awful thing to do in production but when you're just experimenting and trying to see if an idea works without committing to refcells or whatever? why not?
basically a complete borrow checker bypass, with all the awful consequences you'd expect from it.
and of course it could be done manually too like how console currently works
what if uxn devices could be interacted with by repeatedly reading or writing to a memory location? It could either be done manually or even with dedicated ops, something like:
IOR (IORead): Pop source byte address, pop number of bytes to read, pop destination buffer address, then repeatedly read the source byte address to fill the buffer
IOS (IOStore): Pop source buffer address, pop buffer length, pop destination byte address, then write every byte in the buffer to the byte address repeatedly
not sure how i can implement sprites without direct memory access since right now devices can't read anything outside of their own memory and i'd like to not enable that if it can be avoided ;_;
@alderwick a thought after seeing the stuff about vectors on the uxnemu page:
if the first byte of each device's memory is used as a vector, interrupts can be handled by the uxn machine itself: every cycle, check if the first byte of each device is not zero, if so, execute the code at the vector, then set it to zero
@neauoire @xiroux very weird how, in 13 years of using and advocating for FLOSS, I never thought of wording it that way: “the right to repair software”. I think it is a way more effective argument and realistic/achievable action than “the right to create a modified version or to add a new feature”.
Love it, thanks!
documenting what i learn along the way
Server run by the main developers of the project It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!