Stephen Haberman

Hi @HofmannZ ! Thanks for filing the issue! That is an interesting but unfortunate result... The biggest change in ts-proto v2 is that we'd migrated to a different binary serialization...

Selfishly that would be really great & appreciated @timostamm ! :-D Fwiw @HofmannZ since this is important to you, would also be great ofc if you wanted to analyze/dive deeper...

Hey @HofmannZ ! Thanks for the questions/context on the flamegraphs... Honestly when I've done flamegraphs in the past, the issue is either pretty obvious ... or it's not, so I'm...

@HofmannZ realistically the best way for this to move forward is for you to dig in, find the hot spot(s), and open a PR, either to ts-proto or the underlying...

Afaict it's not documented, but there is a test asserting the behavior is desired: https://github.com/porsager/postgres/blob/master/tests/index.js#L643 Thankfully I'm not using the `transform` feature, b/c yeah this is very surprising.

Hi @yallxe ; that does seem odd. Here is the line of code that does this: https://github.com/stephenh/ts-proto/blob/f05de60e4c38c0522669a2ed4202f4cf1267296b/src/generate-nestjs.ts#L57 Tbh I really don't remember why it's there -- my guess is _maybe_...

@AndreyKovalenok the proto over gRPC shouldn't be using `json_name` because the proto binary format encodes the data using the numeric field ids, like `1=bob`, `2=smith` instead of `first_name=bob`, `last_name=smith`. So...

@aceHubert ah yeah 🤔 that is unfortunate; if you could submit a PR that fixes this, that would be great! Thanks!

I tried bumping to 2.x and am getting errors that it is ESM-only, although afaict that's not intentional on their side: https://github.com/ianstormtaylor/superstruct/issues/1287 A recent development in the runtime types space...

I'm running Regolith/sway on Ubuntu 24.04 and looks like this has been resolved? ``` $ sway --version sway version 1.10.1 ```