we have chosen to put most of our research into documents in PDF format.
PDFs are a huge pain to make accessible.
most scientists write their papers in Latex, overleaf, etc., which cannot produce accessible PDFs.
to make such PDFs accessible, one uses Adobe Acrobat, which is expensive and proprietary.
increasingly, we post our PDFs to arXiv, which ~forbids accessible PDFs b/c they can't be compiled from source.
~none of our science is accessible.
artifacts (and file formats) have politics.
@jbigham It seems LaTeX is in the process of implementing this: https://www.pdfa.org/presentation/tagged-and-accessible-pdf-with-latex/
But put me in the "PDFs are not worth the hassle" camp anyway. I know some people in the LaTeX project have been working on a new document container format that still offers precise printing, but that can also reflow text for different screen sizes. The name escapes my memory right now. Maybe it'll be really cool. But for current research publication, I'd settle for HTML.
@jbigham @simulo That's kind of chicken-and-egg, though, isn't it? If someone _wanted_ to output compliant, accessible HTML there are already standards for this: https://w3c.github.io/scholarly-html/
You can be prescriptive without requiring a proprietary file format...
@simulo @profgaelle @jbigham I’m no scientist, but I am a (copy) writer, so I have worked alongside designers for twenty odd years. So believe me when I say this: Design makes a huge difference. The impact on things like perception and readability cannot be over stated. And more often than not it affects content, too. (“Ok, so this info is in the diagram, we can skip the text” or “let’s high light that and skip the diagram” etc.)
Design and typography matters.
@thelovebing @simulo @profgaelle @jbigham
Tech editor here who has also, separately, helped a few nonprofit orgs through the epub process.
I very much agree with your take on this.
@simulo @jbigham
I think it is in great part a desire to make it exactly correspond to what is printed in the paper journal, for better or worse.
My read of the invention of http in particular was that it was meant to help information management needed to run large physics experiements-- many thousand pages of internal documentation, and would probably not have been invented that early to keep track of the very distilled final product
(https://www.w3.org/History/1989/proposal.html speaks of cern user groups etc)
@cbischoff @bbolker @jbigham It is a small team, and I believe they are all volunteers.
The majority of the commits are from one person: https://github.com/latex3/latex3
@jbigham agreed on the thesis, but calling bullshit on "most scientists write their papers in Latex"
I would be hard pressed to believe that most CS or physicists do, much less scientists more broadly.
@salkinium Oh interesting, though got a 404 when I tried the link.
A former coworker wrote an inhouse tool to pull out table data from TRMs, and I remember him complaining about all the edge cases he encountered, even though the tool only had to support documents from our main SoC supplier.
@jbigham hmm, in which way is LaTeX PDF output non-accessible? At least it doesn't produce rasters like, well, Photoshop & other things, so that must be some other problem…
Do you mean text blocks ordering or some other stuff?
@jbigham https://lmgtfy.app/?q=latex+pdf+tags
Don't know how extensive they are but there are ways to do it it seems. Sure they need more exposure, and push for arXiv to consider making it mandatory.
@jbigham just like surely not everyone using Acrobat know or will use the accessibility tags properly anyway.
People still want to fake radio buttons in HTML, so there's a long road ahead…
@jbigham makes me wonder how well LibreOffice handles that in PDF it generates, as I've been using it for years thinking it was just fine.
Same for pandoc I use to generate user manual through LaTeX from markdown… 🤔
@jbigham I now use markdown (specifically quarto) to render work into multiple formats from the same source file.
As you can keep the .tex file while rendering to pdf, for submissions where submitting tex files are that still de rigueur is not a barrier.
But I can also, with the same source document, produce a fully accessible document (even with equations), with structure and metadata.
Making it easier to host files in a more accessible format on arXiv and similar is key
@jbigham I understand this frustration. Sometimes I use Linux for some workaround (including using something that will let me create, markup, and edit PDFs)
I hate the hold that Adobe has on this kind of stuff.
@jbigham I'm curious about how hard it would be to teach screenreaders to read TeX, or perhaps make a tool that takes in TeX and spits out some semantic tagged stream. (Or maybe this is just the TeX -> HTML that already exists.)
And I feel like it should be possible to include TeX in PDFs as an embed of some sort... I know it's a hack, but I just wonder if it's a faster road to accessibility.
@jbigham @freakboy3742 I work in this domain and PDF display and management is a non-trivial exercise.
PDFs held hostage by predatory publishers is a whole other extra barrel of fun.
Alan Arnold
Jeffrey P. Bigham 🔥🔥
Julian Fietkau
Counting Is Hard
quelu 🦔🦉👺
geolaw
Jan D
Scott Jackson
dingodog
Sabrina/Brie
Prof Gaelle Vallee-Tourangeau
Världens bästa Kille™
Virginia Murr
jonny
Knut Morå
Adam Shostack 
Ben Bolker
Colin Bischoff
Tom Ritchford
Sukrit Venkatagiri
cryo2go :unverified:
ahistorical immaterialist
Kristin Branson
Martin Hamilton
Niklas Hauser
Maya
mmu_man
何事西风 
不待人
Tom Renner
Dee Toher
Andrea Bianchi
Nontechie Talk 
Swarmpicker
Rep. Eric Gallager (no "h"!)
Kat the Leopardess
varx/tech
Iain Anderson
Andy Scholand
huesm
pizzapal
Andrew Harvey
Jonathan Dowland