[Blindmath] Accessible LaTeX

[Andrew Stacey]

Thanks for all the interesting replies!

Some of the comments, and the interview that was linked to by someone, made me
wonder why I'm so obsessive about (La)TeX and why I don't want to give it up.
I was particularly taken by that sentence in the interview about wanting to
enter a source that validates, and how he thinks that backslashes and braces
are just so wrong.  To me, that says that the person speaking is someone who
supports TeX-users rather than someone who writes copious documents in it.

For me, TeX is just so right!  And on so many different levels.  I've not
tried to think this through before so I may misstate my thoughts, but here's
a first go at why I like it so much.

Firstly, it is discrete.  When writing a document in (la)tex, then I hardly
notice that I'm writing anything other than the contents of the document
itself.  I don't need to worry about formatting, what it looks like, or
anything like that.  When I do need something more complicated, say
mathematics, then TeX discretely offers a nice simple way to do it.

Secondly, (as someone said in one of the replies) it is a programming
language.  When I need the power of TeX then it's there.  I can define macros
to save my poor fingers: typing \R instead of \mathbb{R} is fantastic!  And
I can make it easy to change global settings, say by changing all composition
of functions from right-to-left to left-to-right, and just know that
everything's going to be alright.  I can do really complicated stuff when
I need to.

Thirdly, as a mathematician then TeX is the epitome of design.  The process
"read, expand, read, expand" is exactly how we parse mathematics!  Keep
expanding the definitions in a proof until you reach something that you
understand.  So it's just beautiful that something so close to what I do
professionally is also at the heart of how I present the results.

So I'm not going to give up TeX without a huge fight.  No one's going to get
me to have input that can be "validated".  Yuk!  What a horrible phrase!
I want my mathematics to be validated, sure, but I'd like the freedom to
really mess up the paper in the meantime.

However, lest the mathml brigade lose heart, just because I'm going to go on
writing TeX for as long as I've a finger left to hit the keyboard with,
doesn't mean that I expect anyone other than me and TeX to read that.  Those
who say that they like LaTeX on the web have probably never tried reading the
source of one of my papers!  What I expect they mean is some sort of
simplified LaTeX with no funny macros and the like.

As far as reading what I write is concerned, then I'm convinced by the MathML
argument.  So for me, an ideal situation would be to have a decent, robust,
(La)TeX to MathML converter.  To my mind, MathML is an output language, like
PDF, PS, or DVI.  So I'm quite happy to compile my LaTeX documents into MathML
to make them more accessible.  If only there were a converter that could cope
with my macros ...

Andrew

PS As an indication of my level of TeX addiction, I even wrote the invitations
to my daughter's birthday party using TeX.