RE: clause 2.2

clause 2.2Karlsson Kent  wrote:
  Depends on what one defines to be a "bug"...  Some of the "bug-fixes"
already
  entertained are 'substantial' ...
Exactly what I said.  When the (imperfect, human) 1.0 spec says two things
in two places or is too ambiguous, we have to make a decision one way or the
other, with the potential to break something.  But we cannot do substantial
changes, unless required by a bug fix, while keeping the version number at
1.0.
  See also the suggestion I submitted today regarding clause 2.3 (cf E62),
  where I suggest not to potentially make (reasonable) current documents
invalid,
  but to have a "for compatibility" sentence instead.
I think you misread the definition of "for compatibility".  As opposed to
"for interoperability", "for compatibility" is not a non-binding
recommendation that is to be obeyed only if you want to retain
compatibility. "For compatibility" is an editorial comment that explains
*why* a design decision was made, but it doesn't change the fact that the
decision was made and is binding.  Compatibility with SGML (incl. the
WebSGML amendment) is a requirement in XML 1.0.  A later version (not a mere
re-edition) may decide otherwise.
--
François Yergeau

Received on Monday, 11 September 2000 11:42:37 UTC