Symplex
Kennt sich schon aus
- Mitglied seit
- 20 Dezember 2011
- Beiträge
- 123
- Punkte
- 0
STATUS: BETA / ALPHA you should know what you are doing !
Requirements:
Amiga (ECS) minimum 4MB fastram with MMU and a Indivision ECS (Core 1.1) installed !
IndivisionECS: http://www.icomp.de/products/indivision_ECS_e.htm
MMULIB: http://aminet.net/package/util/libs/MMULib
Picasso96: http://aminet.net/package/driver/video/Picasso96
IndivisionCore 1.1 https://www.a1k.org/forum/showpost.php?p=494009&postcount=32
Installation:
- Install MMULib + Picasso96 (Version libs:picasso96.library min Version 2.3!)
- Replace incompatible MMU programms with the MMULib equivalents.
- Copy all files of the install archive to the equivalent directorys
FAQ:
Q: I can't boot anymore. I can only start via boot-menü and choosing "no startup-sequence"
A: Press both mouse buttons, the picasso system is not activating in such a situation !
Q: How to uninstall?
A: The Indivision driver can be deleted in the devs:monitors directory. After that only the picasso96 system and the mmu libraries stay inside. Picasso96 is activated on starting a driver in the devs:monitors directory (for example uaegfx, domino ..). Delete all "NEW" drivers installed by picasso96. And please don't touch your PAL, NTSC drivers.
Q: I don't have a indivisionECS installed, but something happens.
A: There is not HARDWARE detection inside. Your should know what you are doing.
Q: Did you support the AGA MK#2?
A: Nobody knows right now, if it is possible. We must wait for it.
Q: Is it possible to play AGA games on a ECS/OCS Amiga ?
A: No its not possible. Maybe you can "play" a RTG game. But it is maybe a little bit slow !
Q: Why should i backup my system?
A: Don't blame me if something goes wrong. I can only test on one amiga with one config! And there is everything ok.
Q: ACATune didn't find my accelerator any more?
A: That's because of the MMU Library. Its blocking by default the communication area to the accelerator config area. Use the MuSetCacheMode Tool to correct this. (Part of MMU.Library package)
MuSetCacheMode Adddress=$b8f000 SIZE=4000 WRITETHROUGH VALID NOCACHE
Test/Developing configuration:
- Amiga 600 / ACA630 / Indivision ECS
Reported working configurations:
2* A600 / ACA630
1* A2000 / Blizzard2060
1 of 2 Apollo Accelerators (030)
Reported NOT working configurations:
1 of 2 Apollo Accelerators (030)
Benchmarking:
Benchmarking is a big problem with the indivision, because it doesn't reflect reality. But you can see, that it is possible to work with it. There is optimization possible, but don't expect factors of 2 or 3. I think its only possible in lower percentage area. Maybe Jens has a few ideas
.
Here are the results of a 030MMU/25Mhz Test: http://dl.dropbox.com/u/54065139/PICT9985_PBENCH.MOV
PBench is saying about 30-40% of the speed of a native Screen A4000/AGA 060. I cant see this is correct (feeling is, that it is slower) but maybe i'm to much PC "windows"-speed infected.
Version History
0.2w 11. April 2012 (Still debugging) MMU DETECTION !!
0.2v 11. April 2012 (Still debugging) 68030 ONLY! -MMU FLUSH !!
0.2u 11. April 2012 (Still debugging) 68030 ONLY! -MMU FLUSH !!
0.2s 11. April 2012
Totally new switching routine... absolute stable on my machine.. Question is it the same on Apollo?
Add this as first command in startup-sequence:
libs:picasso96/indi.card (if it cant find the indivision.. he reports it and you can reboot faster.. i didnt check this later !!!)
0.2r 10. April 2012
More stable switching, Screenmode can be changed.
Additional Picasso96settings file to be copied in the "DEVS:" directory supporting now all indivision.library screenmodes
=> for this and the following versions, please dont modify the RESOLUTION INFORMATIONS anymore (look 2 postings below)
0.2g // 7. April 2012 [Still Debugging for a single user - not posted]
0.2e // 6. April 2012 [Still Debugging for single users]
New INDI switching (alternative) Routine coded. (indi.card only)
-Activated if following file exists: DEVS:INDI_SWITCH_VAR1
0.2c // 6. April 2012
Based on a BUG Report, this version uses a resolution of 800*600 (if it couldnt detect the right one). In the last version it used 1024*768, which didnt work for a lot of people. (indi.card only)
0.2 // 6. April 2012
- Debugging messages
- First support of native Amiga screenmode switching (code switching is not stable, because of timing issues for example startscreen of indivision is coming up and switching back didnt work sometimes)
"Workaround: dont close the native screen and switch via AMIGA N/M if it doesnt work a second/third..... time, until you get back on Workbench."
0.1b // 2. April 2012
- MMU Version
0.1a // 1. April 2012
- No MMU Version
Known Problems:
- One user reports that it take 2 minutes until screen change to indivision take place.
- Mouse pointer not visible (Missing activated tooltype in devs:indi.info something like "SoftwareSprites=Yes")
- Ghosty Mousepointer (looks strange) or call it Ice-Mousepointer (no clue why) => old Picasso Version ! Update to last Aminet-Version
- Screenmode can't be used.. look 2 postings below, how to configure.
