Re: Font-style vs. phrase elements (fwd)

Scott E. Preece (preece@predator.urbana.mcd.mot.com)
Wed, 29 May 1996 09:15:52 -0500


Date: Wed, 29 May 1996 09:15:52 -0500
From: preece@predator.urbana.mcd.mot.com (Scott E. Preece)
Message-Id: <199605291415.JAA20687@predator.urbana.mcd.mot.com>
To: megazone@livingston.com
Cc: abigail@tungsten.gn.iaf.nl, www-html@w3.org
In-Reply-To: MegaZone's message of Wed, 29 May 1996 04:26:29 -0700 (PDT)
Subject: Re: Font-style vs. phrase elements (fwd)

 From: MegaZone <megazone@livingston.com>
 
| >++ I don't see the advantage to this, what am I missing?
| >More structure in the document.
| 
| Is there anything out there now that takes advantage of this?  Or even,
| how does this help?  Just because I've wrapped things in DIV containers, 
| what does that buy me.  "More Structure" is a meaningless statement IMHO,
| what, in practical terms, does it accomplish.
---

Well, how could there be anything that takes advantage of DIV when DIV
is not yet standardized?  I would expect that once DIV is common, at
least some indexing engines will use DIV to subdivide documents into
sections, so that searches can point to a specific section rather than
to the document as a whole, and so that Boolean and proximity searches
can be applied within a section, avoiding catching one word at the
beginning of the page and another at the end.  This kind of stuff has
been part of full-text text retrieval systems from the bginning.

*Plus*, if you follow the suggestion that started this exchange, you
bind a particular heading to a particular extent of text, which makes
meaningful outline views possible.

*Plus*, if you standardize some specific CLASS values for DIVs, you can
have things like abstracts, sidebars, bibliographies, etc., which can be
used by clever browsers and search engines to present material more
effectively.
---
| >That is strange. Why would you have more than 6 deep nesting of <div>s,
| >but no use for <h6>s? (I assume you don't use them now). I would say,
| 
| Because, since H5 and H6 render terribly, I will just continue to use H4
| from then on.  Yeah I'm sure it violates some guideline, but in this case
| readbaility is a practical concern.  H6 renders smaller than regular text
| on most browsers!
---

Giving up all hope of reconstructing the structure of the document (the
nesting of sections) in favor of ad hoc, browser-specifc presentation
concerns.

---
| I also don't believe in the 'headers must be in numerical order' bit.  Most
| documentation I read doesn't do this, the size and/or emphasis on the header
| is related to the contents or the desire to attract a readers attention, not
| on an arbitrary rule of nesting.
---

One of my primary complaints about HTML is that the DTD doesn't enforce
header level sequencing.  It's one of the reasons I often wonder whether
HTML is worth caring about - there's *so little* structure there that
it's really, really hard to get excited about whether proposed additions
are structural or presentation-oriented.

If you lie to the UA about the structure of your document, sooner or
later you're going to hit a UA that's smart enough to be confused and
your document is going to look really weird or be unindexable or appear
to be broken.

scott

--
scott preece
motorola/mcg urbana design center	1101 e. university, urbana, il   61801
phone:	217-384-8589			  fax:	217-384-8550
internet mail:	preece@urbana.mcd.mot.com