Re: Are namespaces actually crypto-entities or crypto-links?

[ht@markup.co.uk (Henry S. Thompson)]

Rick Jelliffe <rjelliffe@allette.com.au> writes:

> ...

> And, more in that direction, someone brought up the idea of using PIs
> rather than overloading attributes (which would have been my preference
> back in the day.)

A minor historical note here (Michael, are you listening?  Please
correct me if so!): I _think_ that that was indeed the near-final
original design, and the reason TBL rejected it!  Basically, again my
memory may be failing me here, because PIs are not scoped.  James (I
think) then came up with the xmlns: solution.

So everything is all intertwingled: if you allow external entity
references (and why not, they're very useful) or indeed external DTDs,
then you need a scoped namespace mechanism to insure against
accidental prefix collision.  Scoping the effect of PIs to entity
boundaries might or might not be a good idea, but would certainly add
a layer of complexity not only to implementation but also to
understanding.

ht
-- 
		    Henry S. Thompson, Markup Systems Ltd.
               Cavers Garden Farm, Denholm; by Hawick; TD9 8LN
                            +44 (0) 7866 471 388
	       Fax: (44) 131 651-1426, e-mail: ht@markup.co.uk
			URL: http://www.markup.co.uk/
[mail really from me _always_ has this .sig -- mail without it is forged spam]