Ka-Ping Yee <kryee@novice.uwaterloo.ca> wrote:
> A week ago, i wrote a message here lamenting the ridiculous proliferation
> of easily-abusable tags[1, 2] in the proposed HTML 3.0 specification[3].
>
> Aside from a discussion about whether <Q> actually has a distinct semantic
> meaning then <BLOCKQUOTE>, responses generally agreed that most of the
> other tags i mentioned were not necessary. Specifically, they are:
>
> CODE KBD AU ACRONYM ABBREV INS DEL S U BIG SMALL
>
> The above tags are unnecessary for reasons explained in [4] and summarized
> at the end of this message.
I agree with you on some of these, but not all.
INS and DEL provide structural and semantic information beyond simple
formatting, and I don't think they're redundant.
ACRONYM and ABBREV are needed for accurate conversion to Braille [1].
They could also be useful for things like spell-checkers and
indexing search engines.
> Summary: why we don't need....
> CODE, KBD: redundant. Use SAMP.
Agree.
The differences between CODE, KBD, and SAMP are so subtle
that CODE and KBD could (and should) be removed with no
ill effects.
> AU: too specific. apply attributes of all kinds to PERSON instead.
Agree.
The rest of the DTD does not provide enough structure to make it
really useful anyway: <AU> can appear anywhere, so there's no way to
tell *what* the <AU> element is designated as the author *of*.
(The general-purpose <PERSON> element is useful though.)
> ACRONYM, ABBREV: redundant. Use DFN.
Disagree. ACRONYM and ABBREV have special semantics not
covered by <DFN> or <EM>. I don't think the CLASS attribute
is appropriate for identifying these semantics, either, because
in order for general-purpose indexers, spell-checkers, et cetera,
to recognize them, those CLASS values would have to be
standardized, and the namespace for the CLASS attribute
should belong solely to document authors.
> INS, DEL: too specific. apply attributes to P instead.
Disagree. These are useful for identifying revisions
to any type of document.
> S, U, BIG, SMALL: purely presentational markup. doesn't belong in HTML.
Agree.
S, U, BIG, and SMALL (along with B, I, and TT) could be folded
into <EM> or a new general-purpose character highlighting
element (C? CLF? FONT? EMPH?), with additional attributes
or in conjunction with stylesheets.
> and i still wonder about:
> Q: too specific. are we going to mark up <PAREN>, <SENTENCE>, etc. also?
I wonder about this too.
[1] <URL:gopher://gopher.mic.ucla.edu:4334/11/ICADD>
<URL:gopher://gopher.mic.ucla.edu:4334/0./ICADD/
info/documents/ICADD.Mechanism.Rubinsky.txt>
--Joe English
joe@art.com