Richard Cornford said:
if TITLE was intended to be a tooltip then it
would have bean clearer to call the attribute TOOLTIP
Who said it was intended to be a tooltip? That may be the way MS chose
to display it but other Operating systems/browsers may use other
methods. Mine displays such messages in an alert in the corner of the
screen, which is the way I'd expect it to appear on this OS.
I suppose there's no reason TITLE text couldn't appear in a bar or
something.
ALT on the other hand seems to be intended as an aid for when the image
cannot be properly perceived as an image.
"text that is rendered when the image cannot be displayed" is what the
specs say.
http://www.w3.org/TR/html4/struct/objects.html
it might mean that the
restricted perception is on the part of the user. That would directly
apply to visual impairment but might extend to other forms of
"disability" relating to perception
Surely such people would rely on more specialist browsers or browsers
with modes/features to affect the way the page is interpreted/rendered
to aid them (such as Opera). There's no need for people with good
vision to have features forced upon them that were intended for
visually impaired people. It's more reasonable to expect someone who
can't 'percieve' images to disable them and have the alt text display
instead (or use whatever modes/settings the browser has to aid them).
the user might appreciate the ability to supplement
their problematic visual reading of the image with direct access to the
ALT text.
But the problem with this is that by supplementing something with its
intended alternative may result in confusion. There are many
occasions where my ALT text bears no connection to the image or its
purpose when rendered in a mainstream browser. To show that text
outside it's intended context makes no sense, however when a text or
speech browser interprets the page it makes sense. (in such cases I use
title="" to suppress the tooltips). An example is "Site navigation" as
the alt text of a logo because without CSS the logo is followed by a
list of links. It's only the CSS that puts the links and logo in two
different places in the screen in a modern graphical browser.
If I had a photo of an object or place then the longdesc attribute
would be a preferred method of helping visually impaired users '
perceive' that image. I'd probably not want to use ALT text for that
since a picture paints a thousand words
it would be better not to impose a particular
interpretation of the intention of these attributes on a browsers but
instead to make the possibilities available as user configurable
settings and impose no more than an initial default configuration.
if the specs were so ambiguous the attributes would be next to useless
as different authors would have different opinions about how they
should be used, and different users would have varing settings.. so how
could you ever know if your alt text would work as intended. Users
would find that the nature of alt/title text varied so much that they'd
probably disable it along with Javascript warnings.
Specs are only as good as they are precise.
