knudsen@ihwpt.UUCP (mike knudsen) (09/26/85)
(Re-Reposted due to more machine-changeover foulups) Is this just cultural coincidence, or was there a hardware reason for the following: In home computers, descendants of the 8080 have aimed at "serious business" TEXT-ONLY applications, whereas the 6800's descendants' machines have featured bit-mapped color graphics and sound to a much greater extent. To list examples of these descendant micros and their computers: 8080 6800 ---- ---- Z80 6502 TRS-80 Models I-IV, Apple II, Atari, Vic-20, C-64 CP/M systems 6809 TRS-80 Color Computer 8086,-8 68000 IBM PC and 1001 clones Apollo, Mac, Atari-ST, Amiga, AT&T 7300 (can run text-only) Sure puts most of the decent graphics into the right-hand side, doesn't it? Yes, there are exceptions -- the Z80-based Exidy Sorcerer and the 8080 CompuColor, and now that business types have discovered graphics you can get hi-res color on the IBM-PCs. But from 1977 thru '83 the picture is pretty lopsided. And it's not just marketing -- the TRS-80 and Apple II were both supposed to handle games and hobby-hacking, AND science and business. As for the latest machines: while you can use an IBM-PC for years w/out graphics, you can't even *talk to the OS* on the 68K machines listed without clicking a mouse! Ever see a text-only Mac? Want to? I suspect part of this is due to the equivalence of processor cycles and bus cycles in the 6800/6502/6809 chips. During the first half of each cycle, the bus is unused by the micro, so the designer could sneak a free DMA timeslot on every cycle for refreshing video graphics from RAM, without slowing down the processor. (This also makes for easier DRAM refreshing). This trick was well-known to 6502 hackers (eg, Hal Chamberlin) and was institutionalized by Motorola in their graphics chip set for the 6809 (the applications note for these chips became the Radio Shack Color Computer). The 8080-type micros, with their irregular bus-access subcycles, could NOT use this scheme; unable to DMA the large RAM areas needed for bitmap graphics, the Z80 boxes had to stick with separate, small, text-only display memories. Interesting how subtle hardware features of a micro can influence the architecture and hence end uses of its computers. Open for commentary -- mike k ihnp4!ihwpt!knudsen
doug@terak.UUCP (Doug Pardee) (09/27/85)
> Is this just cultural coincidence, or was there a hardware > reason for the following: In home computers, descendants of > the 8080 have aimed at "serious business" TEXT-ONLY applications, > whereas the 6800's descendants' machines have featured > bit-mapped color graphics and sound to a much greater extent. I doubt that anyone knows "for sure" how this situation came about, but here's my understanding... The 2-phase bus nature of the 6502-type chips was indeed a big influence in their use on "color" systems. Perhaps equally as important, the 6502 was introduced at the unheard-of price of $25, when 8080's were still going for $150. What the 8080 (and later, the Z80) had going for it was, simply put, "floppy disks". For the "high-rolling" computer hobbyist who could afford to play with floppy disks, the extra hundred bucks for an 8080 was unimportant. And they could also afford the (not very cheap at the time) Western Digital floppy disk controller chips. These FDC's were easy to connect to the 8080, but relatively difficult to connect to the 6502. The bigger bucks bought much faster data transfer rates. And those transfer rates made it possible to write "real" operating systems like CP/M and TRSDOS, and to develop applications which used significant amounts of disk data. Put this all together, and you get two very different kinds of systems. Low cost systems with cassette (or, at best, very slow disks) were built around the 6502; their big drawing card was color graphics. Higher-priced systems with "real" disks were based around the 8080; their big drawing card was performance on any application that used disk heavily, especially for data storage. Applications like word processing and databases. -- Doug Pardee -- CalComp -- {calcom1,savax,seismo,decvax,ihnp4}!terak!doug
robert@nmtvax.UUCP (10/01/85)
In article <> doug@terak.UUCP (Doug Pardee) writes: > >What the 8080 (and later, the Z80) had going for it was, simply put, >"floppy disks". For the "high-rolling" computer hobbyist who could >afford to play with floppy disks, the extra hundred bucks for an 8080 >was unimportant. And they could also afford the (not very cheap at the >time) Western Digital floppy disk controller chips. These FDC's were >easy to connect to the 8080, but relatively difficult to connect to the >6502. The bigger bucks bought much faster data transfer rates. And >those transfer rates made it possible to write "real" operating systems >like CP/M and TRSDOS, and to develop applications which used significant >amounts of disk data. > >Doug Pardee -- CalComp -- {calcom1,savax,seismo,decvax,ihnp4}!terak!doug I have spent more time waiting for a Rainbow, TRS-80, and a Motorola VME-10 than I have my Apple. Don't think that I am protecting an obsolete machine just because I own one. Note that the "slow" apple drives are still used in comparison for many systems. Considering that the processor has to do all the disk handling, its speed is fairly impressive compared to systems with floppy controllers. I still think Atari's smart peripherals were a mistake. "My 'smart' drive makes my computer go "bede-bede" whenever it does anything." Sounds pretty stupid to me... -- "Rock Is a Drug." - Spliff, "Rock Is a Drug" "Living in the limelight the universal dream" -Rush, "Limelight" Robert Kenyon New Mexico Institute of Mining and Technology ...ucbvax!unmvax!nmtvax!robert or ...ucbvax!unmvax!nmtvax!curly!rob
