Damian Poddebniak

> Is it just the warning, or does the formatting not work at all when configured through rustfmt.toml? It doesn't work 😞 > I think running a custom command before...

This is a bit difficult to resolve cleanly. I experimented a bit on `duesee/issue_1185` and we could make it so that we expose only one `VerifiableGroupInfo::process` method. This method would...

Okay, cool. Before I make the PR: Currently I just used [this](https://github.com/openmls/openmls/commit/11f6aed6cc66c48dd681c6229637ee065d52fc10#diff-079771f025c001b816463e46e76a9b06413086780a847cb0d772f27eac3a18a1R34-R59) to experiment around. Do you think we should introduce a variant `WelcomeError::GroupInfo` instead and use ... ```rust pub...

Alright, thanks! I'll go with the `#[from] GroupInfoProcessError` variant then. Luckily, we are still "tight" here so we can delay this question a bit. Being able to move a bit...

Unassigning for now until I have validation straigt.

Maybe all we need is a `#[validate(...)]` attribute in tls-codec so that we can write `#[validate(leaves_are_ordered)]` with `fn leaves_are_ordered(...) -> Result` in OpenMLS? Food for thought: [Parse, don’t validate](https://lexi-lambda.github.io/blog/2019/11/05/parse-don-t-validate/) Spoiler:...

> I think we already encode guarantees in the type system in quite a few places. The validation chain is an example for that. Oh, sure. And that's pretty great...

Why I think this could be helpful: I tried [that](https://github.com/openmls/openmls/commit/dcacd273e31e40181aa710ade50903d63b98604e#diff-9542fb0d7e6c75207bf49d09f575e2620955a9a8560fb76c21f80696c5236848R133) while working on #1140. This way, the type-checker would guide us through all places where a certain type of extension...

> I've notified the team and we'll schedule work to add the -addext "keyUsage=critical,digitalSignature,keyCertSign" as suggested. I'm not exactly sure what the certificate is used for but if it's mainly...

You were faster, hehe :-) The order -- when resolved to a sorted set -- should not be important. AFAIR, imap-types already reorders accordingly when, e.g., iterating over a sequence....