dillon@PAVEPAWS.BERKELEY.EDU (Matt Dillon) (06/20/86)
Anybody know if it's possible to upgrade the ST's 68000 to a 68010 or 68020? -Matt
jdg@elmgate.UUCP (Jeff Gortatowsky) (06/23/86)
In article <8606201839.AA06129@pavepaws>, dillon@PAVEPAWS.BERKELEY.EDU (Matt Dillon) writes: > > Anybody know if it's possible to upgrade the ST's 68000 to a > 68010 or 68020? > > -Matt As far as I've been able to determine the ST does no checking on boot for the type of 68000 installed. Therefore the exception handling routines in TOS will bomb (ie. lockup) miserably. Further (and I've mentioned this before) some dimwit went and put the exception numbers in the upper byte of the exception vectors!!!! Odviously this will give you an 'E' ticket into the Twilight Zone on the 020! Much of the system software in the ST is really JUNK. In Atari's defense however, they never claimed upwards compatibility among future machine archetectures (SP?). Still I'd pay another $50-$75 to get a new version of TOS in ROM (2.0 ?). No, this is not rumored. I just want Atari to know if they ever WERE to decided to re-write some of TOS/GEMDOS themselves (and get it right) instead of DRI's junk, *I'd* be willing to pay for it. Neil? Others???? Comments??? -- Jeff Gortatowsky {allegra,seismo}!rochester!kodak!elmgate!jdg Eastman Kodak Company <Kodak won't be responsible for the above comments, only those below>
robt@molihp.UUCP (Robert L Thurlow) (06/26/86)
In article <477@elmgate.UUCP> jdg@elmgate.UUCP writes: > ... Still I'd pay another $50-$75 to get a new version of >TOS in ROM (2.0 ?). No, this is not rumored. I just want Atari to know if >they ever WERE to decided to re-write some of TOS/GEMDOS themselves (and >get it right) instead of DRI's junk, *I'd* be willing to pay for it. > I, too, would pay more money to get rid of a few problems. It was not great to get the TOS in ROM and try my favorite bug-hunting procedures only to find most of the bugs still there. It was great to liberate that memory, and it was well worth the cost to get the quick reboot time, but there are still bugs there. Hell, I/O redirection doesn't even work sometimes now. And I agree it would be great to get the ability to drop a 68010 in there. (I'll wait awhile for the '020 : I heard yesterday that it is priced at $530 CDN for the 12.5 Mhz part.) BTW, anybody got a definitive list of bugs in TOS? # UUCP : ...!ubc_vision!molihp!robt Robert Thurlow # # My thoughts are my own, and are void where prohibited by law. # # "There was something fishy about the butler. I think he was a # # Pisces, probably working for scale." -Nick Danger, 3rd Eye #
manis@ubc-cs.UUCP (Vince Manis) (06/27/86)
I fail to see the attraction of the 68010. It's not substantially faster, and the few extra instructions it offers are unlikely to do a great deal in speeding up programs. The major virtue of the 68010 is that it supports demand paging, which is useful in virtual memory systems. The existing ST MMU renders this advantage dubious. The 68020 adds a fair number of instructions (notably 32-bit multiply and divide), and is in general a winner. On the other hand, there is little point in trying to add a 68020 to the ST directly, so I'm eagerly awaiting the TT box. Neil Harris, are you listening???? I'm dubious about a $530 CDN price for the 68020, though: the July issue of BYTE has an article on the Definicon 68020 co-processor board for the IBM PC. Including 1MB of memory, a 68020, and a 68881 (FPU), the price they list is US$995. That would cause me to guess that (in reasonable quantities) you can get a 68020 for ~ $150 US.
robt@molihp.UUCP (Robert L Thurlow) (06/27/86)
In article <297@ubc-cs.UUCP> manis@ubc-cs.UUCP (Vince Manis) writes: >I fail to see the attraction of the 68010. Just a little nicer, and several friends have them from those Motorola kits. Some things would be a squeak faster, though I doubt I'd *desolder* the 68000 in my 520ST to use it. >I'm dubious about a $530 CDN price for the 68020, though: the July issue of >BYTE has an article on the Definicon 68020 co-processor board for the IBM >PC. Including 1MB of memory, a 68020, and a 68881 (FPU), the price they list >is US$995. That would cause me to guess that (in reasonable quantities) you >can get a 68020 for ~ $150 US. $530 CDN it is - single quantity from L. A. Varah, as of 86/06/25. Shouldn't stay that way for long though, I agree. # UUCP : ...!ubc_vision!molihp!robt Robert Thurlow # # My thoughts are my own, and are void where prohibited by law. # # "There was something fishy about the butler. I think he was a # # Pisces, probably working for scale." -Nick Danger, 3rd Eye #