Discussions

My post conjecturing the possibility of a better browser

Let's say someone was writing their own web browser that would parse HTML according to any DTD (whatever specified by the DOCTYPE, SYSTEM or PUBLIC).

And this browser would support CSS for all elements. So for most simple things, supplying a CSS style for it would be enough.

Are there any non-standard elements that you'd want it to support (e.g. FIG, AUTHOR, PERSON, WARNING, etc.) in the default style sheet? And what do you think the default rendering of some standard elements should be?

And for elements like FIG that can't be rendered just based on a CSS style, how should they be done? For example, would you like to see tables use internal scroll bars, so that if TABLE wouldn't fit inside the window, the TBODY rows would scroll while leaving THEAD and TFOOT rows at the top and bottom?

(If you want to make up your own element, could you give a DTD fragment for it?)

And let's say this browser had the following features. How many people would be interested in it?

• support for most elements of the HTML4.0 proposed recommendation (which might be an actual recommendation by the time it's done)
• CSS level 1 and most (maybe all) of level 2
• useful rendering of all HEAD contents (LINKs would be displayed, META information would be available, alternate stylesheets would be supported via the TITLE and MEDIA elements)
• useful rendering of REL attributes for A and LINK elements
• provide actual error messages from the SGML parser inside
• user style sheets for all supported MEDIA types (plus ability to over-ride or ignore author stylesheets)
• parse according to any DTD provided
• Unicode support (as far as possible on whatever system it runs under)
• possibly output text to any file or device at the same time as the graphical display (to hopefully support text-to-speech converters over serial connections, e.g. it would send out the TITLE and HREF of links when they are made active)
• possible full keyboard support (necessary if the above is implemented)
• user-controllable USER_AGENT, REFERER reporting
• printer output only for Postscript printers (the browser would use a Display Postscript-like rendering engine)
• possible support for scripting using the event attributes of HTML 4.0 (maybe ECMAScript (it would probably only support the ability to change the STYLE of some element in the document))
• some type of Document Object Model (it would inherently have an internal DOM, but the W3C DOM is very incomplete, and out of necessity I've started a DOM of my own.)
• possible support for Java applets (NOT likely)
• possible support for MathML (NOT a high priority)
• NO support for FRAMEs
• NO support for cookies
• NO support for reading mail or news
• NO support for sound of any kind (not excluding possible future support for built-in html-to-speech)
• NO support for FONT elements or COLOR & BGCOLOR attributes (CSS support for these things though)
• NO support for server-push features (channels, etc.)
• NO support for ratings censorship of any kind (e.g. PICS)
• NO useless features which have nothing to do with browsing HTML documents

And finally,

• distributed as a Java application under the GNU Public License (not necessarily for free)

A browser with these limited (albeit useful) features wouldn't be too hard to do, although it wouldn't happen over night.

The announcement I made a few days later

I am working on a browser. I mentioned some ideas of what I want it to be in the thread "how would you want HTML rendered in the dream browser of the future?" that I started in ciwah.

I'm planning on making the browser rather small and a bit strict. "pedantiWEB" is the name I'm using for it.

I'm doing the browser entirely in Java and it will have to be distributable under the GNU Public License (which means any code I use has to also be redistributable.)

I've already started coding some of the classes needed for the basic SGML elements, and I've got some designs for classes for CSS properties.

For rendering, I already have a Postscript-like graphics class which should make it very easy to do most of the drawing of HTML.

The part that's going to need the most programming work is the SGML parser. I want the browser to have a real SGML parser, so that it can handle both XML and HTML and parse any custom DTDs. I haven't found any Java SGML parsers, so I'll either have to make my own, or try to port James Clark's SP parser from C++ into Java.

I'm also going to spend some time working on a "perfect" DTD which contains many useful elements that currently don't exist in HTML (SECTION, HEADING, PERSON, PLACE, THING, NAME, etc.) I'm hoping to get some help from people in ciwah for this. This DTD may be done as an extension to HTML, or maybe as an XML DTD. I'll hope to get some help on this.

In the next few days I'll be setting up a web page describing exactly what I'll be trying to do with the browser and with the DTD. Although I think it would be nice if discussions about it could go on in the newsgroups, I would set up a mailing list if people here begin to get annoyed. Right now, I think any discussion in the newsgroup or a mailing-list should be about the design of the DTD.

pedantiWEB, a GPL, Java web browser

Last Updated on Wednesday, December 03, 1997.

$Id: discussion.html,v 1.2 1997/12/12 16:36:23 bereza Exp bereza $