[Dave Raggett: Re: HTML 2.0 Call for Review]

"Daniel W. Connolly" <connolly@hal.com>
Message-id: <9406101612.AA07777@ulua.hal.com>
To: html-ig@oclc.org
Subject: [Dave Raggett: Re: HTML 2.0 Call for Review]
Date: Fri, 10 Jun 1994 11:12:00 -0500
From: "Daniel W. Connolly" <connolly@hal.com>
Content-Length: 3499

I'm sending all the review comments I have so far to this list
to be sure they become part of the archive.

------- Forwarded Message

From: Dave Raggett <dsr@hplb.hpl.hp.com>
Message-Id: <9406061225.AA11520@dragget.hpl.hp.com>
Subject: Re: HTML 2.0 Call for Review
To: connolly@hal.com
Date: Mon, 6 Jun 94 13:25:11 BST
Cc: rdaniel@acl.lanl.gov, marca@netcom13.netcom.com, pflynn@curia.ucc.ie,
        weibel@oclc.org, terry@ora.com, mcrae@ora.com, timbl@www0.cern.ch,
        esink@spyglass.com, jay@eit.com, jackson@ncsa.uiuc.edu, yuri@sq.com,
        murray@sco.com, gary.adams@east.sun.com, wmperry@indiana.edu
Mailer: Elm [revision: 70.85]


Please keep me informed as to developments. I am busy right now
preparing for the Internet Society Conference next week, but will
be keen to review the spec after that.

Here is a list of preliminary comments:

    o   Whats wrong with B, I etc inside PRE?

Your comment suggests that these are illegal inside PRE elements, why?

    o   <!ENTITY % pre "PRE | XMP | LISTING">

Doesn't this make it hard to test for obsolete elements with SGMLS?
I would prefer to see them defined in a marked section so that
its easy to switch obsolete elements in/out of the DTD for testing.

    o   <!ELEMENT LI    - O (%htext|%block)+> etc.

The content model allows for text and <P> to be intermingled
as peers. This could be avoided by the following:

        <!ELEMENT LI    - O ((%htext)+|(%block)+)>

The same problem occurs for DD, BLOCKQUOTE and ADDRESS.
I avoided this in HTML+ by requiring a <P> element before any
text, which also makes it easy for browsers to infer missing <P>s

    o   HTEXTAREA - I would prefer to use TEXTAREA
        and make use of content-type mechanism

Why do we need a new element to input hypertext into forms?
It would be much more general to support a content-type mechanism
e.g. via a new attribute on TEXTAREA. Then users would be able
to paste images, hypertext or plain text into form fields.

    o   MENU and DIR will be obsoleted in HTML 3.0
        in favor of a more flexible use of UL

You may want to mention this.

    o   STRIKE used in place of HTML+ <S> why?

The HTML+ spec has for a long time provided <S>..</S> as a physical
style for strikethru. Why are you using a new tag name? For logical
use, I provided ADDED and REMOVED. One needs the ADDED tag to allow
browsers to diffentiate such text, e.g. my browser changes text color.

    o   what if people use dynamic entity declarations?

You should explicitly state that browsers are NOT required
to support such declarations. We may change this for HTML 3 or 4.

    o   no mention of processing instructions

You should require browsers to ignore comments and processing instructions.

    o   COMPACT etc, with NAME rather than:
            compact (compact) #IMPLIED 
        as in SGML Handbook, page 534 for GL element

I find this 2nd mechanism cleaner than the method you employ.

    o   ISINDEX missing HREF attribute

Doesn't Mosaic support this already? It allows documents to reference
a different URL for processing index searches, and will be in HTML 3.0.
- --
Best wishes,

Dave Raggett

- -----------------------------------------------------------------------------
Hewlett Packard Laboratories              email: dsr@hplb.hpl.hp.com
Filton Road                               tel:   +44 272 228046
Stoke Gifford                             fax:   +44 272 228003
Bristol BS12 6QZ
United Kingdom

------- End of Forwarded Message