d88ricwe@odalix.ida.liu.se (Rickard Westman) (03/23/91)
>My dealer told me :) that A1000 users will *not* be able to use WB 2.0. >Is this likely: > >a. Wrong Yes, it's wrong. It is possible to use 2.0 with A1000 *now*, if you have at least 512K RAM at $200000 and have the disk based version of KS 2.0 distributed to developers. (I am not an official developer, so I don't have it.) If users will be able to run 2.0 in a way supported by Commodore is another question. Commodore *could* provide 2.0 to A1000 owners if they wanted to. This may not be the case, even though I hope they will. >b. Because Kickstart 2.x will be distributed only on ROMs. Commodore will probably prefer users running 2.0 from ROM. >c. Because the Kickstart daugtherboard doesn't have enough DRAM to hold > the larger Kickstart 2.x This is true. Only one half of Kickstart 2.0 would fit in the daughterboard RAM. Commodore could provide a Kickstart disk which loads half the ROM image in WOM and half in RAM @ $200000. Actually, the first half would be loaded into WOM automatically. When the boot ROM code jumps into the WOM, it could check the checksum of the other half. If it was wrong it could be because the RAM image was corrupted or never loaded. In either case, the second half would be loaded from disk. Of course, expansion.library and trackdisk.device would have to be located in the $FC0000 part for this to work. I think this would be a good idea, but chances are Commodore wouldn't think of it as worthwhile. In that case I hope they will be kind enough to make it possible for someone else to provide such a disk, by locating the needed libraries/devices, as well as any unused space in the $FC0000 part. -- Rickard Westman, University of Linkoping, Sweden
f90angu@fy.chalmers.se (Andreas Gunnarsson) (03/23/91)
In article <1991Mar22.170907.26385@ida.liu.se> d88ricwe@odalix.ida.liu.se (Rickard Westman) writes: [stuff deleted] >Only one half of Kickstart 2.0 would fit in the >daughterboard RAM. Commodore could provide a Kickstart disk which >loads half the ROM image in WOM and half in RAM @ $200000. Actually, >the first half would be loaded into WOM automatically. When the boot >ROM code jumps into the WOM, it could check the checksum of the other >half. If it was wrong it could be because the RAM image was corrupted >or never loaded. In either case, the second half would be loaded from >disk. Of course, expansion.library and trackdisk.device would have to >be located in the $FC0000 part for this to work. > >I think this would be a good idea, but chances are Commodore wouldn't >think of it as worthwhile. In that case I hope they will be kind >enough to make it possible for someone else to provide such a disk, by >locating the needed libraries/devices, as well as any unused space in >the $FC0000 part. >-- >Rickard Westman, University of Linkoping, Sweden Why not simply put the KS at $200000 and use the $fc0000-$ffffff memory as RAM. The only drawback is that if any program tries to write at $f80000-$fbffff the RAM at $fc0000 would become ReadOnly, but no program should write in that range anyway. I just came to think about another drawback: the biggest memory chunk would be 256K less, but that shouldn't be too much of a problem. ============================================================================== 73 es 88 de SM7TLS f90angu@fy.chalmers.se Andreas Gunnarsson -.-- --- ..- -.. --- -.- -. --- .-- -- --- .-. ... . ..--. ============================================================================== -- ============================================================================== 73 es 88 de SM7TLS f90angu@fy.chalmers.se Andreas Gunnarsson ==============================================================================
doctorj@en.ecn.purdue.edu (Jeffrey W Davis) (03/24/91)
In article <1991Mar23.035000.11891@fy.chalmers.se> f90angu@fy.chalmers.se (Andreas Gunnarsson) writes: >[stuff deleted] > >Why not simply put the KS at $200000 and use the $fc0000-$ffffff memory as >RAM. The only drawback is that if any program tries to write at $f80000-$fbffff >the RAM at $fc0000 would become ReadOnly, but no program should write in that >range anyway. >I just came to think about another drawback: the biggest memory chunk would >be 256K less, but that shouldn't be too much of a problem. > Since running KS2.0 in an Amiga 1000 requires extra hardware (namely memory) anyway, I have another solution. The Frances board (by Brad Fowles) has a simple logic array that re-maps the KS addresses $fc0000-$ffffff into addresses that are available on the Frances memory board. By pulling up one pin on a 74F244 and connecting it to the un-translated address bit, this can be extended to re-map 512K of KS addresses (or more, 1Meg. KS?). Now, the non-developers version (Shipped with A3000) can be copied into the address space pointed to by the Frances translation circuitry. Initiate the ERKS (Enable Remap Kickstart) and