W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2004

Good and better Re: cool news... (was RE: alt text &...

From: Charles McCathieNevile <charles@sidar.org>
Date: Mon, 28 Jun 2004 02:33:47 +0300
To: "John Foliot - WATS.ca" <foliot@wats.ca>, Kurt_Mattes@bankone.com
Cc: w3c-wai-ig@w3.org
Message-ID: <opr993ulv5w5l938@chaals.local>

On the one hand this is a smart piece of work, figuring out how to do  
something that a lot of people want to do, in a way that works for the  
vast majority of systems. Another good reason for using an Apache/PHP  
setup (the IE of the server world...).

On the other hand, it seems to ignore some of the reasons for not using  
images for text. People who are capable of dealing with content that  
includes images, and therefore download them, may not be able to read your  
particular font in the header. People with some vision might use the  
white-on-black  setting offered as a standard display mode by any serious  
operating system, with a font that suits them. Within these constraints  
they are happy to try and look at an image, if it's more or less explained  
and they are just getting an additional idea. With a heading turned into  
an image, they're back to the old problem - control of contrast is dodgy  
(in most cases it should work, because most designers are too smart to  
make headings almost unreadable, but not all).

I am intrigued by how it works for people who download some images (the  
way I work on my phone, and from time to time on my laptop) - to have the  
alternative available unless they decide to try the image. Or vice versa.  
I'll look over the code in more detail.

Yesterday I came across the first example I met of SVG used totally in the  
wild on a completely commercial site (for ordering a taxi). It's a bit  
over a year since I started coming across it in the wild in non-commercial  
settings, where people are often prepared to experiment more - some with  
crazy ideas, some with thinking about the standards and how to implement  
them as they are designed.

Stylesheet switching is starting to offer the user a chance to get a site  
the way they want it.smarter developments in the mobile world are doing  
even more in that direction - after realising that a technically better  
Web isn't as popular as one that includes what is on _THE_ web (the lesson  
of the WAP/iMode work) people have mostly gone back to trying to make a  
technically better version of the existing web. There is, of course, more  
or less infinite material for debate on what is actually technically  
better. And given that there are constraints in the real world, we also  
get to discuss which trade-off is better.

I can see a motivation for trying to use images for text, whatever WCAG  
says, and finding a workaround that solves the same problems is a  
reasonable strategy. A lot of the people who have come up with very smart  
solutions are seriously committed to standards, and are looking to make  
the Web better - and in general they are the kind of people who do. Our  
job, as accessibility people, is to mmake sure we know as much as possible  
about why a checkpoint or a common practice exists, and whether a new  
technique meets the same needs.

On the other hand, a massive number of these needs can be met by simple  
things like CSS' Web Fonts (little used when they came out years ago,  
available now in SVG so likely to get some slow growth). Authors, and more  
seriously the tools that the vast majority of authors use, go on  
supporting something like the easiest common denominator. That is, they  
neither go for something that can work in every tool, only a selection of  
them (there are grounds for making a selection...) nor something that  
takes a very imaginative or architecturally clear view of a better future.

Like Todd Fahrner's Image Replacement technique, this is a piece of  
careful thinking from someone applying real brainpower. I hope it stands  
up better to the curiosities of the real world of accessibility better  
than Todd's approach. But there are a couple of things that make me  
nervous. One of those things is our tendency to accept very quickly  
something that they can offer to designers, without testing it heavily  
first. We don't yet have better alternatives (for all i believe that  
well-implemented SVG is smarter, we simply don't have enough long-term  
experience with it in this group to explain way, how to use it right, and  
what to avoid). People have commercial imperatives, and quite sensibly  
they are prepared to do something good rather than not putting anything  
online until they can do it better. But that should make us redouble our  
efforts to findd something better than what we offered first.

(One of the things I like about this group is that people are doing  
precisely that :-)

avagoodweekend. And don't forget the Aerogard. (Don't worry, that last  
sentence is an obscure and irrelevant reference to my childhood. Geoff  
Deering might get it...)



On Thu, 24 Jun 2004 15:25:29 -0400, John Foliot - WATS.ca <foliot@wats.ca>  

>> The W3C's WCAG states:
>> 	Priority 2 - 3.1 "When an appropriate markup language
>> exists, use markup rather than images to convey information."
>> It can't be clearer than that.  You keep asking how, many keep  
>> suggesting don't.
> While this may not solve all of your problems, Stewart Rosenberger  
> recently
> posted an article at A List Apart
> (http://www.alistapart.com/articles/dynatext/) which may provide  
> something
> of a partial solution to Kurt's dilemma.

Charles McCathieNevile                                  charles@sidar.org
FundaciĆ³n SIdar                                        http://www.sidar.org
Received on Sunday, 27 June 2004 20:34:25 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:28 UTC