W3C home > Mailing lists > Public > www-html@w3.org > February 1996

Re: lynx and image maps

From: Foteos Macrides <MACRIDES@sci.wfbr.edu>
Date: Thu, 01 Feb 1996 10:58:34 -0500 (EST)
To: brian@organic.com
Cc: www-html@w3.org
Message-Id: <01I0P247NRJM005ZC9@SCI.WFBR.EDU>
Brian Behlendorf <brian@organic.com> wrote:
>On Wed, 31 Jan 1996, Foteos Macrides wrote:
>> 	There is nothing in the HTML 2.0 or subsequent protocols which
>> makes image maps via IMG tags adequately accessible to non-GUI clients.
>Sure, but the imagemap functionality on the server-side could.  When 
>a client makes a request to a map resource without coordinates, the cgi 
>program (or the functionality in the server if the imagemap-functionality 
>is built-in) could return a menu of the links available instead of the 
>silly "your browser does not support imagemaps" error message.

	You've made that point and we've discussed it at length before on
the lynx-dev list, but perhaps it's worthwhile to review the issues on this

	Lynx, like the CERN LineMode browser, used to act on image map
links by sending the request without a coordinate pair, which is technically
an error condition, and unless the WebMaster is a programmer and has hacked
the server's image map handling code or script to assume this is a request
from a text client and return a worthwhile default file for text clients,
an error message will be generated.

	Sending a 0,0 coordinate pair is perfectly legal within all
existing http and html RFCs and working drafts.  Indeed, it would be
perfectly legal for Lynx to prompt the user for a coordinate pair,
offering 0,0 as the default, but allowing the user to probe more
exensively in the dark (within coordinate pair limits).

	When someone, such as the person who started this thread, indicates
a desire to use the Web for Global Information Sharing in a manner which
takes advantage of image maps for sighted visitors with GUI clients, but
ensures that the information will be available to all, the suggestion to
modify the server's image map handling code or script is not always helpful.
But anyone, no matter how new to the Web, can be instructed to include a
brief ALT string and make sure that the top left-hand corner of the image
returns a file suitable for text clients.  That can be done by making the
map's default such a file and not having a hot zone in that corner, or the
image could indicate explicitly that "clicks" in that corner will return
such a file (so GUI clients can select it overtly as well).

	The FIG tag in the expired HTML 3.0 DTD eliminated any need for a
"Lynxism" of this sort (though, of course, it was retained for backward
compatibilty).  If FIG is now crippled for image map handling, and MAP
becomes the sole markup for it, here's what can be expected.
<CAVIAT>I do not have a functional crystal ball</CAVIAT>

	The RFC for MAP includes ALT strings in the example markup.  The
Netscape instructional files do not, and are a better predictor of what
can be expected in the "real world".  Instead of a rich, alernative
presentation based on real markup, Lynx users should brace themselves
for seeing:



 Foteos Macrides            Worcester Foundation for Biomedical Research
 MACRIDES@SCI.WFBR.EDU         222 Maple Avenue, Shrewsbury, MA 01545
Received on Thursday, 1 February 1996 10:57:55 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 30 April 2020 16:20:17 UTC