Simon Cruanes

Oh that's cool, I looked the other day and it wasn't working but now it is! Very neat :)

We don't need a ppx I think, we already have printers! Dune expect tests are fairly good already. But I think it's a good idea!

You can have multiple outputs per file, they'll just be concatenated. You can separate them with a printf if you want but there's no inherent need. And yeah I'd rather...

Short reply from my phone. First, improving the testing workflow would be great. Right now it involves fetching a submodule and having protoc in the path, and it's a bit...

My 2 cents on the wire format: - a binary format like CBOR would go a long way in improving the efficiency of encoding and decoding, but it does make...

On paper we could have an option to represent the value as a `bytes * int * int` (just like there's an option to use, say, int instead of int32)....

I'm not sure it's really a bug, the FIFO scheduler isn't meant to be used for super fine grained parallelism. It's there mostly as a simple scheduler for big tasks....

I'm quite interested in the proposed feature. I currently use `__FILE__` and `__LINE__` for tracing-related purposes. I think @dbuenzli is right to suggest the reuse of the existing types. In...

Interesting. I've had to carry around some context recently as well between _async_ spans, so it's not just Eio. I think there needs to be clear ways to: - grab...

Yes it's annoying. I'm not sure how to improve on it (perhaps write a wrapper to `Eio.spawn` that captures the current trace context, spawns a fiber, and sets the current...