ADFFS 2.29 beta

Discuss ADFFS development and download test releases
JonAbbott
Posts: 2938
Joined: Thu Apr 11, 2013 12:13 pm
Location: Essex
Contact:

Re: ADFFS 2.29 beta

Post by JonAbbott »

Vanfanel wrote:Pacmania working here on my Rpi. ROM version is BETA Rpi RC-11 (2013-07-08).
However, it has a LOT of tearing. No smooth scroll as expected. My monitors DO support 50Hz modes (I use Minimig on them).
Isn't there no way around this tearing? Smooth scroll was one of the main computer (Archie, Amiga, etc) trademarks in the 90's.
You shouldn't get tearing as the frame blit is done on the GPU's VSync. Can you confirm you have fake vsync enabled in config.txt.
Vanfanel wrote:On the other hand, Zarch doesn't work at all: I mount the disk, then click on the Zarch icon and I get "Error on data transfer". Maybe I need a Zarch image with "!boot" on it?
Are you launching Zarch with the Obey file (F1040201) from !ADFFS.Obey?

Ginger1 wrote:When I get to point 2, I get an "Application May Have Gone Wrong" error, and when I click on describe, it says "Abort on Data Transfer at &FC201668".
Check what version of CLib you have loaded, ADFFS will try to update it if it's too old which doesn't work on newer OS's that use C.
Ginger1 wrote:Just to investigate, I tried another disk image (Jahangir Khan's squash) which ADFFS does accept and I can then reboot using Boot Floppy.
Did you download the full archive or just the disc image?

You could eMail me the contents of your SD's DOS partition, I'll see if I get the same problem with your ROM: jon at jaspp dot org dot uk
JonAbbott
Posts: 2938
Joined: Thu Apr 11, 2013 12:13 pm
Location: Essex
Contact:

Re: ADFFS 2.29 beta

Post by JonAbbott »

grannyg wrote:Conqueror almost worked. The demo plays ok but aborts on trying to start a game.
It doesn't sound like you have it running under the JIT. I've just checked the original, which works. I've updated the Boot script, which is as follows:

Code: Select all

ADFS
DIR ADFS::0.$

IF ADFFS$CPUID>=&A10 THEN ADFRemapVideoMemory 13 160 ELSE ScreenSize 160
ADFForceVSync 4
WimpSlot -min 640K -max 640K
LOAD Conqueror
IF ADFFS$CPUID>=&A10 THEN GOARM3JIT 43800 ELSE GO 43800
IF ADFFS$CPUID>=&A10 THEN ADFRemapVideoMemory
ADFForceVSync 0
Obviously, that's for the original protected floppy. You'll need to change the LOAD and GO statements for the version you have.
Vanfanel
Posts: 576
Joined: Mon Sep 16, 2013 12:01 am

Re: ADFFS 2.29 beta

Post by Vanfanel »

Jon: I'm trying to load Zarch with the F1040201 obey file, but I can't get it working.
What I do is:
-Launch ADFFS
-Mount the Zarch ADF
-Copy over the F1040201 obey file to the !Zarch directory
-Double click on the Obey file

And I get a "file Zarch not found error", but the executable file "Zarch" is in the same directory as the F1040201 Obey file.
I also tried copying the !Zarch app to the SD and putting the F1040201 obey into !Zarch, with identical results.
What am I doing wrong here?
JonAbbott
Posts: 2938
Joined: Thu Apr 11, 2013 12:13 pm
Location: Essex
Contact:

Re: ADFFS 2.29 beta

Post by JonAbbott »

Vanfanel wrote:Jon: I'm trying to load Zarch with the F1040201 obey file, but I can't get it working.
Did you amend the Obey file? All the Obey files we release are written for original versions of the games, with their disc protection intact. I'm afraid I can't assist with pirated copies of games, as one of the fundamental aspects of the project is to gain permission to release the originals legally.
Vanfanel
Posts: 576
Joined: Mon Sep 16, 2013 12:01 am

Re: ADFFS 2.29 beta

Post by Vanfanel »

Ok, forget about that game, then. I'm not buying a possibly unusable floppy disk containing a game from 1987 that wasn't even released in my country. That's just plain ridiculous. I have the game on Amiga anyway.

About Pacmania, I have the default fake vsync isr option set on my config.txt
Maybe we're using different RISC OS ROM versions? As I said, mine is BETA Rpi RC-11 (2013-07-08), wich is available for download here: https://www.riscosopen.org/zipfiles/mis ... 1373304198
grannyg
Posts: 18
Joined: Fri Dec 13, 2013 3:47 pm

Re: ADFFS 2.29 beta

Post by grannyg »

JonAbbott wrote:
grannyg wrote:Conqueror almost worked. The demo plays ok but aborts on trying to start a game.
It doesn't sound like you have it running under the JIT. I've just checked the original, which works. I've updated the Boot script, which is as follows:
Conqueror is working. I made the mistake of thinking pressing 'Return' starts the game. It just so happens that pressing 'Return' (which brings up the map) always causes an abort or a lockup.
JonAbbott
Posts: 2938
Joined: Thu Apr 11, 2013 12:13 pm
Location: Essex
Contact:

Re: ADFFS 2.29 beta

Post by JonAbbott »

grannyg wrote:Conqueror is working. I made the mistake of thinking pressing 'Return' starts the game. It just so happens that pressing 'Return' (which brings up the map) always causes an abort or a lockup.
I'll check and resolve the issue
JonAbbott
Posts: 2938
Joined: Thu Apr 11, 2013 12:13 pm
Location: Essex
Contact:

Re: ADFFS 2.29 beta

Post by JonAbbott »

Vanfanel wrote:About Pacmania, I have the default fake vsync isr option set on my config.txt
Maybe we're using different RISC OS ROM versions? As I said, mine is BETA Rpi RC-11 (2013-07-08)
I'll test with this build. Many changes have been made to the graphics subsystem on the Pi to better support games, so you may need to update.

I'll confirm later today, could you please zip and email the contents of your SD DOS partition so I have you exact BIOS and RO settings. Also confirm your monitor is set to Auto.
Vanfanel
Posts: 576
Joined: Mon Sep 16, 2013 12:01 am

Re: ADFFS 2.29 beta

Post by Vanfanel »

ok, Jon. E-mail sent. And yes, my monitor is set to auto.
JonAbbott
Posts: 2938
Joined: Thu Apr 11, 2013 12:13 pm
Location: Essex
Contact:

Re: ADFFS 2.29 beta

Post by JonAbbott »

grannyg wrote:Conqueror is working. I made the mistake of thinking pressing 'Return' starts the game. It just so happens that pressing 'Return' (which brings up the map) always causes an abort or a lockup.
It's working on StrongARM, although using the 32-bit JIT on StrongARM produces the same result so it's an issue with one of the 32-bit codelets. I'm seeing a similar issue with Magic Pockets, so am now double checking codelet's that affect the PSR.

EDIT: I'd forgotten to turn off the code that forces the CPU to 32bit, with that off both Conqueror and Magic Pockets work on StrongARM using the ARM11 version of the codelets, so they're not at fault.

However, from debugging on the Pi, both games crash when in an IRQ so it could be the hypervised exit, I'll continue investigating tomorrow.
Post Reply