Draft of binary format: moved to other repository

[SergeyKanzhelev]

PR was moved

This PR was moved here: w3c/trace-context-binary#2

[watson]

Great to see this getting started! 💯

[yurishkuro]

Why not define the binary format using protobuf IDL instead of inventing some custom encoding?

[SergeyKanzhelev]

@yurishkuro

Why not define the binary format using protobuf IDL instead of inventing some custom encoding?

I got a recommendation to not define it as protobuf as it's proprietary protocol/tech. That's said proposed format is basically how it will look in protobuf.

[yurishkuro]

I got a recommendation to not define it as protobuf as it's proprietary protocol/tech. That's said proposed format is basically how it will look in protobuf.

You just got another recommendation :-) Should we have a vote then?

There are many binary format specs out there. I think it would be best if we pick one rather than come up with a custom n+1 spec. There's nothing wrong with having the format defined by a widely used spec imo, it does not force the implementation in any way.

[SergeyKanzhelev]

You just got another recommendation :-) Should we have a vote then?

Before voting - do you agree that we should provide binary encoding rules that basically copies proto algorithm as not everybody can take dependency on proto?

Also I tried to do it with proto and it feels like an overkill. For instance, instead of byte length that is easy to read we will get varint with complex algorithm of serializing/deserializing.

What other binary format specs you can suggest?

[SergeyKanzhelev]

PR was moved

This PR was moved here: w3c/trace-context-binary#2