HTML and Forms Activity Proposals

Introduction

Following a long - probably too long, for which our apologies - period of deliberation and discussion, the HTML and Forms Activities are being relaunched with a new structure, new charters, new deliverables and a new focus. This proposal integrates feedback from the Advance Notice (member-only link) sent to the AC in May 2006 and discussions in the Hypertext CG from August to November. Early feedback from Disruptive Innovations, Google, IBM, Microsoft, Mozilla, Nokia, Opera, Origo Services, Sun Microsystems, Volantis, X-Port, participants in the Hypertext CG, and the W3C Team is acknowledged. In some cases the feedback was contradictory; this proposal attempts to steer a middle path.

The aim of this document is to present some background, explanatory material so that the reason for the proposed changes is made clear.

HTML

Over time, participation in the old HTML Working group dropped. In particular, participation by implementors of HTML-aware user agents (i.e., browsers) shrunk to near zero. In part, this was because the features being added to HTML, or the changes (such as modularization) being made to it, were not seen as interesting by implementors.

Previous non-XML versions of the HTML specification claimed to be based on SGML, but did not enforce the constraints needed to use SGML. In practice, SGML processing was not used for these documents.

Starting with XHTML 1.0, the move to an XML syntax gave benefit for compound documents and for machine treatment of HTML documents (such as harvesting) but did not confer immediate benefit for stand-alone documents. This, plus the forgiving nature of existing HTML browsers, meant that there was no clear benefit for content developers attempting to use XML, no new features to encourage them to switch, and well-formedness errors were not reported so tended to propagate.

To address this, a new Working Group is being formed, to add new features (both to the markup and to the DOM) which current HTML usage indicates will help the content creation community. One source of ideas for this format is the Web Applications 1.0 proposal by the WhatWG. Instead of insisting on XML, the same language will allow two syntaxes; one XML, the other suitable for 'classic HTML' parsers. Being an incremental development from the HTML 4/XHTML 1.1 feature set, it is hoped that HTML user agent developers will wish to actively participate in this group.

Forms

XForms 1.0 has a clear architecture, a good separation of presentation from content, device independence, internationalization, and accessibility. It also requires a completely different processor to HTML forms and is not at all compatible with them.

In consequence, XForms is widely used in the enterprise market, where the choice of User Agent can be controlled, and rarely used outside that environment.

At the same time, incremental improvements such as Web Forms 2 were proposed, extending existing HTML form syntax.

To address this, the HTML Working Group will add evolutionary improvements to HTML (in both classic and XML syntaxes). Meanwhile, a new XForms Working Group will develop XForms Tiny, which accepts the existing HTML forms syntax and uses the XForms architecture. XForms 2.0 processors will be able to read XForms 1.0 syntax, XForms Tiny syntax, and XHTML Forms.

To encourage cooperation, we propose the creation of a task force, made up of Participants from both the HTML and XForms Working Groups.

In consequence, implementors and content creators will no longer be daunted by a step function between simple HTML forms and advanced XForms - but rather, encouraged by a steady curve.

XHTML 2

The XHTML 2 specification has been under development for several years. In trying to make HTML a general XML application, it has produced a number of other specifications as side effects - XFrames, xml:id, XML Events, RDFa. Some of these have moved to other Working Groups, such as xml:id, now a W3C Recommendation. Others have been adopted by other groups (such as XML Events, used in SVG 1.2).

Also, XHTML 2 has found use as an intermediate authoring format, for example in DIAL, and is used for server-side content adaptation. However, the number of client-side implementations has remained low.

To address this, a new XHTML2 Working Group will continue to develop these specifications, with a greater emphasis on client-side implementation and test suite development.

Proposal for Membership Consideration