Comments of HTML+ draft
Tim Berners-Lee <timbl@www3.cern.ch>
Date: Wed, 17 Nov 93 14:39:50 +0100
From: Tim Berners-Lee <timbl@www3.cern.ch>
Message-id: <9311171339.AA00566@www3.cern.ch>
To: dsr@hplb.hpl.hp.com
Subject: Comments of HTML+ draft
Cc: www-talk@nxoc01.cern.ch
Reply-To: timbl@nxoc01.cern.ch
Dave,
A list of points
0. Thanks for the very great effort.
1. Editorial...
1. The draft filename is unfortunate in that "plus" is
missed out,
and the name looks different from the html draft name
draft-ietf-iiir-html-00
draft-raggett-www-html-00
I don't know what we do about that. It is reasobal to miss
out iiir as they are not involved with that yet but it is
their area. They can both go under the new nominal WG.
1.5 The document needs an index. It is difficult, for example,
to find the discussion of LANG on p6 and I never found a
discussion of the INDEX attribute. (Comes from my
reading in a order random.)
1.6 May I suggest as a question of style -- and an aid to
convergence -- that you wrod the spec in a conventional
way as a specification -- with "should" and "shall"
to specify levels of imperitiveness? The "could"s
which are comments about how the spec should be changed
should be compleetly separate -- like in a different font.
That way one can say "I am happy with the spec as it stands
at <date> except...".
2. ISMAP attributes are in the FIG, IMG, etc. I really want to
make them obsolete as quickly as possible.
3. Where is HR? Not in Appendix IV
4. Why does INPUT have no ID? Does Name serve it's purpose - if
so, tehn it should be "ID", else it ID should exist
separately.
5. ISINDEX I would also like to see "obsolescenbt", with the
use of LINK instead.
6. LINK should have PRINT if A has PRINT. LINK should also have
ID.
7. I think you should be more precise about the entity names,
rather than "most browers support". Make two levels if you
have to, but please define them. Perhaps a minimum set and
also a very large set. For example, the upper part of
a left brace is not necessarily the way to go for math...
We want to be able to establish real poratability for
HTMLplus conforming documents -- not postscript-like
kindaportability.
We can add to this recommendations for how to
make extensions ina non-destructive way, as we always have.
8. PRE no longer has "WIDTH". That stops HTML+ being a
superset of HTML.
9. If HTML+ is to be a superset of HTML, it musg include the
<HTML> element.
10. Entity names are case sensitive. Looks like the example
óengus should be Óengus on p5.
11. Do we need ID on TITLE? Specify that the title is not
part of the document contents.
12. p32 -- what is a "conventional hypertext-based index"?
13. p22: Yes, I think a separate METHOD attribute would
be cleaner. POST should be the default.
14. 10.1, p25: next to last para. "In the near future..."
This spec should not be limited. Insist on floats
now. It will be much more difficult later!!!
10.2 p26: The definitions of the formats for sending
forms by HTTP, mail, etc., need to be more clear.
I forget exactly the dates of the Irish meeting, but hope it goes
well anyway.