W3C home > Mailing lists > Public > www-html@w3.org > February 2001

RE: Point of order!

From: Charles F. Munat <chas@munat.com>
Date: Tue, 6 Feb 2001 13:19:20 -0800
To: "'W3C HTML'" <www-html@w3.org>
Message-ID: <005301c09082$790801d0$0100a8c0@aries>
Scott E. Lee wrote
"But, if the site developer uses unsupported technologies, how can an
existing UA read it?  Sheer force of will is not always the road to
acceptance."

Reply:
Although this started as a discussion about client-side includes, I was
actually responding to Dustin's comments, which seemed far more general. You
are right that client-side includes would be a problem for legacy browsers.
There would have to be a way to permit graceful degradation. Or they would
have to be limited to use on dynamic sites where the server did the includes
except when it knew that client-side includes were enabled on the user
agent...

I am certainly not advocating for sites that lock out legacy browsers. In
fact, I go to great lengths to ensure that all my sites are viewable on
*any* browser.

Scott:
"But there is still the UA upgrade problem.  I am using NS4.7 because my ISP
(NetZero) doesn't support NS6.  What would my option be if I wished to view
an
XHTML or XML site?  Even if I did upgrade, what about AOL, IE, Mosaic,
WebTV?
You can't force everyone to upgrade just so you can write tighter code
(Unless
you are MicroSoft).  Or am I way off base?"

Reply:

XHTML works beautifully on NS4 (though CSS doesn't) - certainly better than
sites built with MS-proprietary code.

Netscape 4 and IE 3 represent a problem. By using XHTML Strict you can
design sites that will degrade very nicely on older or non-CSS browsers
(including CSS-browsers with CSS turned off). Then you can use CSS to
improve the presentation for CSS-enabled browsers (at this point, well over
90% of the browsers in use are CSS-enabled). Things get even better when you
combine XHTML with XSLT. You can convert to VoiceXML, WAP, etc. (maybe even
SVG) - but it will be a long time before these are really widespread, and
they will never be commonly used by amateur web site designers (unless they
can be done using a WYSIWYG authoring tool). In contrast, *anyone* can learn
XHTML Strict.

The problem with NS4 and IE3 is that their implementations of CSS are
terrible. Actually, I've found that IE3 usually doesn't present too much of
a problem, but that may change as newer browsers implement CSS2 and then
CSS3 and people start using more of CSS (but I doubt it). The real problem
is NS4. It is still used by probably 5-10% of users, and its implementation
of CSS is truly horrible. On most of my sites, I end up using either the
server or JavaScript to serve a separate stylesheet to NS4. It's the only
way. And I have one site where even that didn't work, and I ended up serving
a page with tables for layout to NS4 and a "proper" page to all other
browsers.

I doubt that anyone still uses Mosaic (though I wouldn't be surprised -
there used to be hundreds of Mosaic spin-offs). As I recall, Mosaic didn't
support tables, and given the number of sites that now rely on tables for
layout, I suspect that Mosaic would be too frustrating to endure. WebTV
presents some problems (and I sometimes serve a different stylesheet to
them, too). It doesn't seem to handle font-sizing properly. I think it bases
h1 fonts on its default h1 size, rather than on the page default. So if you
say font-size: 2em on a heading, you get twice the default heading size (I'm
not positive, but it seems to work this way). Also, beware of text-align.
When I center or right-align a row of links, WebTV likes to stack them.

But - correct me if I'm wrong - when WebTV upgrades they upgrade all the
boxes automatically, don't they? So that doesn't have to be a problem.

Still, I think the time is quickly approaching when we are going to have to
put pressure on NS4 users to upgrade (even downgrading to NS3 would be an
improvement!). The sooner we can do this, the better. As for you, Scott, if
you cannot switch to NS6 (I'm not sure what you mean when you say that
NetZero doesn't support NS6) or IE5.5, why not try Opera 5? It is now free
(if you accept the banner ads) and it works pretty well.

Scott:
"But thinking that the big guys will buy
into new technologies or code styles that may not be viewable by millions is
a
risky proposition at best.  As a designer, you would have to maintain
backward
compatible pages for the "Mom and Pops" out there that either don't know
how, or
don't want to upgrade for whatever reason.  Does that make it easier on the
designer?"

Reply:
As I said, I'm all for backwards-compatibility when workable. But I think
you overestimate the concern of the big guys for Mom and Pop. Look at how
many sites REQUIRE cookies. Look at how many sites use proprietary IE code
or require JavaScript to work properly. Many big sites either don't care or
they serve other (usually non-equivalent) pages to other users. There is no
reason to think that this would be any different with any new technology
(like SVG). The problem, as I see it, with things like SVG or client-side
includes is that many people will use them with complete disregard for those
using legacy browsers. I don't have a solution to that (other than
education), but I don't think the Web should come to a standstill just
because there are a lot of ignorant people out there.

Would it be possible to create a plug-in to permit client-side includes?
That's an interesting question. But it still wouldn't address those browsers
that don't use plug-ins.

Scott:
"Which I am all for.  Reliance on IE, NS/Mozilla by the masses (myself
included)
is scary.  But the new UA would have to be all encompassing, free, easy to
use
and understand, and easily integrated into the OS of the user.  Maybe this
is
what "Ginger" is?"

Since Mozilla is free and can be easily integrated into other software, I
suspect that we will eventually see a slew of new browsers incorporating
Gecko - if Mozilla can ever get it perfected, that is. I, for one, am hoping
that open-source turns out to be our savior. For this reason, I think it is
vital that we keep the pressure on Mozilla to concentrate on absolute
standards-compliance with W3C technologies and to stop wasting time on
useless feature bloat (programmers evidently love to add features and hate
to work on getting the code right).

But the secret is: create the demand. Build the demand and the programmers
will come. And the way to create the demand is to teach users (and
developers) about standards-compliance and interoperability. Not using
technical terms, but by giving concrete examples of what neat things they'll
be able to do once this is a reality. We must convince people that
standards-compliance = power and freedom. And that shouldn't be too
difficult because they do.

Scott:
"The support will probably have to come from an upstart that
designs a plug-in or module that helps the existing UA's to render the
pages.
Then the big guys will either adapt or get left behind."

Well, actually, they the big guys will just buy the upstart. But you're on
the right track, IMO.

Charles F. Munat
Received on Tuesday, 6 February 2001 16:15:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:45 GMT