Re: Binary Messages

[Julian Reschke <[e-mail]>]

Am 11.08.2021 um 02:19 schrieb Martin Thomson:
> Hey everyone,
>
> I just submitted an updated proposal for a simple binary message format for HTTP.
>
> https://datatracker.ietf.org/doc/html/draft-thomson-http-binary-message-01
> Live editors copy: https://unicorn-wg.github.io/oblivious-http/draft-thomson-http-binary-message.html
>
> We're forming a working group for doing oblivious HTTP and this is a dependency of that.  That group might be good for the crypto pieces, but this is pure HTTP magic.  I'd like to talk about adopting this work here.

Thanks for working on this, Martin.

So it seems that the design is something like "the simplest possible
binary format which will support HTTP's base semantics" (and yes, I
agree with that approach in principle).

Things that this will leave behind are:

- reason phrases (currently optional, only supported in HTTP/1.*)
- "middlers" (in case we ever get to them)
- pseudo header fields other than those defined in H2

This might be completely acceptable for the original use case, but maybe
should be revisited if we're looking for a replacement of application/http.

Best regards, Julian