Matthew Hammer

Yes, there are backend-only tests that exercise the logic there: - They run in CI, from here https://github.com/dfinity/cancan/blob/031f31c0f45af72e42416043e1a2415642844d4e/.github/workflows/backend.yml#L36 - They use [`ic-repl`](https://github.com/chenyan2002/ic-repl), which I recommend trying if you have any...

FWIW, I have a usecase that creates "large" trees (each under 500k bytes) and I notice a 10x difference between 0.6 and 0.7 (where 0.6 is 10x faster than 0.7)....

Thanks. Will fix the failing tests (tomorrow) before merging, obviously. I forgot to run `cargo test` locally first.

> Let the application handle that? That would be nice. Is that possible?

> the callee that somebody wants to forward to presumably is not expecting a dynamic value. Let's discuss that, perhaps? To me, it seems reasonable to treat these `dyn` values...

While gradual-typing-like features seem inevitable in the limit, I wonder if something much simpler and more explicit could still work in the interim? My thinking was that the MVP would...

> Typically, you need individual values, e.g. the generic value of a key/value store. Some explicit primitives, e.g., in a Motoko library, would introduce and elim the argument sequences (much...

@chenyan-dfinity Should we have multiple crates in this repo, like other Rust language projects? Or separate repos for each crate? (The downside of separate is that we loose visibility outside...

> Should we have multiple crates in this repo Nervermind that question. I see that we do now. New question: May I do a PR to add `candid_relate` here as...

MVP in #58 with some tracking issues for the missing pieces: - [ ] #68 - [ ] #67