Added RefCell blanket implementation for I2C to allow sharing, see #35 #55
I know I originally suggested this, and I know there are at least two other threads discussing this
I'm concerned about the scalability and the "coupling" of this approach.
That's the scalability issue; the "coupling" problem is that
I'd like us to explore, and exhaust, all the other options before committing to one approach -- I
(*) I'll post some other idea that I have in one of those threads if it has not been brought up
@Rahix Not yet but I will. ;)
The problem I see is that the number of hoops one has to jump through to use Rust on MCUs is accumulating, so every time we say "Well, we can't do that because it's considered unsafe and in Rust we don't do unsafe." to something that is quite common everywhere else that is driving people away from Rust instead of attracting them.
Without commenting on the specifics of this case, I believe that to some degree this is unavoidable. There's something we want to get out of using Rust, and often we need to pay for that in complexity.
That doesn't mean we shouldn't do our best to minimize that additional complexity though.