[
Lists Home |
Date Index |
Thread Index
]
>XML core would be a natural lead group for such an activity, don't you
>think?
(I'm speaking here in a personal capacity, not as a representative of the
Core WG.)
I think the natural group to lead - or at least start - such an
activity would be the people who need the functionality.
There are two problems:
(1) People keep coming to us saying "something has got to be done
about character entities". When we have asked why, we have received
various different answers, some of which are based on factual
misunderstandings, such as that entities can't be used with Schemas.
Others, while not wrong, do not seem (to me) to justify the cost of a
new mechanism, such as that the XML spec does not require processors
to read an external subset. The document just published aims to make
the WG's position clear, so that if there really is a need for a new
entity mechanism then those who need it can make a coherent case for
it without rehashing the same issues over and over again.
(2) There are (as far as I know) no readily-available experimental
replacement mechanisms for those who believe they need them to try
out. I don't think the Core WG should invent and standardise a new
mechanism. If there is to be a new mechanism, it should be proposed
and implemented by others and then the Core WG can, if necessary,
produce a spec for it. Otherwise we will have the usual outrage from
those who think that the W3C is a great conspiracy to make everything
more complicated.
Perhaps the constituencies who believe they need a new mechanism
(XHTML? MathML?) should form some kind of ad hoc working group to
produce a requirements document.
-- Richard
|
