Andrew Twyman

There's no support for that in Djinni currently. The best I can suggest with the existing generator would be to create a const map which can be used to lookup...

Djinni does require C++11, and always uses typed enums. I don't think any change to that is required for the request here. I think the core of the request here...

Huh, I'm amazed that nobody's ever tested this or run into this before. As a non-ObjC-native I'm amazed that an array which can't hold nil is considered reasonable. Same is...

I've got some experimental code which will signal those errors (in resolver's buildMExpr() where we also disallow optional. A drawback is that the resolver code isn't specific to a particular...

Since they're mixed, this will need to wait to be resolved/merged until after #247. This seems like a good idea, but since I don't use CMake myself, I can't really...

I'd encourage you to go ahead and add some test coverage, lest it break constantly. Since it's a new file with no impact on existing users, though, I'm okay with...

Decoupling PRs is a good idea in my book.

@4brunu if you want to send a separate PR for CMake support without the reorg, I'd be open to that, since it doesn't look like #281 is likely to be...

Our use of Djinni at Dropbox is primarily for Android and iOS, though we're certainly using it on Mac OS X as well (using the same Xcode toolchain as we...

In the current implementation, the only reason for records to be immutable is to avoid confusion. If they were mutable, it would be easy for developers to assume they could...