knudsen@ihwpt.UUCP (mike knudsen) (10/07/85)
> In article <> doug@terak.UUCP (Doug Pardee) writes: > > > >time) Western Digital floppy disk controller chips. These FDC's were > >easy to connect to the 8080, but relatively difficult to connect to the > >6502. The bigger bucks bought much faster data transfer rates. And > >those transfer rates made it possible to write "real" operating systems > >like CP/M and TRSDOS, and to develop applications which used significant > >amounts of disk data. > > I have spent more time waiting for a Rainbow, TRS-80, and a Motorola > VME-10 than I have my Apple. Don't think that I am protecting an > obsolete machine just because I own one. Note that the "slow" apple > drives are still used in comparison for many systems. Considering that > the processor has to do all the disk handling, its speed is fairly impressive > compared to systems with floppy controllers. > > I still think Atari's smart peripherals were a mistake. "My 'smart' drive > makes my computer go "bede-bede" whenever it does anything." Sounds > pretty stupid to me... I'll agree with Doug on this -- my CoCo drives are quite fast, even tho the 6809 (a descendant of the 6800 AND 6502, whether or not Motorola admits it) has to handle all the bytes. And yes, those "smart" disks on the Atari and Commodore are slow enuf to keep any "real" DOS off those machines, due to slow load times for programs not in RAM. I believe the C-64 disk is slower than CoCo's CASSETTE! However, I think the point of the 8080 article was that LSI floppy controller chips were avaiable EARLIER for the 8080/Z80 than were the excellent chips that we now enjoy on the 6502/6809 machines, and this led to the biz/text -vs- color grafix dichotomy. mike k
steve@wlbr.UUCP (Steve Childress) (10/08/85)
In article <792@nmtvax.UUCP>, robert@nmtvax.UUCP writes: > > I still think Atari's smart peripherals were a mistake. "My 'smart' drive > makes my computer go "bede-bede" whenever it does anything." Sounds > pretty stupid to me... > -- > As I recall, Atari embarked upon the intelligent peripheral scheme as an answer to the FCC emissions edict. They (mistakenly) chose to answer the problem by using an easy to shield serial bus for peripherals. Thus, the peripherals HAD to be smart as compared to ordinary, parallel- bus resident peripherals. Apple's success in running parallel cables to the floppies despite the FCC requirements demonstrates Atari's mistake. Regards, Steve Childress Eaton IMS R&D Group MS 43 31717 La Tienda Drive Westlake Village, CA 91360 (818) 889-2211 X2148 {trwrb, scgvaxd, ihnp4, voder, vortex} !wlbr!steve or ...wlbr!wlbreng1!steve
stekas@hou2g.UUCP (J.STEKAS) (10/09/85)
> As I recall, Atari embarked upon the intelligent peripheral scheme > as an answer to the FCC emissions edict. ... > > Steve Childress wlbr!steve The reason for intelligent periphs was to minimize the cost of the basic machine. An Atari talks to all it's periphs over a single serial line using a very simple protocol. The design not only minimizes hardware but software as well. Since the smarts are in the peripherals, device handlers are simple and compact - very important when one only has 64K to play with. Atari's 8K OS includes disk, cassette, printer, keyboard and screen handlers, a simple screen editor, transcendental math functions for Basic (or whatever), and the usual boot and initialization stuff. Jim
jimomura@lsuc.UUCP (Jim Omura) (10/10/85)
Just for your interest, I was also under the impression that
the Commodore disc was slower than the CoCo cassette. It isn't
I can't remember the baud rate exactly, but it's a few times faster.
Cheers! -- Jim O.
It's still amazingly slow though. The new 128 is faster.
--
James Omura, Barrister & Solicitor, Toronto
ihnp4!utzoo!lsuc!jimomura
Byte Information eXchange: jimomura
Compuserve: 72205,541
MTS at WU: GKL6ln63fkn@sdcc7.UUCP (Paul van de Graaf) (10/12/85)
In article <667@hou2g.UUCP> stekas@hou2g.UUCP (J.STEKAS) writes: >... only has 64K to play with. Atari's 8K OS includes disk, >cassette, printer, keyboard and screen handlers, a simple screen editor, >transcendental math functions for Basic (or whatever), and the usual >boot and initialization stuff. > Jim The original Atari OS is 12K, ~8K if you leave out the floating point package. The XL series have 16K of OS, due to 2 character sets, diagnostics, and other cruft. Also included is a graphics handler which draws lines, fills regions etc. The main reason the OS is so small is that it has a centralized i/o subsystem. A device driver need only supply open, close, put, get, and status routines to hook into the OS. All the intelligent controllers bought you was a little convience. The drive drives can format themselves, and read and write a given sector, but not much else. In fact many devices have to load in their own device drivers into RAM over the serial line during boot-up. These drivers were either too big to fit in ROM, or developed after the OS was frozen in ROM. In all I'd say smart devices are a good idea if the peripheral bus has sufficient bandwidth and the "smarts" come relatively cheap. Atari met neither requirement. I talked to engineer who worked on the 800 just after it came out. I was attemping to sell Ataris at a Sears store in Mountain View at the time. He said that his managers were scared to death of the FCC licensing at the time, so they decided to go the serial bus route. (The Atari machines were the first to come out after the FCC started meddling.) The marketing people also liked the idea, because it discouraged 3rd party vendors (big mistake). Paul van de Graaf sdcsvax!sdcc7!ln63fkn U. C. San Diego