Archive of UserLand's first discussion group, started October 5, 1998.
Re: Web browsers still don't work
Author: Andrew Duncan Posted: 3/27/1999; 5:54:30 PM Topic: Build interfaces with xml: XUL (cnet article) Msg #: 4602 (In response to 4593) Prev/Next: 4601 / 4603
Hmmm, is XUL really about "bringing the desktop into the browser"? From my read of the Netscape docs, it seems to be about allowing the chrome to be defined in XML, and my first thought was that this would be a great way to define UI elements.
A few years ago I used AppMaker (and others, like CW PowerPlant Constructor) to to the same thing for C/C++ apps -- create windows and dialogs, size them, add buttons, popups, sliders etc.
These days, I want to use the browser as my interface to Frontier, and perhaps the OS. A personal portal is a wonderful thing, which could be made much better if I could reconfigure the interface to suit my own needs. XUL looks like a great way to do that.
Besides, isn't it Microsoft that wants to bring the destop into the browser? The "mucking with" you refer to is an Explorer thing, right? But I agree with the sentiment: if any website fools with my chrome, they're toast, or I'll disable that feature in the browser, or find another browser.
As far I can see, the major benefit of XUL will fall to those developing custom interfaces, for CD-ROM distribution for example. Or people like me who want to build personal portals. OK, this is "bringing the desktop into the browser", but it's *me* doing it, not some idiot web "designer". If it's a screw-up, it's mine, and I can fix it, or live with it, or throw it out.
Or, if I'm building an editorial workflow system for a client, this could be a *really* useful thing to have in my toolkit. This is a vertical market, and IMHO, it's a good alternative to building C or VB or RealBasic apps for writers and editors, probably cheaper to build and the XUL could be on the (intranet) server so that all the clients get upgraded the next time they log on.
Also, I think XUL makes more sense against the background of Mozilla 5 being an embeddable component. It might make sense for Frontier to have the capability to embed that stuff, in which case XML definitions for UI elements makes a lot of sense. It would certainly allow for richer interfaces "within" Frontier, for those that want/need them. There is a lot to like about Frontier's minimalist interface, in fact it's my personal preference, but this scheme could be a good cross-platform MacBird too.
I agree with you that the browsers are broken, but the really broken thing is standards support. Anything less than 100% support equals a broken browser. Let the browser makers innovate by all means, but please, only on a 100% standards-compliant foundation. We won't know the whole story until Mozilla 5 is released, but I suspect that its release will up the ante considerably. Standards-compliance in IE5 sure isn't perfect, in fact it's broken enough now that *sane* webmasters won't take up the new features anyway: lack of compliance already makes site development too expensive, so the only option is lowest common denominator HTML, unless you can mandate which browser your audience must use.
Finally, XUL seems to be a good way to *enable* software engineering in web browser software. I don't think it's some "vision thing", and I've been following the development of Mozilla 5 enough to see how quickly they can rev the interface -- it's fast. And the revs are downloadable, so some kind of "smart update" facility will propagate the bug fixes and improvements quickly too. Is that a bad thing?
I think there's a difference between Netscape and Microsoft: Microsoft is pitching Explorer as a better marketing vehicle (for big portals, ISPs etc), while Netscape is pitching a better browser. We'll see if I'm right, and who wins.
There are responses to this message:
- Re: Web browsers still don't work, Rafe Colburn, 3/27/1999; 7:16:13 PM
- Re: Web browsers still don't work, Dave Winer, 3/28/1999; 5:49:07 AM
This page was archived on 6/13/2001; 4:49:06 PM.
© Copyright 1998-2001 UserLand Software, Inc.