Image descriptions that require math

14 views
Skip to first unread message

Thomas Judson

unread,
Jan 13, 2020, 12:34:12 PM1/13/20
to pretext...@googlegroups.com
I am trying to write an image description for the following commutative diagram (AKA alt-text). I don’t see how to do this without using math symbols. PreTeXt does not seem to complain if I write


<m>\psi: E[x]/\langle p(x) \rangle \to F[x]/\langle q(x) \rangle</m>

or

\psi: E[x]/\langle p(x) \rangle \to F[x]/\langle q(x) \rangle

Right now, I am inclined to leave everything in math mode. Any comments on best practices?

Tom

PastedGraphic-1.pdf

David Farmer

unread,
Jan 13, 2020, 1:43:08 PM1/13/20
to pretext...@googlegroups.com

It seems like a step in the wrong direction to have that as an
image. Doesn't MathJax handle commutative diagrams where the arrows
are only vertical or horizontal?
> --
> You received this message because you are subscribed to the Google Groups "PreTeXt support" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pretext-suppo...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pretext-support/6724CE00-BC43-4727-90D2-6A1A50070B44%40gmail.com.
>

Thomas Judson

unread,
Jan 13, 2020, 1:52:05 PM1/13/20
to pretext...@googlegroups.com
Quite possible.  This image might be a relic from times past.  I will look into it.

Mitch Keller

unread,
Jan 13, 2020, 2:01:28 PM1/13/20
to pretext...@googlegroups.com
David’s suggestion is a good one, but I will throw in a vote for alt text allowing/encouraging <m> tags in hopes that a screen reader can dig into them and process them, but that’s probably a question for the MathJax folks.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pretext-support/E98A6E43-D17C-478D-87EB-B9AA14853EAD%40gmail.com.


Mitch Keller
mi...@rellek.net

Alex Jordan

unread,
Jan 13, 2020, 2:52:19 PM1/13/20
to pretext...@googlegroups.com
Right now an image description goes into the alt attribute of the img in HTML.
The content of the alt attribute should be plain text. For example, I'm not aware
that MathJax has some way to look into an alt attribute of an image and do
anything with it. (Also it feels like an HTML attribute abuse to use something
other than text, but maybe that is just me feeling and not relevant.)

So as long as we are talking about providing an image description to the user
via the alt attribute (that is, doing things in the standard way) then for practical
reasons I would not let math content into a description.

An alternative to the standard way of doing things would be to not use "img"
for the image, and to use something like "object" instead. The child content
of the "object" is supposed to be the fallback content when the "object"'s
@data isn't available. But I think this alternative would be asking for trouble
with different browsers and screen readers not ready to use "object" the way
it was envisioned to be used.




Rob Beezer

unread,
Jan 13, 2020, 3:03:30 PM1/13/20
to pretext...@googlegroups.com
The alt *attribute* in HTML is an *attribute*. So you are severely limited in
(a) what you can put there, and (b) what you can do with it.

For example, if you put a quote mark in the PTX "description", then the HTML
will be wrecked. So the PTX expectation is that "description" carries no
elements at all and will go into @alt verbatim.

An extreme suggestion would be to allow "m" tags and run MathJax Speech Rule
Engine over it on the side and somehow put it back in place in the output.

A more practical answer might to imagine how the six objects of that diagram
would be verbalized, and then say things like "A three row, two column layout.
With arrows going... With row one, column one being E bracket x mod..."
> =========
> Thomas W. Judson, Professor
> Department of Mathematics and Statistics
> Stephen F. Austin State University
> P.O. Box 13040-3040 SFA Station
> Nacogdoches, TX 75962
>
> OFFICE: 316 Math
> TEL: (936) 468-1704
> EMAIL: juds...@sfasu.edu
>
> The views and opinions expressed in this message are my own and do not necessarily reflect the views and opinions of Stephen F. Austin State University, its Board of Regents, or the State of Texas.
>
>
>
>
>
Reply all
Reply to author
Forward
0 new messages