Vicki Pfau

Maybe that should be added to the examples repository after this is merged.

You put the const on the wrong side of the pointer. This just marks the variable itself as const, not the memory it points to, which is useless inside the...

I think the ones that output void* break const correctness, but still "work" because they're explicitly cast to void* (which you shouldn't actually need to do in C, and would...

I had assumed they were void*, not vu32, so my reply was incorrect. @mtheall corrected me. Sorry about the confusion.

This could easily be rectified by just removing kvsprintf and using vsprintf instead, since that's in newlib

You might be able to replace it with nano-vfprintf from newlib then: https://github.com/devkitPro/newlib/blob/e928275566ab8168c078eadee2043a002bd9352d/newlib/libc/stdio/nano-vfprintf.c

Regarding some of these features, I do have some use for a few of these things, like bitwise ops (there are tricks you can do with them that are useful...

Doesn't work; l, r, and b are only defined within the scope of each if/else block, not beyond them. That's why the mention of imperative was brought up at all.

This was definitely _not_ a regression. This is really old code. I guess I just misremembered.

I have a fix in preparation for this, but I'm not going to backport it into 0.10.3. @fleroviux, do you know if the behavior of HALTCNT depends on the old...