[BlindMath] Zoom meeting on accessible math: Thu 11 Feb 6:30 to 7:30 UK time

Sun Feb 7 01:57:31 UTC 2021

Hello Peter,

If only it was so easy. It is often very difficult to get source files out of people.

Surely LaTeX is only a solution to accessibility  for those people who speak LaTeX. (?)

It isn't the first 95% of a pdf that is the problem. Even the least effective current OCR solution can deliver the text (prose) but there are so many things that reading the source cannot deliver. I can read raw LaTeX but I don't wish it on others given how easily we should be making accessible documents in other ways.

For example, Is it reasonable to expect a high school student to be able to untangle the LaTeX that would generate something like the formula used to solve a quadratic equation?
That same expression is easily read by today's screen readers if it is presented in MathML or MathJax via an HTML document.

So many LaTeX codes are strange until we learn how they came to be what they are, and that's before an author goes and invents their own!

One thing I liked about the TeX4HT solution was that it did allow an author to invent their own LaTeX codes and have them reduced back to the proper (raw) commands for proper presentation in MathJax. As a developer of content, I used my own extra commands extensively. "\BMY" has zero meaning to anyone who hasn't read the complete list of nearly 250 symbols I made for my documents, all in the name of saving myself time as I created content.
I've just been cleaning up LaTeX source files over summer for use in HTML via markdown. I've had to write scripts to find and replace the vast number of codes my former colleague used because markdown won't manage non-standard LaTeX. BTW: \BMY just means the capital letter Y written in the bold math font, meaning it is a matrix (usually a column vector actually).
I also used many shortcuts in the regular paragraphs too. Use of the codes meant that the visual presentation of a term was always the same throughout a 300 page thesis. I wrote commands to replicate sets of other commands so that I saved myself time; they just don't make sense to anyone else and let's face it, most have zero value to the blind person reading them.

Does a blind person really care if it is \varepsilon or \epsilon? They should, but not because the symbol looks slightly different to the sighted audience. FYI: anything read by MathJax should use \varepsilon because it thinks of this as epsilon whereas \epsilon leads to utter silence.

There have been several initiatives to clean up LaTeX and make it more human readable as well as conversion to Nemeth, and filtering for improved access using the JAWS screen reader. None has attracted widespread use though.

IMO: Until a blind reader can engage with the equations written LaTeX on par with their sighted peers, they cannot claim to have access. Even the HTML based solutions leave us a little to a lot behind our sighted peers.

Jonathan G

-----Original Message-----
From: BlindMath <blindmath-bounces at nfbnet.org> On Behalf Of prayner--- via BlindMath
Sent: Sunday, 7 February 2021 2:18 PM
To: Blind Math list for those interested in mathematics <blindmath at nfbnet.org>
Cc: prayner at unimelb.edu.au
Subject: Re: [BlindMath] Zoom meeting on accessible math: Thu 11 Feb 6:30 to 7:30 UK time

