W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2000

Re: text:graphic [2]

From: Al Gilman <asgilman@iamdigex.net>
Date: Sun, 08 Oct 2000 10:35:49 -0400
Message-Id: <200010081413.KAA1119182@smtp2.mail.iamworld.net>
To: <w3c-wai-ig@w3.org>
At 01:11 PM 2000-10-08 +0100, jonathan chetwynd wrote:
>If you are using Explorer and windows98 you might like to visit:
>http://www.signbrowser.org.uk/2k/testing/champagne.html
>can anyone advise me how prevalent this type of linkage is acroos the web,.

It is not prevalent, but there is a standard.  That is if you spell the
file: URL correctly.  The page you cite uses a Windows syntax filespec
rather than an Internet-standard file: URL but the capability is there
across all platforms if you use the standard syntax.  

However file: URLs as resource references have addressing problems.  It is
hard to have all the people who are going to use the same resource have it
at the same filepath in their local file system.

If a file: URL used in a Help or diagnostic page supplied by the Operating
System vendor made such a reference into a system-reserved area then the
reference could be sure that the right stuff will be in the right place.
But people don't do things that way.

Recognizing objects by their earmarks is more effective than by their
location.  

Don't get yourself sidetracked on trying to bend the service out of shape
to conserve bandwidth.  You should be focused on how effective the
verbal:visual mapping resource is in the imagery it presents to convey an
idea.  A search engine can be instructed to care about the size of what it
finds, and only bomb you with megapixels when there is no plausible
alternative.

Compare and contrast what you get with GuruNet with what you get with
Webster.  Each has its place, but if you want images, you aren't going to
find a good dictionary that covers the variety of words, and something that
works off found resources is a better bet.

If SVG succeeds, and if we persuade an appreciable fraction of the SVG
sourcing authors to include descriptions [svg:desc elements] for the
obvious, articulable objects in their material, we would have a resource
for you in spades.  You grope the descriptions starting with the word you
want to explain and with your thesaurus in the other hand.  Then you filter
what you found and figure out (it could be a ring of visuals around the
offending word) how to present the top contenders in what you found.
Voila!  Visual GuruNet.

Al
Received on Sunday, 8 October 2000 10:13:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:13:50 GMT