Robin Freyler

Currently this feature is blocked by not being able to have `&self` or `&mut self` methods of a struct as `const fn`. This might be available sooner or later though....

Technically this is possible. However, the current implementation of `with_x(self, new_value: T) -> Self` bitfield calls the `&mut self` taking methods under the hood. While this is not technically needed,...

I'd accept PRs to add this as long as they are not exploding the complexity of the macros. Note thought that from what I remember, the implementation uses traits extensively...

> HI @Robbepop, > > Thank you for taking the time to check this. > > I've added your suggestion in [910b17d](https://github.com/Robbepop/modular-bitfield/commit/910b17df9adfe7dbfd67e0bc0c37ae1cc2ac06e7). Can you also confirm that this still fixes...

Sounds like a good idea. Not yet implemented though.

I think a proper syntax for this could be something like: ```rust #[bitfield(skip(from_bytes, getters))] pub struct MyBitfield {} ```

Generally the modular bitfield structs should not impose overhead besides longer compilation times due to usage of proc. macros. This library was designed to not introduce overhead anywhere. Bitfield structs...

That's a nice idea! However, from the current tech stack it is unlikely that conventional references will work here. However, it might be possible to create reference-like types that represent...

Question: Do we actually need those?

Thank you for clearing my confusion about rsmt2 ... So in order to implement my solver for rsmt2 I need an actual implementation of an SMT parser first. But at...