From standards@mercury.projectcool.com Tue Dec 7 18:21:57 1999 Return-Path: Received: from mercury.projectcool.com ([206.14.73.2]) by smtp3.fas.harvard.edu with SMTP id SAA28167; Tue, 7 Dec 1999 18:21:46 -0500 (EST) Received: from fas.harvard.edu by mercury.projectcool.com with SMTP; Tue, 7 Dec 1999 15:33:38 -0700 From: "L. David Baron" Date: Tue, 7 Dec 1999 18:19:05 -0500 (EST) Message-Id: <199912072319.SAA15869@is04.fas.harvard.edu> To: standards@mercury.projectcool.com Subject: RE: [stds] Are we loosing it? Sender: Precedence: Bulk List-Software: LetterRip Pro 3.0.4 by Fog City Software, Inc. List-Subscribe: List-Digest: List-Unsubscribe: Status: R On Tue, 7 Dec 1999 14:19:28 -0800 , Chris Wilson (cwilso@MICROSOFT.com) wrote: > The future of markup appears to be all about inventing new tags and stirring > up a big ol' pot of tag soup - witness Netscape's XUL, for example. I'm not > trying to imply they've done anything wrong with XUL - just that they have > invented their own set of semantics and a schema to go along with it, and > that's just fine (since it is legitimate XML). However, there are > absolutely no brakes anymore on inventing tag sets, and no compelling > reasons to converge on a standard tagset for an area unless the need itself > is compelling. Too bad, IMO. I agree with you on this (for a change :-). I don't think the future of the web is in arbitrary XML tagsets. Why not? User agents won't have any pre-defined understanding of them and therefore won't know what a sensible default presentation would be. People are likely to define only a one or a small number of presentations (in CSS, XSL, or whatever) for their markup language, and therefore it won't be device-independent. Since there will likely (despite the current inaccessibility of many web pages) be an increase in the number of alternative Web devices in the near future, it will become harder and harder for the authors of an XML application to suggest sensible presentation for all possible devices. This is why CSS is so neat. In a markup language that the user-agent *does* have knowledge of, the user-agent can have sensible defaults (often modifiable by the user), and the author of a document can make presentational suggestions that can enhance the presentation in the media with which the author is familiar, but don't destroy it in other media. The future of the Web (as I would like to see it, anyway) lies in markup languages such as XHTML, MathML, and SVG where user agents can have a default presentation. (In CSS, such a presentation is described by a UA stylesheet and/or user stylesheet. I don't know how it is described in an XSL-based formatting model. This is my biggest objection to XSL.) However, I do disagree about your use of XUL as an example. I don't think XUL is intended as a language for web content. It's a language for designing the chrome of Mozilla and any other programs that use the Mozilla source code as a base. Using an XML application for such a language was an obvious choice since Mozilla already had the code needed to parse XML. In my opinion, XUL is too limited to be of much use for web pages (at least in its current form). Just because arbitrary tagsets / schemas don't make sense for use on the web doesn't mean they aren't useful for internal use within a company or a project. That's essentially what XUL is. -David L. David Baron Sophomore, Harvard (Physics) dbaron@fas.harvard.edu Links, SatPix, CSS, etc. WSP CSS AC ------- The Web Standards Project http://www.webstandards.org To unsubscribe mailto:standards-off@mercury.projectcool.com OR Send a blank email to: standards-off@mercury.projectcool.com Hosting donated by Project Cool, Inc. http://www.projectcool.com