>a. 2600 Jr. serial #860746015 draws lines;
> 2600 Jr. serial #870014229 doesn't.
>b. The lines appear in column 24 or 25. (This assumes you break the
> screen up into blocks the size of the space between invaders.)
You sure its 24 or 25? I'd think it would only happen in 25, because
column 25 (counting from 1) is the first column of the second copy of PF1,
which apparently isn't getting updated in time for the first pixel of PF1
on one system but is on the other. The lower serial number model responds
to PF1 writes one cycle (possibly as low as one pixel) later than the other
2600jr and my 4-switch 2600.
Try some other games that rewrite the playfield (nonrepeating,
nonreflected) - Circus Atari, Super/Breakout, Centipede/Millipede, Pac-man
("dots") and see if the weird one acts as such for any of those. It'll
only mess up if the program updates the second copy of PF1 on the last
possible cycle. If those mess up, you can assume your console is bad; if
they work, it's probably a quirk of your particular model.
>c. The lines only appear when there are invaders in column 5.
>d. The lines only appear on the same row where there are invaders in
> column 5. (That is, if the two leftmost columns of invaders have
> five and three invaders in them, respectively, there will be five
> sets of lines as the first columns passes through, but only three
> when the next column passes through.)
Again, that happens because the value in bit 7 of PF1 remains unchanged
from its first copy (column 5) to its second copy (column 25).
>e. If I shoot the bogus lines, the game thinks I have hit an invader.
> This sometimes leads to half-size (width) invaders. The fun part
> is that I can rapidly shoot the lines to make the invaders speed
> up, then move very slowly (by underflowing the invader counter).
> (See Erik's previous note about a similar bug.)
Well, that's because the invaders aren't really stored as invaders; they're
stored as simply raw images of the playfield registers. When a collision
is detected, it checks the missile's coordinates, and zeroes the playfield
bit corresponding to that spot and one on each side of it (so if you hit
the left half of an invader, it zeroes that spot, the already-empty spot to
the left, and the filled spot to the right.. if you hit the right half of
an invader, it zeroes the filled spot to the left and the already-empty
column to the right. Notice that I need not do error-checking for the
edges of the playfield - it's impossible to shoot an invader in the first
or 40th column.)
>So, the questions are:
>
>1. Can anyone tell me what might be wrong with my 2600 Jr.?
Was it made in Taiwan or Korea? The only thing I can think of is a
lower-quality (slower) part used for the RAM inside the TIA that holds the
register contents. (Still, anyone wanna try the game on a Supercharger
plugged into a 7800 or Coleco-with-2600-adapter and see what happens then?)
>2. Am I the only one that liked the orignal player shape better? 8)
> Since the invaders can't look like the arcade version, there's no
> reason for the player's cannon to do so.
I'd rather go for as much authenticity as possible within the 2600's
confines, but I'll leave this up to majority vote =)
--
Archives available at http://www.biglist.com/lists/stella/archives/
E-mail UNSUBSCRIBE in the body to stella-request@xxxxxxxxxxx to be removed.
