[Mailman-Users] What would your dream Mailman web interface look like?

[Stephen J. Turnbull]

Executive summary:

1.  Some aspects of accessibility (providing text alternatives for
    non-text media) can be treated like translation (and will increase
    the burden on translation!)

2.  Frameworks need to help point out the "pain points".  Like: "Yo! 
    there's an ALT-less image here that you can tag!"

3.  Frameworks can and should take advantage of embedded comments in
    non-text media.

4.  I personally don't have much clue about partially sighted (color
    issues or extreme myopia), or those who lack the dexterity to
    mouse around.  However, frameworks can do a lot of the work for
    improving accessibility for those less dextrous (tab navigation),
    and maybe for partial sight?

David Andrews writes:

 > I know what you say is true.  Nevertheless, it makes me sad.
 > Twenty percent of the population has some sort of disability, yet
 > accessibility just isn't taught in computer science courses.

The 85% of the world which strongly prefers to use a language other
than English bothers me more, to be honest (especially since that's
100% of my dayjob users).  Funny thing, American universities don't
teach Japanese and Chinese to their CS students.  Sad, isn't it?

The only accessibility tool for the web that I'm familiar with is the
ALT attribute for IMG and other non-text elements of HTML.  I should
think that "filling in the ALT blanks" would be amenable to the same
kind of volunteer effort that provides our natural language
translations.  I hope that a similar kind of separate effort can deal
with many of the accessibility issues.

Obviously, the incentive will have to be somewhat different.  Unlike
translators who generally get involved for their own convenience and
have some English ability that grows stronger with the practice, blind
people aren't going to be able to write the descriptions of images or
video that they need for themselves in many cases, and it's unlikely
that they'll improve dramatically with practice. :-(  But it should be
quite possible to recruit sighted volunteers for the task.

I'm not sure how to deal with the partially sighted, although I
suppose use of relative dimensions and some CSS is helpful.  But this
is exactly where frameworks can help.  I'm even less sure what to do
about colorblindness, and people who lack sufficient dexterity for
"pointing and clicking".

 > The Justice Department has already said that the web is a place of
 > public accommodation, and the ADA applies.  It is only a matter of
 > time before they issue specific regulations.  So, in the near
 > future, anyone producing publicly facing web sites will need to do
 > this!

No, they won't -- they can always shut down.  And I suspect that's
exactly what will happen to most volunteer sites if they try to apply
the ADA standards to them.  I can't be happy about that.  I live in
Japan, and I assure you that public policies that equalize benefits by
reducing the average suck -- especially for the less-well-off.

 > Using a current, "industrial-strength framework" is not a guarantee
 > of accessibility, and passing the buck to them will ultimately not
 > hold water.

It had better at least reduce the leakage to a trickle!  If it doesn't,
accessibility isn't going to happen.  Content accessibility really has
to be a matter of a separate volunteer effort (especially since every
ALT attribute increases the burden on our translators!!), with most of
the "pain points" being automatically pointed out by the framework.

And things like checking for color and sufficient size of clickable
elements and the like should be automatable.

Realistically, programmers are *not* going to do this in general.  In
the case of Mailman, we'll do some of it, but it's hard enough to
create a usable site for ourselves, let alone the sighted and nimble
in general -- I doubt accessibility is something we'll get to in
Mailman 3.0 (except to the extent that general usability principles
provide a strong basis for accessible pages, as apparently happened
with Mailman 2).

Really, you accessibility advocates should be looking for
opportunities to organize this kind of effort.  The translators prove
that such volunteer efforts are practical, and most translators seem
to be willing to donate effort to apps they don't use when necessary
(obviously, experienced users are preferred).  Translators might also
be willing to do this kind of thing.

But translation only really took off with the development of gettext
-- earlier systems were just too painful all around.  So you should
also be looking for ways to improve the frameworks.  For example, most
image, audio, and video media now provide for textual comments
embedded in the stream.  While most content authors do not (yet)
provide useful comments, some do and others would be encouraged to do
so if the frameworks automagically extracted them and provided them as
default ALT attributes.

 > By the way, the web UI for Mailman 2.X is very accessible -- at least 
 > for blind persons.

That's nice to know!

 > If anyone has an actual site I can get to, I will take a look.

I'll talk to the developers here at PyCon and search the mailing list
for URLs.  A few real lists are running with Mailman 3 + Postorius +
HyperKitty already, and I'm pretty sure there are a couple of demo
sites.

Regards,
Steve