devkitPro

I looked at the [url=http://libnds.devkitpro.org/a00100.html]libnds documentation[/url] and the [url=http://devkitpro.cvs.sourceforge.net/viewvc/devkitpro/libnds/source/arm9/sprite.c?revision=1.13&view=markup&sortdir=down]code from sprite.c[/url], but I can't seem to find oamAffineTransforma...

IIRC, libnds has an opengl-isch interfase, and the examples came from the nehe tutorials. so I would suggest nehe's opengl tutorials: http://nehe.gamedev.net/

after you know the beginnings of 3d programming, you can try it on the ds.

seeing as the nds has no usb port, my guess would be no.

PAlib has a lot of issues with staying up to date with libnds, resulting in PAlib not working correctly when libnds updates. you can either move to a previous version of devkitARM* ( NOT recommanded , everything but the last version contains bugs and isn't supported) and hope somebody updates PAlib ...

1. the ds (and gba?) hardware uses 8*8 tiles, and so grit defaults to that. however, you can tell grit that the tiles are in another format, like 16*16, and grit will convert 1 16*16 tile into 4 8*8 tiles so that you can easily use it with the ds. the options are -Mh and -Mw. look at http://www.cora...

the sub engine doesn't have a framebuffer mode.

see here for all the posible modes:
http://www.dev-scene.com/NDS/Tutorials_Day_2#2D

thats the part that he modified.

the 'real' ndstool line is in the ds_rules file, which is included in the beginning. if you look there you can see what to define for getting icon and author and stuff.

or by setting the sprites visibility to false, which is what oamClear() does.

also, oamInit() calls oamAllocReset() in the end, so I think it would be ok.

1. the dx parameter of the oamSetMosaic() function doesn't work (the dy parameter does work). 2. there is a function to set the mosaic for the main sprites, but there isn't a function to set the mosaic for the sub screen. I've created a function that you can copy, (modify) and paste to sprite.h. do ...

well, one of the first things oamInit() does is a call to dmaFillWords(), so my guess would be yes. I suggest you try it out and see if anything weard happens.

but why aren't you using oamClear() to clear/reset the oam?