Since I got another calculator to play around with, I decided to make a simple, 2 player game over a link cable like nibls, where you both try to avoid each other.

The program successfully manages to decide if it is player 1, or player 2 and will exit if another player isn't found within a certain amount of time (IDK exactly; about 7 seconds?). However, once you get to the part where you actually play the game, the other player moves, but doesn't change direction, even if they press a key to do so. In addition, it moves in a direction different from the default direction, so some type of data has been received.

Here is some pseudo-code for what I'm trying to do:

;establish what player
draw screen
keyloop:
;adjust your own direction based on key presses
::if player1
::::halt to ensure that the other calc is ready to receive
::::send both low ($D3) to tell the other calc you are ready (it should be waiting in a loop for both to be set to low)
::::send your direction (sending 0-3; this should be a cheap way to avoid complicated sending)
::::start waiting for the other calc to send both low
::::halt to give the other calc enough time to send it's direction
::::read in the data from the port & save it to player 2's direction
::if player2, do the same thing, just in the opposite order
::advance the "snakes" & update screen (and check to see if someone has won)

Any help would be greatly appreciated; I have posted the entire source code below (converted from calcz80 to computerz80 code).

EDIT: This is the part of code that has the key loops, the part that is most likely causing trouble, below it is the entire source.


Code:



Code:

I can't be sure, but it looks to me like you think the link hardware works in some way it doesn't..
http://wikiti.denglend.net/index.php?title=83Plus:Ports:00

Yeah; in that code, I am using the link port slightly incorrectly in that I need to XOR the data (i.e. direction of the other player) I'm receiving with 11. The old version simply waited until $D0 was sent from the other calculator, instead of $D3, which happened once the other program got to a b_call or reset it's own port. Thus, the calc always thought that the other player was moving up.

However, I'm still having some issues now, but they don't seem consistent. Sometimes, the snakes move a while before stopping (as in about 20 pixels; it isn't constant), but sometimes, they freeze immediately and never move. At first, I thought that the problem was using HALT, because it doesn't wait constant time and might not wait very long at all, but the problem persisted when I switched to using a series of PUSH HL / POP HL commands

As a related note, the guide I was initially using for the link port tutorial is wrong because it says that when reading from the link port, you use bits 2 and 3, but after a lot of experimentation with calcsys and several RAM clears, I found that you use bits 0 and 1 for both sending and receiving...

Most linking tutorials have some inaccuracies. Read this page on WikiTI for definitive, correct information instead:

http://wikiti.denglend.net/index.php?title=83Plus:Ports:00