execute the 2.0 ROM image. The only other modification that is necessary is to disconnect the RESET line from the latch that DISables the remapping circuitry. This nearly drove me crazy until I checked the schematics and realized that when the processor asserts a RESET, the kickstart re-mapping is turned off (2.0 asserts a RESET shortly into the ROM). Now, once you have written your own program to shove the KS2.0 into Frances memory and initialized everything correctly (trash the AFM prog)... voiala! You can run 2.0! --- Now for the rest --- A similar address circuit could be made to slip in between the 68000 and socket that would translate KS addresses to an address you have available! This way, as long as C= distrubutes the 2.0ROM on disk... anyone will be able to use it. Disclaimer: Keep in mind that if you intend to put the above idea into commercial production that I will sue... unless you clear it with me FIRST! (SERIOUSLY) *********************************************************** * Jeff Davis * Relax! And get into /// * * doctorj@en.ecn.purdue.edu * the STRESS!!! /// * * * \\\///030 * * * -Gigahertz!- Amiga\XX/ 882 * *********************************************************** -=[ In Stereo Where Available ]=-
<LEEK@QUCDN.QueensU.CA> (03/25/91)
In article <1991Mar24.070304.16395@en.ecn.purdue.edu>, doctorj@en.ecn.purdue.edu (Jeffrey W Davis) says: >> >Since running KS2.0 in an Amiga 1000 requires extra hardware (namely >memory) anyway, I have another solution. > >The Frances board (by Brad Fowles) has a simple logic array that re-maps >the KS addresses $fc0000-$ffffff into addresses that are available on the >Frances memory board. By pulling up one pin on a 74F244 and connecting >it to the un-translated address bit, this can be extended to re-map 512K >of KS addresses (or more, 1Meg. KS?). That works. You'll need new boot rom to be able to boot up normally. > >Now, the non-developers version (Shipped with A3000) can be copied into >the address space pointed to by the Frances translation circuitry. >Initiate the ERKS (Enable Remap Kickstart) and execute the 2.0 ROM >image. > >The only other modification that is necessary is to disconnect the RESET >line from the latch that DISables the remapping circuitry. This nearly >drove me crazy until I checked the schematics and realized that when the >processor asserts a RESET, the kickstart re-mapping is turned off (2.0 >asserts a RESET shortly into the ROM). > Did you got it to work ? >Now, once you have written your own program to shove the KS2.0 into Frances >memory and initialized everything correctly (trash the AFM prog)... voiala! >You can run 2.0! > >--- Now for the rest --- > >A similar address circuit could be made to slip in between the 68000 and >socket that would translate KS addresses to an address you have available! >This way, as long as C= distrubutes the 2.0ROM on disk... anyone will be >able to use it. > Easier way is to stick a 2.0 rom into the expansion port and over-ride the internal decoding circuits over that range. You might have to generate *Dtack on your own, but that's easy. Solution #2 is to hack the WCS board. I did some work last summer. Most of the hardware was done. Someone else was doing the new boot rom. Unfortunately the new rom has some minor bugs, I an not an assembly expert and was in the position to fix it. I have been waiting for the 2.0 to come out so that I can test it out. >Disclaimer: >Keep in mind that if you intend to put the above idea into commercial >production that I will sue... unless you clear it with me FIRST! (SERIOUSLY) Did you come up with the idea first ? Why do you have the authority ?? Sorry. I posted about the Lucas/Frances hack idea at least half a year in comp.sys.amiga.hardware first. (Summer of 1990 Someone tried it and it works) Check it before you sue. I hate lawyers and people suing others for someone else's idea. (Sounds like Apple) If you sue anyone for this, I make sure t your case is thrown out of court and the other party wins. I don't mind people putting it in anywhere as long as it is at a reasonable price for the process. Straightly speaking it was there all along in the original design. It is not any major innovations, not anything the lawyers would risk. Flame off. Now back to regular progtramming and hardware hacking... > >*********************************************************** >* Jeff Davis * Relax! And get into /// * >* doctorj@en.ecn.purdue.edu * the STRESS!!! /// * >* * \\\///030 * >* * -Gigahertz!- Amiga\XX/ 882 * >*********************************************************** > -=[ In Stereo Where Available ]=- K. C. Lee Elec. Eng. Grad. Student
peck@ral.rpi.edu (Joseph Peck) (03/25/91)
