Andrew Hlynskyi

> I've tried the following, but it didn't even be parsed 😭 You're not alone, my opinion is that the problem is in an approach that was chosen far times...

@amaanq could you add grammar tests for cases that were issues before? To avoid regressions in the future.

I've added the `postponed` tag to the original PR and asked to don't continue to push such PRs to master.

> I think fuzzing scala + seeing the perf benefits as before would mean it's ok That's not the reason, I asked to postpone it because I want to check...

**TODO:** * Needs to be coordinated with #2438

There is a lot of interest in queries (de)serialization and a more robust solution will be implemented that covers more use cases with better deserialization performance.

@criminosis Thank you for the PR and the idea, but I added a proposal #2541 with a more generic design.

@ErinvanderVeen It looks like you forget to commit a file `bits/stdint-uintn.h`.

When I tried to thought about a good implementation for #1942 I thought about several things: 1. It would be great if it would be possible to improve queries compilation...

> > @ErinvanderVeen It looks like you forget to commit a file `bits/stdint-uintn.h`. > > it's a glibc header It seems it isn't portable and don't exists for Emscripten `emcc`...