By Murray Maloney, murray@sco.COM
edited by Dan Connolly, $Id: htmlplus-maloney.html,v 1.1 2006/10/11 05:45:45 connolly Exp $

HTML 2.0

The final discussion of this workshop was led by Dan Connoly of HaL. Dan is trying to codify a DTD for HTML 2.0 so that browser developers and information publishers will have some guidelines. As it stands today, use of HTML markup on the WWW is somewhat inconsistent and there is no formal description of the language to serve either group.

Dan drew a matrix on the board for discussion. The vertical axis listed several issues or features which need to be resolved or agreed upon. The horizontal axis listed current and future versions of HTML:

the June 1993 IETF draft
the unspecified state of affairs as of November 1993
the current draft under review
minor revisions of 2.0
the widespread deployment of HTML+ features such as tables and figures
the widespread deployment of math markup

Unfortunately, due to time constraints the group never completed its review of Dan's matrix, nor a determination of the status of each issue/feature and the time-frame in which it should be considered part of standard HTML. However, through the magic of electronic communications, Dan has provided us with a summary of the views and consensus that he believes was reached in that meeting and at other venues at the conference.

 Feature          1.3       1.x        2.0       2.x        3.0       3.x

 Images            S1        S1         S1        S1         S1       S1
 Forms             -         I,P        S2        S2         S2       S2
 Tables            -         P          P         P,I        S3       S3
 Figures           -         -          -         I          S3       S3
 Mathematics       -         -          -         P          I        S3?
 Character Sets    -         -          -         P          ?        ?
 Style Sheets      -         -          -         P          ?        ?
 Typed Links       P         P          P,I       ?          ?        ?
 Scripting         -         -          -         ?          ?        ?
 lack of 

S0 S0 O O O O _____________________________________________________________________________ read into SGML - - - - - - editor _____________________________________________________________________________ Omittag - - - - - - _____________________________________________________________________________ Legend: P = Proposed for inclusion in HTML I = Public implementations (at least 2) available S0 = Standard, level 0. Required for minimal conformance. S1 = Standard, level 1. Required for level 1 conformance. S2 = Standard, level 2. Required for level 2 conformance. S3 = Standard, level 3. Required for level 3 conformance. O = Obsolete

Descriptions of each of the items in the issues/features column:


Although forms were intended for standardization in HTML 3.0, many browsers have implemented the design specifications and made them a defacto standard of the language if not the entire community of browsers.

Effective with the release of HTML 2.0 and browsers which claim support for that DTD, HTML forms will be considered a standard part of the language and must be supported to claim conformance.


As per Dave Raggett's HTML+ DTD, still under development, but quite stable at this point.


As per Dave Raggett's HTML+ DTD, still under development, but quite stable at this point.


Still somewhat unstable.

Character Sets

The details of encodings of characters from multiple languages etc., including the interactions with the SGML declaration have not been fully resolved.

Style Sheets

Several proposals have been made, but no implementations are yet widely available.

Typed Links

These give ways to express sentiments such as "document X preceeds document Y," or "document X is part of an aggregate document, whose TOC is Y."


Arbitrary programs inside documents. Not likely as an enhancement to HTML, but an apparently necessary part of the WWW architecture.


When a document has a LI start tag, but no LI end tag, we can say that LI is just an empty "marker" element, or we can say that the LI end tag is implicit. The latter requires support of the SGML OMITTAG feature.

Lack of <P>

Current practice and the original design of HTML is to use the <P> tag as a seperator rather than a container. As desirable as this might be considered by many, it is not in keeping with SGML or facilitating object oriented approach to information management within documents.

Effective with the release of HTML 2.0 and browsers which claim support for that DTD, <P> will henceforth be used in combinatiuon with </P> to create a container for paragraphs.