporware > Without wishing to start a computer war, a friend of mine has an >Amiga emulator for his Atari 1040ST. It is a commercial package, thats been >around since 1987, released by Metacomco. This is undoubtedly bogus. First of all, there wasn't much Metacomco code in the original AmigaOS anyway; what was there was simply part of the DOS subsystem. They might try to sell you a version of Tripos for the ST, but it wouldn't be an Amiga emulator. >The manual states that it runs around 7 times faster than a real amiga, disk >drive i/o is much faster, Again, not possible. In pure CPU loops, the original ST would go a tad faster than an A1000, may 11% maximum (the difference between 8MHz and 7.16MHz). The PC inspired floppy interface is slower on the ST. >but the whole thing slows down when it comes to emulating the specialised >chips (some of them not all). To run the Amiga OS, you are by definition using most of the features of the Amiga custom chips. >It requires the atari blitter to be installed, which runs approx 40 times faster >than the amigas blitter which somewhat Was the Atari blitter even available in '87? In any case, this is pure ST propaganda -- Stalin would have been proud. The ST blitter is roughly 1/4 to 1/2 the speed of the Amiga blitter doing operations the ST blitter supports. Amiga's blitter runs at double bus speed, 14.3 MHz, when cranked up. If you have fast RAM, the CPU and blitter can both be going at once. The ST blitter shares the bus with the 68000; it runs the same memory cycle as the 68000, and you can have only one on the bus at any given time. Additionally, the ST blitter supports 16 operations between two operands to one destination. The Amiga blitter supports 256 operations between three operands to one destination and additional tricks, like line draw. What it does well, it does faster than a 68020 at 14.3MHz. Most of the copper tricks cannot be emulated on the ST at all. >Can anyone confirm or deny that this Metacomco product really >exists? Either someone is exaggerating ALOT, or Metacomco is >falsly advertising, because I don't believe this. This is the first I have heard of any such product. I would consider that very strange for any product of any significance to the Amiga market that's supposedly been around since '87. My guess is that this is totally bogus. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "What works for me might work for you" -Jimmy Buffett
jesup@cbmvax.commodore.com (Randell Jesup) (03/13/91)
In article <19794@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: >In article <1991Mar12.022622.25956@mintaka.lcs.mit.edu> rjc@geech.ai.mit.edu (Ray Cromwell) writes: >> Without wishing to start a computer war, a friend of mine has an >>Amiga emulator for his Atari 1040ST. It is a commercial package, thats been >>around since 1987, released by Metacomco. > >This is undoubtedly bogus. First of all, there wasn't much Metacomco code in >the original AmigaOS anyway; what was there was simply part of the DOS >subsystem. They might try to sell you a version of Tripos for the ST, but >it wouldn't be an Amiga emulator. Plus even if there as, most of the code is copyright by Commodore. They could at best to a separate port of tripos (without copying what they did for commodore). And that only covers dos, which is about 1/8 of the rom. >>Can anyone confirm or deny that this Metacomco product really >>exists? Either someone is exaggerating ALOT, or Metacomco is >>falsly advertising, because I don't believe this. > >This is the first I have heard of any such product. I would consider that >very strange for any product of any significance to the Amiga market that's >supposedly been around since '87. My guess is that this is totally bogus. I second that guess. Also, Metacompco has been out of business for a while, I believe. The only thing I know of that they released for the ST was the bcpl compiler (and they're welcome to it). Ask him to show it to you (and make sure it's not a trick program - take a real amiga floppy with programs on it). BTW, I don't think the ST floppy controller can read amiga-formatted disks (very different low-level format that most controller chips puke on). Or, more amusing, try to have him show you the packaging, with the name "metacompco" on it, or an ad _from metacompost_ (not just a silly rumor column).... Sheesh. -- Randell Jesup, Keeper of AmigaDos, Commodore Engineering. {uunet|rutgers}!cbmvax!jesup, jesup@cbmvax.commodore.com BIX: rjesup The compiler runs Like a swift-flowing river I wait in silence. (From "The Zen of Programming") ;-)