I was recently informed of a new bug in my old game, sybil's tail. after doing some research into it, I discovered the cause. I had a global variable set up that could be set at any time to cause the screen to "flash" once, on the next frame. presumably to conserve tokens, I wrote it like this:

flash=flash and rectfill(0,24,128,104,7)

this line calls rectfill if flash is true, and assigns the result of rectfill to flash. as long as rectfill returns a falsy value, we're good.

now. I understand this line will send me to coding hell. but it seems, as of 0.2.2c, rectfill and rect return 0, which is not a falsy value. if this is an intentional feature, I will own up to my sins, patch the game, and move on-- I'm not going to demand the API honor such a silly corner case. but given that circ and circfill still return nil, I suspect it is in fact a bug

if I could get a confirmation whether rectfill returning 0 is intended behavior, that'd be much appreciated. I may still patch the game, but if the problem isn't going to go away, I'd probably opt to do it sooner rather than later

It looks like pset() returns 0 now, too. No other drawing function seems affected.