🚀 go-pugleaf

Joachim Schrod writes:

> Subject: The definition of `DTD' (was: Re: WWW: a HyTime application?)
>
> YEAH!
>
> Now we now, what a DTD ****really**** is, don't we?
>     ``parameter-entity-replacement closure of the
>     	document type declaration's
> 	document type declaration
>     subset.''
> (Please note the nice doubling of the term DTD^H^H^Hdocument type
> declaration. Isn't it almost literate?) This explanation is easy
> readable and immediately understandable. For all those folks out
> there who have always said `SGML is way to complex to grasp, we just
> want to handle our documents in a structured way' -- bah, that is not
> true, we just should tell them!
>
> Really, I'll save this article (Message-ID: <9301192217.AA00521@dylan.local>)
> and will show it everybody who asks me why SGML is not more widely
> recognized and more widely used than today.
>     Hmm, perhaps he or she will know the reason then...
>
> --
> Joachim

Sorry, Joachim.  I didn't invent the vocabulary, I just try to use it as
correctly as I can.  In fact, there is some shorter/easier-to-understand/more-
precise vocabulary (all three at the same time!) being developed and used by
some people (including me), but for the benefit of those out there that haven't
been exposed to the new jargon but do know the old, that's the way I wrote it.

	o  I didn't try to define or even to explain "what a DTD
	   ****really**** is".  Once the commment wording is corrected,
	   "DTD" is not involved--only "document type declaration".
	
I've drafted an article on "what a DTD really is"; if it's rejected for
publication, I'll probably post it on the newsgroup.

	o  There are things called "document type declaration subsets";
	   every document type declaration has exactly one.  "Document
	   type declaration subset" has another meaning which precludes
	   its unqualified use in this particular context, since I was
	   trying to avoid ambiguity.  (I'll avoid boring everyone with
	   a discourse on the two meanings of the phrase, unless there
	   is a general clamor for it.  :-) )

	o  Goldfarb doesn't seem to like to distinguish between a character
	   string containing entity references and the character string
	   that results when those references are replaced by the replacement
	   text of the referenced entities.  I do; I find that many people
	   who don't make the distinction consciously get tripped up by
	   faulty intuition.  "Parameter-entity-reference-closure" means
	   the string that results when the parameter entity references are
	   replaced by the content of the referenced entities, any parameter
	   entity references in the resulting string are similarly replaced,
	   and such replacement recursively continued until there are no
	   parameter entity references.  "Closure" used in this way is common
	   mathematical terminology and is usually puzzled out correctly by
	   non-mathematicians; I know of no other concise term.


Dave Peterson

xyvi!davep@uunet.uu.net