Re: [quicwg/base-drafts] consider making the varint encoding unique (#2299)

[Martin Thomson <[e-mail]>]

I'm pretty sure that this was a case where @ekr said that a canonical encoding would be silly.  That's certainly my view.

I believe that this would be a fairly significant hit both in terms of code complexity and performance.  If you think that you can make the case, I think we might start with a PR [here](https://github.com/martinthomson/quic-int) to show that a) the code is simple enough, and b) the performance is tolerable.  Then there is the point about churn.

The requirement on minimal encodings for frame types exists so that a frame decoder can be more efficient, as @mikkelfj says.  I'm reluctant to add that requirement in too many places though (even in the h3 cases we are currently considering).

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/2299#issuecomment-451788047