Re: [media] Making Sites Accessible Makes Sense For All Customers

>      Lynx was less than 1% of browsers

Lynx users will always be underestimated; a fairly standard question
is "what User-Agent string should I fake?".  That's because a number of
people have encountered sites that adapt to Lynx by refusing to 
talk to it (or by giving a badly degraded page).

It is particularly difficult to get statistics from well designed web
sites as they will have a very high cacheability and will not even
force end to end cache revalidation.

Sites that only count home page hits will not detect the number of
people who give up on them.

>      JavaScript was used on 37.45% of sites

On a number of occasions Microsoft have advised that Javascript be
disabled in their browsers in security bulletins, pending the release
of a hot fix (many people will not have installed these).  I believe
some companies have company policies to disable Javascript.  Most IE5
security problems have probably been with scripting or with ActiveX
controls incorrectly marked as safe for scripting.  IE5's untrusted
security zone disables scripting.

As well as not being formally documented anywhere, DOM0, which is what
most people tend to mean much more than ecmascript, when they say
"Javascript", would require major re-writing of Lynx.  This is good for
commerce because it raises barriers to entry to the browser market!
(The original concept of HTML - and how it differed from its contemporary
solutions such as PDF and early versions of MS Word - was that there should
be very few barriers to entry, particularly for users, but also for
browser designers.)

Amongst UK e-commerce sites, I expect to get blocked by navigation
features that fundamentally require Javascript more than 50% of the time.
Even parts of the BBC have done this recently.

I've had a short correspondence with a blind (I think newly blind)
person on the Lynx mailing list and they say that there is no way that
they can afford JAWS.  I'm trying to encourage them to look at this list,
but I don't know if I've succeeded.  (They got stalled by a site that
implemented MD5 authentication as a Javascript handled form rather than
within the HTTP protocol.)

Received on Wednesday, 14 February 2001 18:54:09 UTC