Re: XML Blueberry

[Benjamin Franz <snowhare@nihongo.org>]

On Fri, 22 Jun 2001, Tim Bray wrote:

> At 01:26 PM 21/06/01 -0400, John Cowan wrote:
> > And
> >upgrading parsers to Blueberry is intentionally trivial: it's basically
> >about expanding a few tables.
> 
> Well, the upgrade is just a software upgrade, and it doesn't really
> make any difference whether you're changing one bit in a dispatch
> table or doing a major re-write.  It's not clear that upgrading
> the installed base of XML parsers is trival.  No... it is clear
> that upgrading the installed base of XML parsers is not trival.

I would contend that given the 25 year old trend of migration _away_ from
mainframes and the almost frightenly fast adoption of XML 1.0 worldwide,
that significantly more pain is already attached to breaking
interoperability for XML than in NOT allowing NEL as an accepted
whitespace char.

Also, it is clear that XML while readily _viewable_ in text editors is
actually something that should not be _generated_ by text editors but by
specialized tools written for the purpose. And _given_ that to do XML at
all correctly requires new software to be written, I would contend that
much less pain is attached to forcing IBM-mainframe legacy shops to
migrate to XML 1.0's view of the world when they want to play XML than
forcing the rest of the world to migrate to IBM's view of the world.

-- 
Benjamin Franz

  Programs must be written for people to read, and only 
  incidentally for machines to execute.
                             ---Abelson and Sussman