No, I mean it will only import 12 rows and 7 columns of tiles even if the pic is the full 96x64 pixels.  All pics can be read from archive.

Nice to hear about the update, and nice to know about pics and rectangles (I also like being able to do 16*16=32 multiplication)
Quigibo, could you check the documentation for the rectangle routines? It sounds like an added ^r does nothing and that RectI( always draws to the backbuffer.
Keep up the good work!

Fine by me, I had guessed the correct uses of the commands, but I was hoping he would clarify it for confused others.

I just tried (or rather am trying) the fractal program. Two words: amazingly awesome!
Keep up the good work!

Sorry about the typos in the commands list, I'll fix it next update.

@Builderboy, yeah, its really easy to get 4 lvl gs for the fractal using mod 3, I had it originally, but i thought it was overkill.

Its pretty interesting actually.  Let me explain using decimal and then you can extrapolate to hexadecimal.

So say you have two 4 digit numbers, always positive.  If you do a multiplication, what is the largest number you can possibly get?  Well, that's simple to find:

9999 x 9999 = 99980001

Notice how we end up with an 8 digit number.  Since our variables can only hold 4 digit numbers, like the input, we have to decide which 4 digits to keep and which ones to throw away.  We have now have 3 methods to choose (only the underlined part is kept):

9999 *   9999 = 99980001
9999 ** 9999 = 99980001
9999 *^ 9999 = 99980001

There is a little more to ** when you have negative numbers because its signed instead of unsigned, but this is basically how all of these commands works.  Don't forget, you won't get these answers on the calculator because my examples are in base 10 for demonstration.

I know that this may be off-topic but exactly what should we expect in 0.4.0?