Re: [stella] button questions

Subject: Re: [stella] button questions
From: KirkIsrael@xxxxxxxxxxxxx
Date: 17 Jul 2002 16:06:11 -0000
> On Tuesday, July 16, 2002, at 10:37 PM, Chris Wilkson wrote:
> Once you've substituted BMI, try putting 1 back into D6 of VBLANK 
> just like you'd done before.  I bet that works.  I think more 
> recent versions of my code have an okay joystick read routine, 
> including the buttons.

I thought I tried it both ways (setting D6 to 1, keeping it at 0)
after getting the BMI mojo working but neither seemed to effect the 
behavior I observed on my test, which polled the button every 
screendraw...but I'm not positive what I was hoping to see, so 
it wasn't a very good test.

> Eckhard further explained that the latch bit can get more 
> complicated if your kernel only checks for button presses every, 
> say, third screen draw (as I understood him).  If you throw down 
> the latch, you can preserve the button press until you deal with it.

Ah. Huh.

So you got the impression that it's safest to slap a #%01000000
not just a #0 in there?

My JoustPong program, which I assembled by inspection of other code 
and experimentation, not always real understanding, has a pretty 
functional routine, I only want to register each button press once,
so I have a flag that amounts to "ButtonWasPressedLastScreendraw"...
if the button is released this flag is cleared, if the button is 
pressed it checks the flag and only does the flap if the flag was
> But I'm not doing anything that fancy, and am not *quite* sure how 
> that's helpful yet.  Insert "young grasshopper" reference (meaning 
> me in this case).

You and me both.
Of course I've been "procrastinating" on JoustPong with this work
on 2600 101, as well as YANMAPGE (Yet Another Newbie Making A Player
Graphic Editor)--actually, I have a pretty slick javascript page,
one for players, one for playfields, and it works *terrific* locally,
but when I moved it to a real server I found out that IE was making 
a round trip for all of the damn 8x8=64 squares (or whatever) if the 
user changes the bgcolor, even though they all referred to the same 
damn .gif file! If I can't get a caching routine working letting them
all refer to the same loaded file, I might have to be more DHTMLy and 
change the color of table <TD>s (possibly behind transparent GIFs as

PROCRASTINATION "Hard work often pays off after time, 
                  but laziness always pays off now."

Archives (includes files) at
Unsub & more at

Current Thread