Checking duplicate TPs is onerous

[dtikhonov]

Unlike HTTP/3 settings, receipt of duplicated transport parameters must be treated as an error:

                                                    An endpoint MUST
   treat receipt of duplicate transport parameters as a connection error
   of type TRANSPORT_PARAMETER_ERROR.

Why the inconsistency?

The checks for duplicated transport (or settings) parameters are expensive, because one has to maintain an arbitrarily-sized set of possibly large numbers.

I propose we switch the duplicate transport parameters enforcement to "MAY treat as error."

[ianswett]

I'm fine with MUST not send duplicates, MAY treat receipt as an error.

[MikeBishop]

Fine by me.

[janaiyengar]

Yup, seems sensible to me.