Dave CROCKER <dhc2(_at_)dcrocker(_dot_)net> writes:
The Other Dave C's highlighting the possibility of an "abstract" API
is also worth considering.
Yes. Today I would prefer to define abstract APIs in the IETF and let
implementers or other organizations map them to their language or
environment of choice.
The IETF is not the best place to go for implementation wisdom these
days, unfortunately. Specifying APIs could make more API designers come
to the IETF, which would be good, but I fear the IETF will just produce
poor APIs until a sufficient mass of clueful people have joined. I
don't want to see another GSS-API.
A compromise could be to specify one abstract APIs on the standards
track and allow multiple rendition of the abstract APIs in various
programming languges. Then if someone dislikes a particular concrete
API enough, they could propose a better approach.
/Simon
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf