[nabs-l] [Acbs-l] using Braille for math

[Bill Dengler]

Thanks for the comprehensive reply!
I’m interested in being a Computer Science major as well, and have seen your name in Github commits somewhere.
Math ML and such are incredibly frustrating. VoiceOver on the Mac jumps all over the place when reading Wikipedia’s and Stack Overflow’s implementation, so I’ve switched them both to output plain TeX.
You say that "UEB is extremely subpar for math.  Less symbols means reading faster.  Less symbols means keeping your equation on one line."
Yes, I understand that... but are there any benefits for me to learn and use UEB? As you say "your braille portion is for you, and you're the only person who needs to read it." so writing in a standard notation is obviously not important (as long as I can transcribe back to TeX!)
I am working to get a Baum Vario Ultra 40, which has a notepad and the ability to disable Braille translation. I plan to use the notepad (with translation disabled) for math, in combination with hardcopy Braille.
I have seen discussion that Braille math presents things spatially, which aids in understanding. With a display, you obviously only have one line at a time, so isn’t the spatial portion lost? Clearly you can only get that with hardcopy Braille... but as you said a Perkins would not work in a class setting! As you know, my only experience with Braille math has been the Math Window, which was a total disaster. As soon as I got to Algebra, I started writing in calculator notation on the computer, eventually moving to (La)TeX in pre-calculus.

Thanks,
Bill
> 
> On Sep 1, 2016, at 5:51 PM, Austin Hicks <camlorn at camlorn.net> wrote:
> 
> Hi,
> 
>     I've left you both on this as Christopher was interested in my response as well.
> 
> 
>     I'm Austin Hicks, and you may (or may not) know me as Camlorn.  I have my BS in computer science.  About 4 years ago, I decided that the problem I wanted to solve was 3D audio.  I started by writing OpenAL bindings, but that failed.  Not to be deterred, I taught myself digital signal processing from some online books that have LaTeX in the alt text of all the images, then used my newfound knowledge to write a now-mostly-finished replacement for the old DirectSound stuff but also cross platform and somewhat more powerful.
> 
>     If this all sounds like gibberish, you can just read it as I know math.  I want to know more math--I consider my knowledge very insufficient.  But of the blind people I know, I'm ahead of all of them by a great deal.  Which explains why Christopher forwarded this to me.
> 
>     I apologize in advance for the length.  This is pretty much everything I know about doing math as a blind person.  Feel free to send additional questions.
> 
>     Anyway, to take your concerns one at a time:
> 
>     LaTeX is absolutely what you should be using.  You may get some mileage out of a product called Mathtype, an add-on for word that allows you to type TeX into the document.  If you couple this with NVDA and Mathplayer, you can proof your work.  That said, the only reason I ever used it is that I never bothered to learn LaTeX fully, and only know TeX offhand (TeX is the math parts only, used on places like Wikipedia).  LaTeX compilers should give you errors, so proofing is less of an issue there (Mathtype doesn't, it just guesses, which is why you need Mathplayer).
> 
>     LaTeX puts you in a very, very good place for college.  You can get it out of Wikipedia and anything else they run (i.e. Wikibooks), for one.  Until recently, this used to be the default, but they changed it to MathML, so now you need an account.  Arxiv is a source of academic papers that you can access  for free and legally.  Many authors post LaTeX versions of their papers, to the point that there is at least one group who are trying to use it to build new tools for OCRing math to LaTeX.  But the biggest selling point is that, in college, you're going to have tests and quizzes.  And, since they're using it anyway for their research, many professors make their tests with it.  Lastly, many pages using Mathjax (like Stackoverflow, for example) can have Mathjax set up to show TeX source.  I get to take credit for that one,  though I didn't do the coding.
> 
>     I've mentioned Mathplayer and MathML.  These are the new "solutions" for this kind of thing. They're a waste of time, in my opinion.  By all means try Mathplayer, but everyone I know who is post algebra 2 hates it and goes back to LaTeX almost immediately.  The people working on it are primarily funded to get it working for the typical U.S. K-12 student, who you are already beyond.  In addition, you already know more math than James Teh, the head developer of NVDA.  If you end up trying it and liking it, then you'll be the first at your level of math that I've met, and really I need to figure out where we send feedback that might make them get that they're leaving people like me behind.  The only use I ever found for it was dealing with the fact that I wasn't using a proper LaTeX compiler, but you should try it because for all I know you'll love it.  Some people do, just, usually they're barely doing algebra.  The one good thing you can potentially get out of it is that it can get Nemeth going to a braille display, if you have one.  But the speech is worse than LaTeX.
> 
>     I have seen the math window.  It is a joke for anyone beyond basic addition.  Run and run far.
> 
>     Instead, what you really want is a Perkins Brailler and a bunch of paper for it.  The web site appears to be this:http://www.perkinsproducts.org/store/en/braillers/210-classic-perkins-brailler.html?gclid=CPPulpLZ7s4CFdBZhgodHNQB2Q <http://www.perkinsproducts.org/store/en/braillers/210-classic-perkins-brailler.html?gclid=CPPulpLZ7s4CFdBZhgodHNQB2Q>
> 
>     55 words a minute braille reading speed is typical, and about what I have.  You may be faster, and you're possibly the fastest I've personally heard of.  Braille is just intrinsically that slow.  But for math, nothing beats the perkins brailler.  You can type on it faster than a sighted person can write with a pencil with some practice, and you can get almost 40 characters a line if you buy the larger paper (you can get paper for it via office depot, btw, just buy cardstock. Don't order directly from them till you're sure they're a good deal).  It's like a braille type writer, and you can go as fast as the average sighted person types.  This isn't exactly portable, nor is it quiet.
> 
>     How I did math in high school was someone translated Nemeth for me.  How I did math in college was mostly LaTeX plus the perkins brailler for anything that needed to be broken up if I did it via speech.  To do this reasonably, you want to do as much on the computer as you feel comfortable with.  Then, use the brailler to write down anything you need to access quickly.  Finding something on the braille page is way, way faster than trying to find it in the text document, and comes with the additional benefit that you keep your place in the text document.  For things like matrices, the brailler is the only convenient way to work with them because suddenly 2D navigation is important.  When the expressions get too complicated, you can switch fully and then transcribe back.  This is annoying, but at least it's not for every problem.
> 
>     As for the code, you want to start by learning Nemeth through algebra 2.  This is something like 20 or 30 symbols only.  UEB might eventually take over, but it won't be while you're still in school.  In addition, the NFB or someone lobbied and I believe got Nemeth into the specification, so the U.S. probably won't switch off it ever.  UEB is extremely subpar for math.  Less symbols means reading faster.  Less symbols means keeping your equation on one line.  Even if we all switched to UEB tomorrow, it's going to be another 15 or 20 years before you can't get Nemeth: we can't just drop all the students that grew up learning it.
> 
>     Then, you're going to butcher it.  This will probably not be intentional.  But your braille portion is for you, and you're the only person who needs to read it.  You can leave off some symbols while still being unambiguous, for example x_2 can be done improperly as x2, and you just know from context that that's never the same as 2x.  Whenever you go to college and no one can teach you new symbols, you invent your own.  There are official things for calculus that are more efficient than mine, but no one could teach me; as a consequence, I use lim(x, from, to, expression) for limits, int(a, b, expression)dx for integrals, etc.
> 
>     The only place I ever used a notetaker in college (the hardware) was math classes.  Because the perkins brailler is not portable or quiet, I needed an alternative.  To my knowledge, no notetaker understands Nemeth.  But they have 3 features that can combine to let you fake it 100%: the speech can be turned off, they can be configured to write brf (braille format) files, and you can configure them to let you input braille directly.  If you do these three things, you can do Nemeth with the refreshable braille display.  If you go this route, push ahrd for a 40-cell display; even with a 40-cell, things still wrap, and wrapping is very expensive when you only have one line of the page at a time.  As a bonus, it'll even print from the braille printer correctly.  As an extra bonus on top of that, someone sighted can read it with a little practice, should you send the file through the print printer (certain things like + and parens happen to match up. A convenient coincidence, but I don't think intentional).
> 
>     Should you keep going, you eventually reach where I am.  Computers can do all the gruntwork.  Integrate this expression, sure. Solve that equation, yep.  Do this linear algebra thing, no problem.  Want to work with alternate algebras? Sure!  There are actually accessible solutions for these things, but you're not at the point where using them is easier than doing the math, nor will you be allowed to.  But for things like actual research projects, people don't have to do a lot of it by hand.  Now that I'm out of college, I don't pull out the brailler much.  It's much easier to go ask something like Sympy to do this calculus problem, or to go find out how someone else solved it.
> 
> On 8/30/2016 12:22, Christopher Kchao wrote:
>> Care to field this inquiry? Obviously, no pressure - I just thought you'd be optimally equipped to address some or all of his concerns.