Oh, so *that's* it, a frickin' page boundary....I must've
pushed things around when I added in all the 16 bit math.
it's funny, I used to be really paranoid about page boundaries...
I made up a perl script to analyze and find cross-page jumps,
but then I realized it was other, more basic programming errors
(usually forgetting the # when using constant values) that was
causing my issues and I stopped using it...obviously issues w/
stuff happening in the VBLANK are a lot more reselient to page
boundaries than my tightly wound Kernal...
Guess I should do some .ORGing to put my kernal at the begining of
page, and then JMP to that location as the last instruction at the very
end of the VBLANK? Is that a good method for dealing with this,
easier than rejiggering a kernal?
Thanks for looking at this! I was begining to despair...
>
> Hopefully this will sort in the archives correctly. This isn't a true reply as the message isn't in my inbox anymore.
>
> > So I'm having trouble with my kernal...but I can't understand why I'm seeing
> > it *now*, everything I've been changing is in VBLANK (the fractional
> > positioning, and I just added in a routine to rebound off the floor at 1/2
> > the descending speed) and the Kernal, which seemed pretty solid before,
> > should be unaffected.
>
> It seems the problem is your kernel loop is taking longer than you expected. Counting the cycles at the maximum cycle count shows the last branch happening at cycle 73. The loop branch crosses a page boundary so that's 4 cycles vs. 3. Doing this produces an extra scan when all your objects are on the same scan line.
>
> Take care,
> Dennis
>
> ----------------------------------------------------------------------------------------------
> Archives (includes files) at http://www.biglist.com/lists/stella/archives/
> Unsub & more at http://www.biglist.com/lists/stella/
--
KirkIsrael@xxxxxxxxxxxxx http://kisrael.com
"Sisyphus has a sense of playfulness [...]
you have to look at it from the rock's point of view."
--Pointy Haired Boss, Dilbert (TV)
----------------------------------------------------------------------------------------------
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/
