RE: The Philosophy of @alt (was [html4all] several messages about alt)

Henri Sivonen wrote:
> I think it isn't worthwhile for me to address your other points when
> it seems you don't believe that the primary functions on a camera and
> a photo hosting service are capturing pixels and hosting them
> respectively.

I accept that. This however is a re-direct on your part that skirts the real

Those images are not presented in a vacuum - they use HTML as a binder to
gather and present the files as web content.  If all you want is an FTP link
for people to look at pictures then we have no debate: present your tree
view of the FTP directory complete with cryptic file name(s) and browse away
(there's a paved cowpath for you). But that's not what the majority want

The moment you add the additional layer of the HTML container, then we are
talking about a different beast altogether, as you are seeking to make those
cryptic file names meaningful to those who can visualize (the majority) by
adding essentially a GUI interface - yet the current proposal says
"Pfttt..." to those who cannot work with a GUI interface (the minority), and
who instead require semantic logic and structured, complete disclosure of
information.  The current proposal of optional @alt suggests that "cryptic
file names" or worse *NO INFORMATION* should be deemed acceptable to some
users, simply because the majority *might* not provide such, and that
automated systems should not take some responsibility for providing any kind
of information. It suggests that somehow the very same genius
engineers/developers who will unveil heuristic image analysis some day in
the future cannot figure out a better solution than making an established
convention today optional moving forward.

> As for social engineering, if you cannot acknowledge what the people
> whose behavior you are trying to modify are seeking to do without your
> intervention, you can't develop successful social engineering
> strategies except by accident.

Whereas you cannot come to terms with the fact that if you cannot or do not
seek to change peoples behaviour, and in fact remove any moral or social
pressure to do so, that change will *never* happen.  

Throughout the majority of the 20th century, most people saw cigarette
smoking not only as socially acceptable (movies, television), but even
"normal", "sexy" (Marlboro Man) or "cool" (Joe Camel) - that's the way it
was, and using the paved cowpath logic, that's the way it should have
stayed.  Times have changed though, albeit slowly, and there is still
resistance today from some people who do not want to quit smoking, and the
pipeline continues to be filled with new, young smokers, despite the
overwhelming evidence that now shows smoking is bad for you.  What really do
you not understand about social engineering?  

You cannot change everyone's thinking, especially over-night, yet societally
speaking smoking is no longer considered "normal" or acceptable, and those
who continue to do so are now ostracized by "rules" that infuriate some (as
they stand huddled 10 meters away from the building entrance at "break
time"), but have - over time - created the climate where the greater
goal/benefit is beginning to emerge (and it warrants mentioning that this
change has taken longer than 10 years - far longer than the popularization
of the world wide web).  Restaurants no longer have ash-trays, newlyweds no
longer order embossed matchbooks, new cars do not have ashtrays or cigarette
lighters, etc. etc.  You don't just throw up your hands and say "...oh well,
some people will never stop smoking, so let's stop trying." ("...oh well,
some content authors will never bother to add alternative text, so let's
just make it optional sometimes.") It is now 'against the
rules/non-conformant' to smoke in restaurants, and doing so has consequences
- oh physically you can, but socially you cannot. Should social engineers
stop trying to get people to quit smoking, even when it is clear that for
some smokers, they do not want to be told that smoking is bad?

Sure, we will continue to get some bad alt text (some people who quit
smoking start up again), but gradually the pendulum swings to where we want
it, and suggesting to make @alt optional in the future disrupts the progress
that has been made to date.  You overlook the context of time.


Received on Tuesday, 15 April 2008 18:59:51 UTC