*** If you find typing errors, you can take them with you. ***
*** Wer Tippfehler findet, darf diese behalten
***
Requirements:
Amiga (ECS) minimum 4MB fastram with MMU and a Indivision ECS (Core 1.1) installed !
IndivisionECS: http://www.icomp.de/products/indivision_ECS_e.htm
MMULIB: http://aminet.net/package/util/libs/MMULib
Picasso96: http://aminet.net/package/driver/video/Picasso96
IndivisionCore 1.1 https://www.a1k.org/forum/showpost.php?p=494009&postcount=32
Installation:
- Install MMULib + Picasso96 (Version libs:picasso96.library min Version 2.3!)
- Replace incompatible MMU programms with the MMULib equivalents.
- Copy all files of the install archive to the equivalent directorys
FAQ:
Q: I can't boot anymore. I can only start via boot-menü and choosing "no startup-sequence"
A: Press both mouse buttons, the picasso system is not activating in such a situation !
Q: How to uninstall?
A: The Indivision driver can be deleted in the devs:monitors directory. After that only the picasso96 system and the mmu libraries stay inside. Picasso96 is activated on starting a driver in the devs:monitors directory (for example uaegfx, domino ..). Delete all "NEW" drivers installed by picasso96. And please don't touch your PAL, NTSC drivers.
Q: I don't have a indivisionECS installed, but something happens.
A: There is not HARDWARE detection inside. Your should know what you are doing.
Q: Did you support the AGA MK#2?
A: Nobody knows right now, if it is possible. We must wait for it.
Q: Is it possible to play AGA games on a ECS/OCS Amiga ?
A: No its not possible. Maybe you can "play" a RTG game. But it is maybe a little bit slow !
Q: Why should i backup my system?
A: Don't blame me if something goes wrong. I can only test on one amiga with one config! And there is everything ok.
Q: ACATune didn't find my accelerator any more?
A: That's because of the MMU Library. Its blocking by default the communication area to the accelerator config area. Use the MuSetCacheMode Tool to correct this. (Part of MMU.Library package)
MuSetCacheMode Adddress=$b8f000 SIZE=4000 WRITETHROUGH VALID NOCACHE
Test/Developing configuration:
- Amiga 600 / ACA630 / Indivision ECS
Reported working configurations:
2* A600 / ACA630
1* A2000 / Blizzard2060
1 of 2 Apollo Accelerators (030)
Reported NOT working configurations:
1 of 2 Apollo Accelerators (030)
Benchmarking:
Benchmarking is a big problem with the indivision, because it doesn't reflect reality. But you can see, that it is possible to work with it. There is optimization possible, but don't expect factors of 2 or 3. I think its only possible in lower percentage area. Maybe Jens has a few ideas
.Here are the results of a 030MMU/25Mhz Test: http://dl.dropbox.com/u/54065139/PICT9985_PBENCH.MOV
PBench is saying about 30-40% of the speed of a native Screen A4000/AGA 060. I cant see this is correct (feeling is, that it is slower) but maybe i'm to much PC "windows"-speed infected.
Version History
0.2w 11. April 2012 (Still debugging) MMU DETECTION !!
0.2v 11. April 2012 (Still debugging) 68030 ONLY! -MMU FLUSH !!
0.2u 11. April 2012 (Still debugging) 68030 ONLY! -MMU FLUSH !!
0.2s 11. April 2012
Totally new switching routine... absolute stable on my machine.. Question is it the same on Apollo?
Add this as first command in startup-sequence:
libs:picasso96/indi.card (if it cant find the indivision.. he reports it and you can reboot faster.. i didnt check this later !!!)
0.2r 10. April 2012
More stable switching, Screenmode can be changed.
Additional Picasso96settings file to be copied in the "DEVS:" directory supporting now all indivision.library screenmodes
=> for this and the following versions, please dont modify the RESOLUTION INFORMATIONS anymore (look 2 postings below)
0.2g // 7. April 2012 [Still Debugging for a single user - not posted]
0.2e // 6. April 2012 [Still Debugging for single users]
New INDI switching (alternative) Routine coded. (indi.card only)
-Activated if following file exists: DEVS:INDI_SWITCH_VAR1
0.2c // 6. April 2012
Based on a BUG Report, this version uses a resolution of 800*600 (if it couldnt detect the right one). In the last version it used 1024*768, which didnt work for a lot of people. (indi.card only)
0.2 // 6. April 2012
- Debugging messages
- First support of native Amiga screenmode switching (code switching is not stable, because of timing issues for example startscreen of indivision is coming up and switching back didnt work sometimes)
"Workaround: dont close the native screen and switch via AMIGA N/M if it doesnt work a second/third..... time, until you get back on Workbench."
0.1b // 2. April 2012
- MMU Version
0.1a // 1. April 2012
- No MMU Version
Known Problems:
- One user reports that it take 2 minutes until screen change to indivision take place.
- Mouse pointer not visible (Missing activated tooltype in devs:indi.info something like "SoftwareSprites=Yes")
- Ghosty Mousepointer (looks strange) or call it Ice-Mousepointer (no clue why) => old Picasso Version ! Update to last Aminet-Version
- Screenmode can't be used.. look 2 postings below, how to configure.
*** If you find typing errors, you can take them with you. ***
*** Wer Tippfehler findet, darf diese behalten
***
