> I think it is fine to require from a caller that they are aware that a > byte (or byte array) must be passed to i2c functions. Given the (maybe) > two problems I pointed out above making it a bit harder to pass non-byte > data to these functions doesn't sound like a bad idea to me. > > Obviously your mileage varies, but I personally like having an explicit > type in the API. I guess we have to agree to not agree and let Wolfram > decide if he likes your change or not. I am on Uwe's side here. I like it being explicit and think the casts as they are now are the smaller problem.
Attachment:
signature.asc
Description: PGP signature