Niklas Saari
Niklas Saari
> Do you mean the derive implementation? That wouldn't help because you can't rely on the caller using the derive implementation. I was thinking that would it be possible to...
I see. I will use the similar method than in `PER`.
Okay, `SET` seems to work now, also with `personnel` test! I wasn't that big change after all, while the code looks a bit messier now.
It should be rebased. Those changes should be related to those two other PRs, since codec depends on them.
> What about basic and canonical decoding? Basic decoding is default, but should be still improved. If we want strict canonical decoding, that should be easy to add. There isn't...
> Perhaps, I know people want the opposite, because there are protocols that are stream based, so you may have a buffer which contains a message + seven bytes of...
Ehh.. I just noticed that I used `BigUint` for length determinant on decoding size but `usize` on encoding... 🤦 I guess `usize` is the best way,
I can try to make a suggestion on separe PR which combines all current errors from the main for encode error/decode error modules.
Hey. I have almost done the separate PR. I had to do some other things over the past weeks. I will get back to it next week.
I will have also 3 standard implementations which tests this codec further (needs 4 separate crates to implement them, as one workspace for now, due to cyclic dependencies), and I...