BLuRry

When I look at the trace, I see this prior to the loader crash:

A0 38 0C 01E3 ..RB..ZC 8b0 : STA $c00c
A0 38 0C 01E3 ..RB..ZC 8b3 : STA $c000
A0 38 0C 01E3 ..RB..ZC 8b6 : JMP $b600
A0 38 0C 01E3 ..RB..ZC b600 : BBS5 $a0,$R
A0 38 0C 01E3 ..RB..ZC b602 : LDY #$a0
A0 38 A0 01E3 N.RB...C b604 : LDY #$a0
A0 38 A0 01E3 N.RB...C b606 : LDY #$a0
A0 38 A0 01E3 N.RB...C b608 : LDY #$a0
A0 38 A0 01E3 N.RB...C b60a : LDY #$a0
A0 38 A0 01E3 N.RB...C b60c : LDY #$a0
A0 38 A0 01E3 N.RB...C b60e : LDY #$a0
A0 38 A0 01E3 N.RB...C b610 : ???
A0 38 A0 01E3 N.RB...C b611 : CMP $c3
A0 38 A0 01E3 N.RB...C b613 : BBS4 $ce,$R
A0 38 A0 01E3 N.RB...C b5d9 : ???
A0 38 A0 01E3 N.RB...C b5da : ADC $3173,Y
F6 38 A0 01E3 N.RB.... b5dd : ROL $73,X
F6 38 A0 01E3 ..RB.... b5df : AND $(36),Y
12 38 A0 01E3 ..RB.... b5e1 : BVS $b614
12 38 A0 01E3 ..RB.... b5e3 : BRK #$00
BRK at $b5e5

Obviously BBS5 is not a valid opcode in this case -- so why would flobynoid jump to this address when there's no program code there? I see it activate 80 col and 80 store prior to this -- I didn't munge the softswitch implementation for aux reads did I?

-B

Antoine Vignau

The version 9.8 available at http://www.brutaldeluxe.fr/public/ boots
well... feel free to compare with yours,
antoine

BLuRry

Well, I didn't support 2mg 5.25 images. Ok, so added support for 2mg 140kb floppy images (I'm assuming they are always prodos-ordered?)

Same crash unfortunately -- Is there a good breakdown of how the bootloader works in this game or is it kept secret? ;-D

-Brendan

BLuRry

Well, I didn't support 2mg 5.25 images. Ok, so added support for 2mg 140kb floppy images (I'm assuming they are always prodos-ordered?)

Same crash unfortunately -- Is there a good breakdown of how the bootloader works in this game or is it kept secret? ;-D

-Brendan

Deckard

> Same crash unfortunately -- Is there a good breakdown of how the bootloader works in this game or is it kept secret? ;-D

My patched version works with Jace:
http://boutillon.free.fr/Underground...pplewin.dsk.gz

The original boot1 doesn't check the address fields...

JM

BLuRry

LOL! yeah, I forgot that there was a patched version for AppleWin already.Still though, what is the major difference? I'm willing to tweak things to suit the original version if it's not a major difference.

Thank you for sending up this link to the patched version -- it's great to see it working 100% with mouse support in JACE -- and because Flobynoid uses clamping to get the mouse position, the paddle is perfectly synced with the mouse cursor making the game extremely playable!!

Now if I can fix a few annoying miscellaneous crashes I'll be really happy with this beast.

Michael J. Mahon

BLuRry <brendan.robert@gmail.com> wrote:
> When I look at the trace, I see this prior to the loader crash:
>
> A0 38 0C 01E3 ..RB..ZC 8b0 : STA $c00c
> A0 38 0C 01E3 ..RB..ZC 8b3 : STA $c000
> A0 38 0C 01E3 ..RB..ZC 8b6 : JMP $b600
> A0 38 0C 01E3 ..RB..ZC b600 : BBS5 $a0,$R
> A0 38 0C 01E3 ..RB..ZC b602 : LDY #$a0
> A0 38 A0 01E3 N.RB...C b604 : LDY #$a0
> A0 38 A0 01E3 N.RB...C b606 : LDY #$a0
> A0 38 A0 01E3 N.RB...C b608 : LDY #$a0
> A0 38 A0 01E3 N.RB...C b60a : LDY #$a0
> A0 38 A0 01E3 N.RB...C b60c : LDY #$a0
> A0 38 A0 01E3 N.RB...C b60e : LDY #$a0
> A0 38 A0 01E3 N.RB...C b610 : ???
> A0 38 A0 01E3 N.RB...C b611 : CMP $c3
> A0 38 A0 01E3 N.RB...C b613 : BBS4 $ce,$R
> A0 38 A0 01E3 N.RB...C b5d9 : ???
> A0 38 A0 01E3 N.RB...C b5da : ADC $3173,Y
> F6 38 A0 01E3 N.RB.... b5dd : ROL $73,X
> F6 38 A0 01E3 ..RB.... b5df : AND $(36),Y
> 12 38 A0 01E3 ..RB.... b5e1 : BVS $b614
> 12 38 A0 01E3 ..RB.... b5e3 : BRK #$00
> BRK at $b5e5
>
> Obviously BBS5 is not a valid opcode in this case -- so why would
> flobynoid jump to this address when there's no program code there? I see
> it activate 80 col and 80 store prior to this -- I didn't munge the
> softswitch implementation for aux reads did I?
>
> -B

Actually, a bank switching problem could easily cause such a symptom.

-michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon

BLuRry

I was thinking the same thing -- I reviewed the code and found that I lazily didn't implement the LCWRITE switches to only take effect until the second write access. But even after implementing that it still crashes the sameway. :-( Is there a good test suite for this sort of thing?