RE: proper use of abbr and acronym

Dave  J Woolley, at 16:49 -0000 on Fri, 17 Nov 2000, wrote:

> 	Most authors won't even know the elements exist; of the
> 	rest, most will only be interested in what font style
> 	and size they produce (regardless of whether the content 
> 	really is an abbrevation or acronym)!

Sure, most don't know the elements exist.  But I do.  And I care; maybe
I'm the last of a dying breed :)

> 	This seems to be a good case for automation; if you markup
> 	the first occurrence, an automated tools can then scan for
> 	the remaining ones.  One could then use a special style sheet
> 	to make them stand out to ease the removal of false positives.
> 	(I believe good index generation systems take a dictionary of
> 	indexing terms rather than requiring every occurence ot be
> 	marked.)

There is always a balance needed between automation and your editing
environment; in my opinion, ducplicate code, no matter where, should be
minimized, and handled by modularity/referentiability in the language (in
contrast to having the editor work).

> 	[DJW:]  Where is the market for this; without an apparent
> 	market, the big 2 aren't going to add this to their user
> 	agents.

Just because a need is not currently perceived, doesn't mean one won't
arrive in the future.  Clean, functional design always helps, and can
often open up new things to do with dfn's.  Anyways, anyone can contribute
to Mozilla, or Konqueror, etc, etc, so it's not only "two big companies"
designing browsers.  But we shouldn't really get into that :)

> 	In practice, I think you need the title for text to speech,
> 	as such tools are unlikely to look at previous uses and 
> 	there might be cases where the same character string is used
> 	in more than one way.

True, in practice, this seems to be the only purpose behind an abbr or
acronym element.  It's hard otherwise think of semantic meaning behind an
abbr or acronym.  Of course, this doesn't mean there couldn't be a
practical use in the future.

> 	Hyperlinking every occurrence is noisy; I don't think it
> 	is a good idea.

Ah, but this noisyness is due to the user-agent, not any implementation
requirement.

> 	Note that you have used "tag" where you mean "element".

I tend to use the term "tag" to mean the textual representation of an
element, or, in the verb sense, to "elementicize" some data.

-- 
Frank Tobin		http://www.uiuc.edu/~ftobin/

Received on Friday, 17 November 2000 14:26:41 UTC