E-mail List Archives
From: Jukka Korpela
Date: Nov 22, 2002 12:33AM
- Next message: Michael R. Burks: "Help!"
- Previous message: Terence de Giere: "Re: Forms"
- Next message in Thread: None
- Previous message in Thread: None
- View all messages in this Thread
Terence de Giere wrote:
> Some technology (such as a UNIX computer) may not understand
> the Windows character set used for the page although I saw
> nothing in the form that this would make a difference here.
I'm going to comment on that in some length and technicality.
But first something completely different about the form:
I noted that the form contains "placeholder" data in
text input fields. This has often been discussed, and I think
the consensus is that such technique should not be used;
the note about it in the WCAG 1.0 document begins with
"until user agents...". And at present, the technique just
causes difficulties. The browsers that could not
cope with empty input fields are extinct. But, on the other
hand, for fields like phone numbers, the expected format
should be explained in the normal content of the page
(e.g., "Work phone (in xxx-xxxx notation)"; actually
xxx xxxx notation would be better, but I digress).
The "character set" thing is a tricky issue, starting from
the phrase "character set" itself; it is unfortunately the common phrase,
even implied in the charset keyword). In the current
case, we do not even know what the "character set (that is,
character encoding) actually is, for the page,
since on the WWW, it is the actual HTTP headers that are decisive,
if present. If a server sends a document e.g. with the header
Content-Type: text/html; charset=iso-8859-1
then it shall be, and actually will be, interpreted as iso-8859-1
encoded, no matter what <meta> tags there might be in the document
In general, there's no reason to use windows-1252 encoding on
If you're not using any of the "Windows special characters",
you can simply state charset=iso-8859-1. If you are, you
should consider whether you really need them, despite the
problems caused, and if the answer is yes, use some other
approach like character references or utf-8 encoding.
I think it needs to be noted that the character encoding of
a document affects the encoding of form data, too. So if
you switch to utf-8, for example, then the form handler
needs to be able to process utf-8 encoded form data.
I'm afraid this is excessively technical, and on the other hand
not technical enough. But the fact is that when some authoring
tools automatically insert <meta> tags that make a claim about
character encoding, this can cause quite some trouble. Using
window-1252 will mostly not cause problems, but when it does,
they could be really confusing, especially to people who
already have problems in surfing.
Jukka Korpela, senior adviser
TIEKE Finnish Information Society Development Centre
Diffuse Business Guide to Web Accessibility and Design for All:
To subscribe, unsubscribe, or view list archives,