Re: text:graphic was: XML? (no, caching)

Here's a thought experiment.  Start with what is available in GuruNet
<http://www.gurunet.com/>.  This is a universal "Whazzat?" [What is that?]
tool for text on the screen.  It appears to use screen reader technology
because it is not limited to Web stuff.  Think what it could do if it
understood a user preference for imagery (for all I yet know, it does).

The key is getting the corpus of graphics associated with relevant text.
But this is the metadata version of LONGDESC in a nutshell.  One can get
software to invert the relation.

Al

See also <http://ptech.wsj.com/archive/ptech-20001005.html>.  Walt likes it.

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,.
>I cannot say i have come across it. I had imagined that security issues
>might have precluded usage of links to the hard drive and never bothered to
>try it, how lazy can one get. Of course it does not help if you dont have
>the library, and one still needs a possiblity to downlaod if the default is
>absent.
>
>I recognise that W3C/WAI does not have a magic wand, however if we could
>design a suitable document identifying what was needed, it would help other
>standard setters to identify their requirements. We need an easy to use and
>cheap solution, similar to 'alt tags' in implementation. The web page author
>might not even need to be aware of the change, though WAI will need to push
>the use of plain english for the forseeable future.
>
>My intention was both to not define the problem too tightly and thus allow
>others to express their current understanding, and enquire what efforts
>WAI/W3C (including members and the web) is making to ensure that personal
>graphics libraries might be as accessible as character fonts. This does not
>require the authorization or even design of a 'standard graphics font'.
>There is absolutely no need to tie this project up in expensive and
>unnessary text or applications. Every time one selects a graphic by whatever
>means, a conversion from text(ok this could be binary) to graphic is taking
>place.
>
>I would hope we could identify, what changes are essential to allow the use
>of a local database of graphics and yet retain the advantages of the web?
>Are we sure we need a change to HTTP and could we define it?
>
>Imagine if we had to design a browser that added alt text, the problem seems
>insurmountable yet we know another way was found. Designing a browser that
>displays a known graphic with each word, is simple and solves the problem
>but unfortunately many of the benefits of the web are lost. It remains an
>'excluded' product,
>
>If standard graphics libraries are to be used, can we help define the
>problems and make it easier for library creators to implement.
>The libraries need to be editable by the user(or carer), unlike standard
>character fonts. Yet this can create chaos...
>
>If the local desktop is missing graphics ultimately something like AKAMAI
>might be required. Most LD users might be happy in the first case if some
>graphics were missing, or slow to download.
>
>It might need the design of a particular browser similar to lynx in its
>limitations.
>eg: selected download of graphics, and 0-30 words per page limits(see Dr
>Seuss).
>or would this also exclude unnecessarily.
>
>I probably could go on, but my wife has been very patient this morning, our
>guest are arriving any minute and realistically does this make any sense?
>
>jonathan chetwynd
>
>jc@signbrowser.org.uk
>IT teacher (learning difficulty)
>& accessibility consultant
>

Received on Sunday, 8 October 2000 09:25:46 UTC