given that the LaTeX source itself is highly accessible I wonder if the simplest solution to this is a package that includes the LaTeX source as an external media type in the generated pdf? I've not looked much at pdf and included content beyond what's possible with sounds but this seems an option which could be implemented without climbing inside all the packages. Anyone looked at this?
regards
Peter
Tony Malykh via BlindMath writes:
>I did my own little investigation into this a while ago. Here is my 
>understanding. I'll be happy if anyone corrects me if any of my points 
>are not valid.
>1. Tex was invented like in  1960s - and they had no idea about 
>accessibility back then. Tex is really just the engine, and there are 
>tons of packages written on top of that engine. Neither the engine, nor 
>the packages were written with accessibility in mind.
>2. There appear to have been some efforts to introduce Tex PDF 
>accessibility. For example, see:
>     *
>https://tex.stackexchange.com/questions/550511/what-is-the-best-package
>-for-accessibility-tagging/550523#550523
>     * https://github.com/karel-brinda/tex-accessibility
>     * https://github.com/integr-abile/axessibility
>However, these packages seem to suffer from one common problem: they 
>break visual layout  except for the simplest cases. That's the reason 
>why conferences don't use this package.
>3. In order to have a working accessibility Latex package, it appears 
>that you need to make changes to nearly every otehr Latex package to 
>make it to work with accessibility package. The number of tex packages 
>that are widely used is enormous, therefore therethere's enormous 
>amount of work to be done with no one to pay for, and let's be honest - 
>with marginal benefits - there are not too many blind people in the 
>world willing to read math papers.
>4. I also stumbled upon the discussion with one of ML conference 
>organizers (couldn't find a link now), and it seems like they were 
>willing to work in that direction, that is enforce all the papers to be 
>accessible, as long as there is a working accessibility package.
>So my conclusion was that Latex PDF accessibility is not likely to be a 
>reality in the nearby future. Even if NFB or any other 
>blindness-related group forms an advocacy group to push for accessible 
>Latex PDFs, it still feels unlikely  that widespread accessibility 
>could be achieved, primarily becuase of complexity of the system.
>Too bad. Again, hope someone can correct me.
>--Tony
>
>On 2/6/2021 7:10 AM, Petr Pařízek via BlindMath wrote:
>> Jonathan wrote:
>>
>>
>> <<<<<
>>
>> Many teachers use LaTeX for course materials. I'd like to have 
>> real-world
>>
>>> examples of LaTeX math, and accessibility problems and solutions. 
>>> Please send to this list or to me privately as you wish.
>>> >>>>>
>>
>>
>> Recently, someone has told me that out of the many existing TeX to 
>> PDF converters out there, there doesn't seem to be such one that 
>> would be able to label the math expressions with corresponding 
>> alternative text in the PDF document. Sadly, I'm unable to verify 
>> whether this is indeed the case or whether labeling graphics with 
>> alternative text is somehow problematic in PDF documents. If 
>> alternative text is no problem in formats like HTML, then I have no 
>> idea why it should be a problem in PDF. And if it is not a question 
>> of issues with alternative text itself, then I wonder why the 
>> converters can't use it for labeling the math expressions properly 
>> (which are actually stored as graphics). Honestly, I'd love to know 
>> myself what the answer to the question is but I don't know anyone who 
>> knows enough both about the actual PDF format and about the available 
>> TeX to PDF converters and who could indeed give me the answer.
>>
>> Petr
>>
>>
>>
>
>_______________________________________________
>BlindMath mailing list
>BlindMath at nfbnet.org
>http://nfbnet.org/mailman/listinfo/blindmath_nfbnet.org
>To unsubscribe, change your list options or get your account info for BlindMath:
>http://nfbnet.org/mailman/options/blindmath_nfbnet.org/prayner@unimelb.
>edu.au BlindMath Gems can be found at 
><http://www.blindscience.org/blindmath-gems-home>

--
Peter Rayner
Academic Lead - Climate & Energy College <http://www.climatecollege.unimelb.edu.au>
Clean Air and Urban Landscapes NESP hub <http://www.nespurban.edu.au> School of Earth Sciences, University of Melbourne, 3010, Vic, Australia mobile +61 402 752 379, skype: petermorag zoom id 4431343191, join at  <https://unimelb.zoom.us/j/4431343191?pwd=a1E5Z3JEOTRVQUJsaVdRbVUvR1QyZz09>
mail-to: prayner at unimelb.edu.au TWITTER: @raynerstrings

_______________________________________________
BlindMath mailing list
BlindMath at nfbnet.org
http://nfbnet.org/mailman/listinfo/blindmath_nfbnet.org
To unsubscribe, change your list options or get your account info for BlindMath:
http://nfbnet.org/mailman/options/blindmath_nfbnet.org/a.j.godfrey%40massey.ac.nz
BlindMath Gems can be found at <http://www.blindscience.org/blindmath-gems-home>