Lists Home |
Date Index |
Gerald Bauer wrote:
> Well, using SVG to replace XUL is wishfull thinking.
Not to replace, to implement. Getting people to agree on what's needed
in a UI system is wishful thinking. I sure hate every single GUI toolkit
that I've tried to use.
> SVG is markup for rich 2D graphics. XUL is markup for
> rich widgets (e.g. windows, wizards, dialogs, trees,
> datagrids, splitters, menus, toolbars, and so on).
Rich widgets aren't created out of thin air, they're created using rich
> If you build a complete cross-browser widget set
> than you'll end up with the whole DHTML
> today all over again.
You're thinking in terms of a world in which IE6 or NN4 are the targets
and the future. If that's case, then we're all doomed anyway, or at
least stuck with antiquated tagsoup and gif rendering engines.
> Just imagine when every SVG page
> includes its own widget library in SVG. Why not attach
> a copy of Windows with every page instead?
You have a talent for exaggeration. For one it's still better than
having to download twelve different XUL implementations, and also HTTP
caching has been invented for a reason. Besides, widget implementations
tend to be fairly small in my experience.
> Isn't a
> binary transfer more efficient?
It often is, but that's an orthogonal permathread.
> By the way, why include MathML in your standards
> portfolio when you can easily use SVG to render
MathML has its semantic value, which is worthwhile in itself. Beyond
that, yes you could use SVG to render MathML.
I'm not saying that having a standard XUL is bad idea, just that if
you're targeting desktops then it'll be really hard to get people to
agree, and given the power of desktops why not get them to disagree in
I see value in a simple, stupid, XULish thing for the Mobile (and other
limited devices) domain. I believe Jun Jicrosystems has something for
that, called Jsomething (yeah I know that doesn't help, don't blame me