#readingsystems — Public Fediverse posts

Also in English:

There is a problem in accessible math publishing, which could be solved by reading system developers (and advocacy work).

I held a presentation on accessible math authoring and reading in Daisy consortium's techincal meeting in June. The Daisy consortium collects and distributes knowledge on accesible publishing all around the world.

Accessible math is in good standing! MathML is part of the HTML standard and many browsers support it. A screen reader can access the MathML code and relay the math equation as braille or it can be read aloud. For example, math equations authored with Wikipedia's math editor are accessible, since they are use MathML.

The situation is the same in accessible publishing (for example EPUB ebooks) - unless you are a sighted person!
In actuality my presentation was about how the situation is sort of upside-down (to the usual way of things): Usually it is the screen reader users in particular who face inaccessible content. Now the different reading systems supporting EPUB don't support MathML visually as well. Depending on the reading system, the equation can be barely understandable, while in others the reading experience is a lot better.

Firstly, this is a problem, because you can't talk about math! If the equation looks different to the other person than in braille for the other, you are at a dead end.

Secondly, publishers, who would publish mathematical content, won't do it if the result doesn't look correct or good. This could reduce the use of MathML, which should be instead increased, because of its accessibility.

It was important to bring this up, because many had not realised the situation! The assumption was that, because the browsers now support MathML, the same is true for the reading systems. This is not the case. Reading systems might be updated later than the browsers and they could also be developed from a version of the reading system, which is different than the one used in the browser. Math is also not first in mind when it comes to reading system development.

Most important of all: There was an assumption that it should work. I finished my presentation by saying that generalisations and assumption rarely hold up, and a lot has to be tried and tested. Luckily, many reading system developer are more than happy to get bug reports, which I should get on now!

#mathematics #MathML #accessibility #DigitalPublishing #math #ReadingSystems

Also in English:

There is a problem in accessible math publishing, which could be solved by reading system developers (and advocacy work).

I held a presentation on accessible math authoring and reading in Daisy consortium's techincal meeting in June. The Daisy consortium collects and distributes knowledge on accesible publishing all around the world.

Accessible math is in good standing! MathML is part of the HTML standard and many browsers support it. A screen reader can access the MathML code and relay the math equation as braille or it can be read aloud. For example, math equations authored with Wikipedia's math editor are accessible, since they are use MathML.

The situation is the same in accessible publishing (for example EPUB ebooks) - unless you are a sighted person!
In actuality my presentation was about how the situation is sort of upside-down (to the usual way of things): Usually it is the screen reader users in particular who face inaccessible content. Now the different reading systems supporting EPUB don't support MathML visually as well. Depending on the reading system, the equation can be barely understandable, while in others the reading experience is a lot better.

Firstly, this is a problem, because you can't talk about math! If the equation looks different to the other person than in braille for the other, you are at a dead end.

Secondly, publishers, who would publish mathematical content, won't do it if the result doesn't look correct or good. This could reduce the use of MathML, which should be instead increased, because of its accessibility.

It was important to bring this up, because many had not realised the situation! The assumption was that, because the browsers now support MathML, the same is true for the reading systems. This is not the case. Reading systems might be updated later than the browsers and they could also be developed from a version of the reading system, which is different than the one used in the browser. Math is also not first in mind when it comes to reading system development.

Most important of all: There was an assumption that it should work. I finished my presentation by saying that generalisations and assumption rarely hold up, and a lot has to be tried and tested. Luckily, many reading system developer are more than happy to get bug reports, which I should get on now!

#mathematics #MathML #accessibility #DigitalPublishing #math #ReadingSystems