W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2002

Can I use... (was RE: Media - Suit Over Airlines' Web Sites Tests Bounds of ADA)

From: John Foliot - bytown internet <foliot@bytowninternet.com>
Date: Wed, 9 Oct 2002 07:31:01 -0400
To: "Phill Jenkins" <pjenkins@us.ibm.com>, <w3c-wai-ig@w3.org>
Message-ID: <GKEFJJEKDDIMBHJOGLENGEHKCMAA.foliot@bytowninternet.com>

(where Phil paints himself into a corner...)

> Can I use HTML 2.0, 3.2, 4.01, or XHTML?

If it's VALID code, use any of the above.  All have been developed to be
backwards compatible.  XHTML and HTML 4 are more robust than HTML 2

> Can I use W3C's SMIL?

Please do.  Don't forget to provide alternative content for users who cannot
access the info.  This can be as "simple" as a text file, although if you
are creative enough to be using SMIL I'm sure you can come up with a more
creative alternative than that... <smile>

> Can I use W3C's CSS2?

Degrades beautifully.  Not all visual browsers will support every part of
CSS2, but since you've marked up your content symanticly correct, the
message and content will still come through loud and clear, although some of
the web pretty may be lost. Bob will be happy though, it will work
beautifully in LYNX...

> Can I use any W3C specification?

Yes.  Be aware that all of the specs may not be fully supported in all
browsers, but that's up to the browser manufacturers.  As a content
developer, my job is to develop compliant code, not compliant browsers.
That said, I do test in multiple browsers to ensure that what I am serving
up looks "relatively" the same across different applications.  But if a
certain browser doesn't support a specific element "correctly" I make a
value judgement but generally tend to avoid turning myself into a pretzel.
Netscape 4.x sucks... get over it.

> Can I use any public standard? (i.e., gif, jpg,etc.)?

Sure... PNG too if you want (SVG support ain't there yet, although you
*could* use it).  Again, the W3C Standard states that you should provide a
text equivelant for any none textual object, so don't forget your
*meaningful* ALT text and possibly LONGDESC attributes.

> Can I use JavaScript? Why not, it's a public ECMA standard.

If you want to use JS for light-weight, non essential enhancements, have a
party.  Mouseover behaviour spin your wheels? Go for it (use onFocus and
onBlur as well for users who do not use their mouse).  But if you want to
use JS for a Mission Critical function do so at your own risk.  Using JS for
form validation is unreliable - it can be spoofed by the end user (potential
security risk here).  I recently saw a site which used JS for their
site-wide search function, with <a href="javascript:search()"> as the means
to implement the search.  This doesn't work in all browsers, so their
mission critical function was disabled to browsers that don't support JS.
All critical scripting should reside on the server to ensure consistant and
predictable behaviour, that's just a smart decision on your part... With
literally hundreds of free server-side scripts available on the web today
(CGI, PHP, ASP, etc.) there realy is no excuse to use client side scripting
for any mission critical function. (Don't have access to your server?  There
are hundresds of remotely hosted scripts as well, although most are banner

> Why do you insist that the "problem" or "burden" is with the authors and
> not the developers of Lynx that don't support all of the W3C and other
> public standards?  Again I argue that it would be more effective to fix
> Lynx than to fix the millions of applications already written.

While I always urge users to upgrade to the latest software whenever
possible, it may not always be possible, sometimes through stubborness,
although often because of external factors beyond the end users control.  I
have a friend who uses speech recognition technology to access the web (he's
a quadraplegic).  While he has access to both Dragon Dictate and Dragon
Naturally Speaking (the *newer* version), he prefers Dictate as it actually
has better hands-free functionality, as well as a lower CPU overhead - an
important consideration when using a 2 year old lap-top.  When Dictate was
released, the predominant browser WAS Netscape, and so the software has
hooks built into it which *work* better with NN than with IE.  So he uses
NN, not by choice but by circumstance.  It's easy to say "upgrade", although
sometimes it's not so easy to do...

That said, if the millions of applications are poorly written, why should
they be tollerated? (and truthfully, they are not applications, but rather
web pages)  In law, there is the concept of "fruit from the poisoned tree" -
any evidence which originates from a "tainted" source is suspect and not
allowed, even if the eveidence is essentially correct and true.  Fixing LYNX
won't fix broken web pages... besides I don't author pages for specific
browsers, I author them for the web.  It's true, if a user continues to
insist on using NN4.x then they may not get *everything* I serve up, or it
may be a tad funky, but at the end of the day they still will be able to
understand the content of the page(s), because they make logical sense if I
use the standards.  Valid code, scalable text, logical markup using
appropriate elements such as <h1>, <li>, avoiding excessive tables for
layouts, etc. Continuing to perpetuate support for broken browsers extends
the broken browser's lives - a bit counter productive don't you think?

> and why do *you* get to decide that "a 16.5 baud modem" is obsolete? or
> that a 300 baud isn't?
> If I code to HTML 4 will your DOS 2.1 machine get it?

Structurally, your content should make sense, although all formatting will
be lost (this or course assumes that the DOS 2.1 machine only supports a
minimal browser... LYNX maybe?).  But since you are using CSS2 anyway, your
HTML 4 document will have no formatting attached to it since all the "web
pretty" is in your style sheet, so in theory, yes it would "get it".

(yet another opinion)

Received on Wednesday, 9 October 2002 07:31:20 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:20 UTC