Re: numbers, resizing, etc.
To be honest, you have tested this thing much more strenuously than I have. I fully expect that all of these bugs will be reproducible here, and I'll schedule time next week to look at them. Recently eradicated bugs have been about an even split between stupid programming errors and toolbox confusion. And while in both cases the going is slow, I'm far from discouraged!
How hard would it be to display this PICT instead of the current About Alex dialog?
Dunno, but I'll commit to figuring it out and including anything like this you can come up with (medium priority). That globe is *cool*. As an FYI, I first displayed the globe in 16 colors, and had just a little bit of trouble reading some of the text. Switching to 256 colors improved things dramatically. Is your 16 bits equivalent to my 256 colors, or are you capable of one (or more) steps up from my display quality?
Re: JPEG, etc.
I think I recently threw out some such program. Currently, I'm reduced to displaying your images through TeachText. Can you recommend any shareware to fill the gap? I'm not eager to fork out for a copy of PhotoShop.
Re:try varying the algorithm by beginning with globs instead of sparks. That is, instead of tossing single pixel sparks, toss 5 by 5 globs so that the neo-continents get off to a meatier start. Variable-sized globs?
I've moved this up to near the top of the wishlist. What do you mean by variable-sized globs: user-specifiable size, size ranges (e.g., 1-5 sparks per glob, randomly chosen), or something else?
Re: Macro pixels, etc.
Each cell location is 10 pixels square in recent shipments. This seemed a reasonable compromise during the first level of world mapping, but is probably a poor long-term size. My current code uses 14 pixels on a cell side (this has the added benefit of making the cell numbers resolve much better), which seems to be a more viewable size. While the size is adjustable by changing one parameter at compile time, I'm guessing other design issues may encourage us to pick one size and stick to it. If 14 is it, we'll try 49 micro-pixels per cell first. This is 4 micro-pixels per physical pixel. Keeping this ratio a whole number lets me avoid any tricky/expensive mapping issues.