In article <1991Mar24.070304.16395@en.ecn.purdue.edu> doctorj@en.ecn.purdue.edu (Jeffrey W Davis) writes: [chomp chomp chomp ...] >Since running KS2.0 in an Amiga 1000 requires extra hardware (namely >memory) anyway, I have another solution. > [swallow swallow swallow ...] >--- Now for the rest --- > >A similar address circuit could be made to slip in between the 68000 and >socket that would translate KS addresses to an address you have available! >This way, as long as C= distrubutes the 2.0ROM on disk... anyone will be >able to use it. > >Disclaimer: >Keep in mind that if you intend to put the above idea into commercial >production that I will sue... unless you clear it with me FIRST! (SERIOUSLY) So you claim to have a patent or copyright on basic MMU technology? Seems pretty unlikely to me.... Besides, I would then counter sue you. I thought of something similar a year ago that would allow kick start roms to be put in a cartridge ( a la 2600 ) and let you plug in whatever version you wanted and switch to it on the fly.... no case opening or chip switching required. Soooooo, tthhhpppt!!!! to you. Maybe your (SERIOUSLY) should switch to a few smileys..... :) > >*********************************************************** >* Jeff Davis * Relax! And get into /// * >* doctorj@en.ecn.purdue.edu * the STRESS!!! /// * >* * \\\///030 * >* * -Gigahertz!- Amiga\XX/ 882 * >*********************************************************** What a great legal system we have.... Joe Peck peck@ral.rpi.edu
doctorj@en.ecn.purdue.edu (Jeffrey W Davis) (03/25/91)
In article <!a5flmb@rpi.edu> peck@ral.rpi.edu (Joseph Peck) writes: >In article <1991Mar24.070304.16395@en.ecn.purdue.edu> doctorj@en.ecn.purdue.edu (Jeffrey W Davis) writes: > >[chomp chomp chomp ...] > >>Disclaimer: >>Keep in mind that if you intend to put the above idea into commercial >>production that I will sue... unless you clear it with me FIRST! (SERIOUSLY) > >So you claim to have a patent or copyright on basic MMU technology? >Seems pretty unlikely to me.... Well, OF COURSE I'm not talking about MY patent on MMU tech. this time :-) But I do have a specific product and design in mind... it would probably go into the public domain should the need arise though (no C= support on A1000). >Besides, I would then counter sue you. I thought of something similar >a year ago that would allow kick start roms to be put in a cartridge >( a la 2600 ) and let you plug in whatever version you wanted and switch >to it on the fly.... no case opening or chip switching required. > >Soooooo, tthhhpppt!!!! to you. Now your idea is totally different! Sooooo, tthhhpppt!!!!! back at ya... WAIT a sec... this isn't alt.flame!?!? > >Maybe your (SERIOUSLY) should switch to a few smileys..... :) Okay, MAYBE I forgot the smileys... >What a great legal system we have.... > Well... I'm not sure how great it is... >Joe Peck >peck@ral.rpi.edu *********************************************************** * Jeff Davis * Relax! And get into /// * * doctorj@en.ecn.purdue.edu * the STRESS!!! /// * * * \\\///030 * * * -Gigahertz!- Amiga\XX/ 882 * ***********************************************************
<DXB132@psuvm.psu.edu> (03/25/91)
>In article <1991Mar24.070304.16395@en.ecn.purdue.edu>, u >doctorj@en.ecn.purdue.ed >(Jeffrey W Davis) says: >> >>Now, the non-developers version (Shipped with A3000) can be copied into >>the address space pointed to by the Frances translation circuitry. Well, a f80,000 version can be copied into that address space. The A3000 version (taken literally) won't work, because it has A3000-specific code. But this is going to become very unimportant when someone makes a relocatable version of 2.0, which is already easy to do. As a simple example, you can make a RamKickable (disk resident) module version of (say) trackdisk.device with a simple program that compares the developer ($200,000) and 3000-Kickstart (f80,000). By the way, the 2.0 trackdisk, when used with the 2.0 disk.resource, works fine under 1.3. At some point we'll have a version that loads as an executable file, whereever you have free memory (although in practice not, because it would fragment your memory). Now if you steal my super-genius just-has-to-be-original ideas I'll sue your mother!!! :-) [Remember : ideas are worth NOTHING. If you want to make money how about making a product] -- Dan Babcock
barrett@jhunix.HCF.JHU.EDU (Dan Barrett) (03/27/91)
In article <91084.021755DXB132@psuvm.psu.edu> DXB132@psuvm.psu.edu writes: >...compare the developer ($200,000) and 3000-Kickstart (f80,000). Holy crow!!! Do you really mean that the developer version of Kickstart costs TWO HUNDRED THOUSAND DOLLARS, and that the "release" Kickstart for the A3000 will cost EIGHTY THOUSAND FRANCS??? Wow, that's realy steep! I think I'll stick with 1.3! :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) Dan //////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ | Dan Barrett, Department of Computer Science Johns Hopkins University | | INTERNET: barrett@cs.jhu.edu | | | COMPUSERVE: >internet:barrett@cs.jhu.edu | UUCP: barrett@jhunix.UUCP | \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////////