cmcmanis@stpeter.Eng.Sun.COM (Chuck McManis) (07/11/90)
There has been a lot of misinformation flying about about 2.0 and its compatibility or incompatibility on this net. Since I now have "release" 2.0 I assume for the moment that it is done and that developers can talk about it. (If the A3000 introduction doesn't count as a public demonstration nothing does :-) [inside joke]) So let's start with this one from Sean : In article <3456@crash.cts.com> (Sean Cunningham) writes: >KS2.0 can't work with the A1000 because it's only got 256K of writable control >store...KS2.0 takes up 512K, and requires TWO eproms in the A3000. KS2.0 will >be a DUAL ROM set (it'll be interesting to see how they make it work with >A2000s). This is infact incorrect on two counts. The first is that 2.0 won't work on a 1000 due to the smaller WCS. In fact, Commodore has devised a scheme which developers use which can "kick" 2.0 into the first autoconfig'd fast ram board. This continues to work, even with "release" 2.0. This makes the A1000 boot sequence Kickstart 1.3->Kickit 2.0 (on a 1.3 disk)->2.0 Workbench. The disadvantage is of course that the three disk boot is more tedious than the two disk boot before, and even on warm boots you have to use two disks. Now doubt someone will make a CPU daughterboard that has a 68000 socket and a 2.0 ROM socket but that would be a hack. The second point where Sean is incorrect is a bit more insidious because it implies that 2000 and 500 owners won't be able to upgrade. This is also not the case. The A3000 requires two ROMs because it has a 32bit data bus and would really like to read in data 32 bits at a time. There are no 32 bit ROMs so the 3000 uses two 16bit Roms to store the data. On the 2000 and 500 though they have a 16 bit databus and yes, I believe there are 512K X 16 ROMs which presumably plug into the same 40 pin socket as the existing 256K X 16 ROM. If not, I _know_ there are 272048 EPROM's that could be used instead in a daughterboard arrangement. If you are wondering why the delay, this may very well be it. (developing the daughterboard) Finally, on 2.0 and A1000 compatibility and "why can't I use it...?" There is no reason why you cannot use 2.0 on a 1000. The caveats are that is will be more inconvienient (two disk warmboot) and all of the new Amiga 2000s and A500Ps are shipping with 1M of chip RAM. Eventually these systems will compose a larger installed base than the 512K A1000 and A500 systems (I know it will be a while) but some applications (mostly games) will begin to require 1M of chip ram. This will probably be 2 years down the road but it will happen eventually. Some productivity applications such as word processors will ask for a new video mode and fail if they can't get it. These applications won't work on the A1000 or unexpanded A500s either. Again, it will be awhile since the installed base will force 512K/noECS support. The bottom line though is that as people become familiar with 2.0 and the number of 1M+/ECS systems grow they will begin to take advantage of those features more heavily and 1000 owners and older 500/2000 owners will be forced to choose between upgrading and foregoing. All the introduction of 2.0 and ECS means is that _it will_ happen someday. How long that will be is anyone's guess. My guess is 2 years. -- --Chuck McManis Sun Microsystems uucp: {anywhere}!sun!cmcmanis BIX: <none> Internet: cmcmanis@Eng.Sun.COM These opinions are my own and no one elses, but you knew that didn't you. "I tell you this parrot is bleeding deceased!"
new@udel.EDU (Darren New) (07/11/90)
In article <138659@sun.Eng.Sun.COM> cmcmanis@stpeter.Eng.Sun.COM (Chuck McManis) writes: >There is no reason why you cannot use 2.0 on a 1000. The caveats are Is there also not a caveat that the disk to kick from 1.3 into 2.0 is not available? Will CBM release KS2.0 on floppy at all? Will CBM release the program needed to kick such a floppy? I have a friend from which I could probably get the programs (illegally), but I suspect that the vast majority of loyal A1000 owners won't have such an option. Just the ability to kick 2.0, however inconveniently, so that I could test my programs under 2.0, would be nice. Give me a year, and I'll be out of here and have a job where I can afford a new A3000, so the question of obsolecence in 3 years is not a problem for me. -- Darren
cmcmanis@stpeter.Eng.Sun.COM (Chuck McManis) (07/12/90)
In article <1990Jul11.232831.28400@cbnewsm.att.com> (Neil Weinstock) writes: >Hmm, that'd be 8 Mbits. I was not aware any such thing was available, >certainly not in production quantities. I'd be happy to be proved wrong, >though... Yeah, I flinched too after I multiplied it out. However, remember we are talking _roms_ and not either RAMs or EPROMs. With 1u Technology you can build a diode in a little bit over 1.8 u square. With two level metal CMOS that is potentially 250,000 bits per square mil and 8M bits in about 33.5 sq mils. Drivers, row and column decoders and the selection circuits will add some space but it isn't inconceivable at all. Just remember you don't need any transistors/cell in a masked rom and it makes more sense. I suppose someone could always call Toshiba or someone and ask. :-) -- --Chuck McManis Sun Microsystems uucp: {anywhere}!sun!cmcmanis BIX: <none> Internet: cmcmanis@Eng.Sun.COM These opinions are my own and no one elses, but you knew that didn't you. "I tell you this parrot is bleeding deceased!"