Re: interoperability in the face of extensibility

"C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com> writes:
> Or at least I never suffered from the kinds of interoperability
> problems some people reported.

I think it helps to bear in mind that for a large subset of the
community of web users, the browser is an application development
platform, not a tool for reading documents published by other human
beings. (I’m not asserting that the distinction is a crisply drawn
border where you could say catagorically that a URI was definitely on
one side or the other; I render DocBook documents with some embedded
JavaScript to support annotations which is more the latter but also a
little of the former.)

If the browser is an application development platform and if ad revenue
is in getting the sizzle in the application, then one is motivated to
chase performance and interactive functionality to the last decimal
place. I am not surprised that subtle differences in the way browsers
support extension elements or CSS renderers support --x-property versus
--y-property bite in this context.

It also seems like a very long way away from the kind of problems
pragmas could introduce in ixml grammars.

                                        Be seeing you,
                                          norm

--
Norm Tovey-Walsh
Saxonica