Re: [Id-event] Making SETs distinct as JWTs

[Benjamin Kaduk <[e-mail]>]

On 03/03/2017 08:03 AM, Brian Campbell wrote:
> How much should theoretical (but likely) implementation defects play
> into protocol design? I don't know the answer. It would be nice to
> rely on the requirements of the RFC 7515-7519 group and, as Mike
> suggests, push to get deficient implementations fixed. But practically
> speaking that might amount to willful ignorance.

When possible, protocol designs can allow for well-behaved ecosystem
participants to enforce the behavior in question, as in TLS's GREASE to
detect intolerance to adding new codepoints in various fields.
And of course, making the required behavior critical for normal
functionality is a good way to get it implemented everywhere, but that's
not really an answer so much as an observation.

-Ben