bendk

> > * then fairybridge should define blanket FFI trait impls for its types > > Thanks, but I'm still mildly confused - in the description, you mention "Make UDL...

Just pushed the implementation of this. I split it up into different commits for readability, but I'll probably squash them together when we merge. All tests are passing and I'm...

> We probably want my PR into Bens fork: [bendk#3](https://github.com/bendk/uniffi-rs/pull/3) - without it it is not possible to have a single error type defined in crate `X` but used in...

I made #2150 to try to land the custom type changes before the remote type changes. The remote type changes are blocked by #2130. I think we're pretty close to...

> @bendk @mhammond did you see my comment [#2087 (comment)](https://github.com/mozilla/uniffi-rs/pull/2087#issuecomment-2128711332) above? any plans on fixing support for workspace setup with UniFFI? I think my plan above is pretty straightforward to...

Closing this in favor of #2303.

@mhammond I just added an upgrading guide for users. I'm not sure how to hook it in to the new doc system though. Does this mean updating the `gh-pages` branch...

> > I'm not sure how to hook it in to the new doc system though. Does this mean updating the gh-pages branch after we merge this? Maybe you could...

> Instead, just the term "custom" appears, and without much experience with UniFFI it might not be obvious what is meant by the term (eg, it might be easy to...

Using two terms for the two different types makes a lot of sense. What if we leaned into that more and used a syntax like this? ``` uniffi::custom_type!( custom: HeaderMap,...