[Irl-dean] Email fonts

[Barry McMullin]

On Fri, 9 May 2003, Mark Magennis wrote:

[I'm mindful of Matthew Ovington's very fair comment about this
kind of thread; but I also feel there is a significant issue
worth commenting on here...]

> Ah true Bob, but to be fair Brenda didn't use a fixed size font, so you can
> increase it to whatever size your eyesight prefers. Do you think it is
> reasonable to expect people to be able to do this in their email reading
> software? There is another issue with Brenda's email though, and with your
> reply. Both are HTML rather than plain text. I've noticed that there are
> discussions on the WAI list to the effect that HTML should not be used for
> email since it may cause problems for some people. I have never had the time
> to try to understand the arguments around this but I am curious since, as
> Barry often points out, properly marked up HTML should be a good solution
> for universally accessibility. Also, I notice that many people within NCBI,
> including blind people, use HTML email. So I'm wondering what the problems
> might be in practice. Can anyone tell me whether we should try to avoid
> using HTML mail in general and on this list in particular?

It seems to me that *in theory* HTML email might be more
accessible for at least some users.  In practice I doubt this
actually works that well.  Most email clients offering HTML
composition seem to use almost exclusively *presentational*
features of HTML rather than the *structural* features on which
accessibility hinges.  That's precisely the effect Bob
encountered, and which actually caused an accessibility *problem*
for him.  Further, HTML seems to me to be a cumbersome vehicle
(at best) for interactive messaging (where one wants to easily
intersperse quotations and annotations with prior material).

As against that, the key potential accessibility advantage that
HTML messaging seems to offer is text reflow (particularly when
forcing a larger font size) without horizontal scrolling.
Similarly, for some users with speech output, reflow to relatively
long line lengths can be beneficial. But all this does assume a
client that have been specifically configured to match the users'
specific requirements. 

Another factor is that many email clients are extremely bad at
composing plain text email (both outlook express and netscape
messenger fall into this category).  A common defect is to send
paragraph blocks with no line breaks.  This is apparently (?)
some sort of attempt to support reflow - but it is completely
outside of all relevant standards, and typically breaks if the
message is translated into HTML (for online archival).  Thus you
get the dreaded HTML page with a <pre> line hundreds of
characters wide requiring horizontal scrolling to view.

My conclusion?  There is no clearcut answer here - and that is
why this particular list has not (yet) taken any policy line.
(We do have a policy line against attachments of any type, but
that is another day's discussion...)

I'd be delighted to hear opinions one way or the other: but I do
now suggest that further discussion be taken off-list!  If
something substantive comes from off-list discussion I'll happily
summarise back again.

Best,

- Barry.

-- 
Barry McMullin
http://www.eeng.dcu.ie/~mcmullin/