WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: Server-side image maps definition?

for

Number of posts in this thread: 3 (In chronological order)

From: Hoffman, Allen
Date: Mon, Aug 29 2005 4:14PM
Subject: Server-side image maps definition?
No previous message | Next message →

I read that looking for "input type"image" indicates an image map is on
a page. M ai missing something, or is this inaccurate.


Allen Hoffman





From: Patrick H. Lauke
Date: Mon, Aug 29 2005 4:45PM
Subject: Re: Server-side image maps definition?
← Previous message | Next message →

Hoffman, Allen wrote:
> I read that looking for "input type"image" indicates an image map is on
> a page. M ai missing something, or is this inaccurate.

input type="image" does not necessarily indicate a (server-side) image
map. Yes, when clicked with the mouse the browser submits the current
form and passes on X and Y coordinates...something which can't be done
via the keyboard or alternative input devices. However, many times image
inputs are simply used as graphical submit buttons, where X and Y are
actually irrelevant and there's no special behaviour depending on where
a user clicked. In that situation, they can actually work for keyboard
users as well...it just submits the form, but doesn't pass any X and Y.

--
Patrick H. Lauke
___________
re

From: Jukka K. Korpela
Date: Tue, Aug 30 2005 10:40AM
Subject: Re: Server-side image maps definition?
← Previous message | No next message

On Mon, 29 Aug 2005, Patrick H. Lauke wrote:

> Hoffman, Allen wrote:
>> I read that looking for "input type"image" indicates an image map is on
>> a page. M ai missing something, or is this inaccurate.
>
> input type="image" does not necessarily indicate a (server-side) image map.

It depends on how you define "(server-side) image map". It is certainly
true that the program code on the server that processes the form data
may completely ignore the x and y coordinates - just as it can if you have
used <img ... ismap> in HTML.

> Yes, when clicked with the mouse the browser submits the current form and
> passes on X and Y coordinates...

Correct.

> something which can't be done via the
> keyboard or alternative input devices.

Well, browsers just don't support alternative devices - and admittedly it
would be difficult to react e.g. to the user's eye movements; it would
surely be possible with the current state of the art, but it would require
suitable hardware and software.

> However, many times image inputs are
> simply used as graphical submit buttons, where X and Y are actually
> irrelevant and there's no special behaviour depending on where a user
> clicked.

That's probably much more common than using the x and y values.

> In that situation, they can actually work for keyboard users as
> well...

Not as well. Image submit buttons have several problems; most of them are
probably described at
http://www.cs.tut.fi/~jkorpela/forms/imagebutton.html
To begin with, the very reason for using image submit buttons (authors
want to make their buttons look cool) is a strong reason for _not_ using
them, even if we ignore the technical difficulties. If it's a button, it
should look like a button.

> it just submits the form, but doesn't pass any X and Y.

It typically passes x=0&y=0 (which is theoretically inadequate but
manageable in practice).

--
Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/