Re: lynx-dev lynx as Test Instrument

[Marco De la Cruz]

Note that your query is probably better suited
for comp.infosystems.www.authoring.html

Martin McCormick <address@hidden>:

>        The people who are putting together our distance learning web
> pages have asked me to help them make the pages more accessible to
> people who are blind.  It is going to be a tall order because many of
> those pages are being generated with Domino or MS Frontpage which is
> kind of like raising a kid on the mean streets and then hoping he will
> somehow do the right things.

Personally, I think proper HTML can only be done
"by hand". MHO.

>         With the exception of not being able to launch multimedia
> players, one should be able to navigate a site perfectly well with

Lynx can launch external programs with no
problem, i.e. you can certainly hear a sound file
if you have the player and the appropriate setup.

> lynx and to read all the text pages and even download pdf documents
> which can be run through a converter program which renders any ASCII
> text that exists within the pdf file.

OK...

>         I am maintaining that if it works with lynx, it will work with
> IE or Netscape or any of the hand full of other browsers that students
> might bring to the table.

Well, yes, but one should be careful:

http://128.100.80.13/marco/alita_text.html
http://128.100.80.13/marco/alita_graphic.html

Both look identical in Lynx, and pretty much the
same with any text browser, and both validate
for HTML 3.2 (using the W3C validator thingy),
and yet...

Of course, I purposely set out to sabotage GUI
browsers in the first case, but it just comes
to show that things can go also wrong with
"text-friendly" pages (this is the exception,
not the rule).

If you design for a GUI browser and it looks
OK in Lynx it's probably OK (I do it the other
way around, though, but I ironically tend to put
too many images sometimes, Lynx is so fast! :)

>         Lynx should illustrate those places where text is missing.

Yeah, well, just make a habit of including an
ALT with IMG, even an empty ("") one if the picture
is just meaningless eye candy (which might make
you ponder, why is it there in the first place?)

>         This and the "bobby" application should give our web folks a
> couple of good tools with which to gauge the things they are trying to
> do.      

Try also: http://www.flfsoft.com/html/html_validators.html

Run your pages through a few of these, they complement
one another.

>         I would like to hear any comments on this approach because it
> is an opportunity to maybe break a lot of bad habits and get on the
> right path.

Avoid any Java/Javascript, and mark the links --
I usually put brackets around them. One can usually
do without frames, but if you must have them give
them a meaningful name, and add a link to the "main"
frame in the <noframe> area.

Avoid using more than one column of text because screen
readers dictate on a per-line basis. Always have
clearly-labelled links to the next/previous pages,
if any, and to the main page. Beware of tables: if
you put them make sure they "degrade" gracefully in
Lynx.

There's probably other stuff I'm missing,
alt.comp.blind-users can provide more feedback,
and there are many web pages geared towards
accessibility. I commend you on your interest
in doing this sort of thing...

Cheers,

_________________________________
address@hidden
Gunnm: Broken Angel
http://128.100.80.13/marco/main.html