SCS2 Migration Plan

[Markus Demleitner]

Hi DAL,

On Tue, Aug 04, 2026 at 11:02:03AM -0700, Stelios Voutsinas wrote:
> One more question to iron out the details I have is whether it is valid for
> a service to co-locate both dialects under the same accessURL?
> If this is allowed, there might be some confusion due to the different
> error conventions, but I don't know how common a use case that would be or
> if the spec should take a position one way or another.

I have not thought this through.  Ages ago, it was thought that by
passing in a VERSION parameter, such a co-location would be simple
and safe.  We have shelved that idea for many reasons, and I believe
plannig for it is asking for all kinds of trouble, in particular
raising the expectation that multiple versions live next to each
other beyond a (notionally brief) transition time.

My stance would be: If you think you can reliably auto-detect the
protocol version, I'd not come after you.  But I'd neither design
protocols such that that is easy nor would I discuss such a design in
the spec.

URLs are cheap.  Guessing is hard.  I'd use two URIs.

Thanks,

          Markus