Re: what about XPath? was: Can everyone be happy?

>- fix the namespace prefix matching -- of course this is the hardest part
>because it will either break the existing recommendation or the existing
>implementations.

GIVEN that the XPath folks have said they did not intend to diverge from
the intent of the Namespace spec, and that the Namespace folks have said
they did not intend to support relative syntax (never mind absolutization)

AND GIVEN that nobody has reported an XPath implementation that has taken
the absolutization wording seriously,

MOVED: That we declare the divergence an erratum, say "Oops, but that's the
risk you accept in using Version 1.0 of anything", adopt the Forbid
behavior, and stop spinning our wheels.


We've established that no solution will make everyone happy. Forbid seems
to generate the least total unhappiness, multiplying individuals by
strength of reaction.

______________________________________
Joe Kesselman  / IBM Research

Received on Thursday, 22 June 2000 10:45:44 UTC