W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 2002

RE: magnifiers vs relative font-sizes

From: Charles McCathieNevile <charles@w3.org>
Date: Tue, 16 Jul 2002 00:02:56 -0400 (EDT)
To: "Slaydon, Eugenia" <ESlaydon@beacontec.com>
cc: WAI GL <w3c-wai-gl@w3.org>
Message-ID: <Pine.LNX.4.30.0207152309500.19689-100000@tux.w3.org>

Direct answer: Yes, explain that it is more expensive to code for Netscape
4.x as well as the Web at large. At the same time, you can explain why in
some cases it is more expensive to code for other specific systems as well as
the Web at large. But in my personal opinion it is better strategy to be a
Web designer than a Netscape designer or an Internet Explorer designer.

The rest is fairly philosophical, things I have said before (as I read it), a
bit unstructured and too long. Stop here or enjoy as you prefer...

Summary: Web designers should be professional experts, and should be prepared
to correct clients' misunderstandings.

If they accept a job to do something that in their professional judgement
won't work properly, it would be a surprise if they produced a good result
(and show that their judgement isn't as good as they might hope). But that is
a matter to sort out with the individual client - we do a disservice to the
industry by claiming that something can be used by everyone just because it
is common practice.


I appreciate that the general public don't like the idea that things don't
work "perfectly" on their browser, and that people paying for websites are
reluctant to provide something that the general public thinks is broken. I
know that it isn't the browser that gets blamed even if it is at fault.

The task of making sites is, therefore, a lot more complex than it should be.

In many countries there is a legal obligation to provide accessible services
- in Australia the law is that this applies to anyone providing services,
although private organisations (unlike government ones) may be able to gain
exemption on the basis of unreasonable hardship - and other countries set up
their own rules.

As Eugenia points out, when websites don't work for customers they go
elsewhere. I recently tried to pay a phone bill online (since the alternative
was calling a number that oculdn't be reached from outside Australia), and
when their site didn't work I simply didn't pay the bill for two weeks. I
have large phone bills, and that costs Telstra a measurable amount.

And clearly when the presentation of a site is so unclear as to make it
difficult for people, it has the same problem. Some ability to use graphic
presentation as a communication tool is critical to the success of the web on
"common" desktop computers (probably still more than half the market).

Knowing how to combine these requirements is of course what makes some
designers better than others. The very best are worth a great deal of money,
the very worst probably cost their clients much more in terms of foregone
income and bad public impressions than the client ever hoped to gain.

Many browsers are known to cause problems with different Web content.
Netscape 4 in its many versions, Internet Explorer 3 (Windows) and 4.5 (Mac),
are browsers that have caused me major problems because of the way they
implement (or not) CSS. Other problems have arisen through the way that
browsers implement HTML, XML, HTTP, and so on.

There are some strategies that can be fairly easily applied to serve content
that makes up for a known bug to browsers which claim to be a particular
version. (People often set their user-agent header to be something untrue.
But I believe that people who do it can be relied on to recognise that they
deserve to be treated as the browser they claim to be vis-a-vis bugs, and can
solve the problem themselves.)

Of course this requires a server which can do this. But running a courier
business requires a means of transport too - I think that it is rightly the
responsibility of the developer to ensure that  they sell their client a
solution which can meet the needs, or tell the client why their solution is

Some people will always agree to do work they are embarassed to put their
name to, or try to make the best of a bad situation - that's the nature of a
market economy. But it is going to remain true that some tools simply do not
support the best practice of accessibility, and that some solutions are not
accessible to people with disabilities, however well they meet other needs.

The opposite is also true. Something that works for people who are blind
doesn't automatically work for all people with disabilities, and I have seen
things claiming to be accessible which clearly are not because many people
would not be able to use them at all.

So we need to be prepared to recognise, and indeed become expert in, the
limitations as well as the possiblities of the Web. This will put us in a
position to advise clients better, and do a job that meets their more
realistic expectations.

The Web is not a medium designed to provide layout that doesn't change one
millimetre, and only a fool (or someone who hadn't considered the wide range
of devices used to access the Web, which actually includes very many people
who are not fools) would hope otherwise. It is a medium that allows for
greater accessibility to a wider range of people than almost any other we
have available. It is a medium in development. At the same time it is a
critical tool for many citizens of developed and developing countries.

Fewer than half the world has ever made a telephone call, let alone used the
Web, but it may be surprisingly soon that more people do the latter than the
former, given the relatively low extra cost and high extra potential benefit.
When the majority of people on the Web are using mobile semi-disposable
devices, and micropayments on the Web have caught up with the ease and
acceptance of micropayments by telephone, the marketing demographics will
shift considerably. We would all like to be sure that we understand what is
happening and how to design for the new Web. The best way to get there then
is to be there already...


On Mon, 15 Jul 2002, Slaydon, Eugenia wrote:

>I design for the general public and my clients use any and all browsers and
>will demand that it looks equally good in all. (I'm sorry - the site isn't
>working in which version of Netscape on the Mac? Uuugh.) :)

And elsewhere:

Then the only solution would be to state that we can't code for Netscape 4.x
and to design accordingly, otherwise we will fail on that checkpoint. It
isn't a case of working around and using relative fonts - this is a case of
relative fonts being destroyed (use them and your page format is blown -
they don't work). But I hate to do that. I feel like I'm alienating a large
audience. Granted the page will still "work" but since the font that is
supposed to be a title may end up smaller than the text - I can't say it is
Received on Tuesday, 16 July 2002 00:02:56 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:42 UTC