[accessibility] Making PDFs produced from Latex accessible - part1
Ross Moore
ross.moore at mq.edu.au
Fri Dec 9 02:00:45 CET 2016
This is part 1 of an email discussion from several weeks back.
All involved agreed to have it posted to this list.
Begin forwarded message:
From: Boris Veytsman <borisv at lk.net<mailto:borisv at lk.net>>
Date: 31 October 2016 5:02:58 AM AEDT
To: Ross Moore <ross.moore at mq.edu.au<mailto:ross.moore at mq.edu.au>>
Cc: <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>>, <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>>, <karl at freefriends.org<mailto:karl at freefriends.org>>
Subject: Re: Making PDFs produced from Latex accessible
Hi, Ross
Thanks for your very interesting letter. I am glad to hear you are
working on this. Let me know if you need any help in testing,
development or funding.
I would like to discuss another use case, which is relevant to my
work. Consider a LaTeX class author. Suppose she wants to make a
class that produces accessible PDFs. As a document class author, she
has much better control than the document author: for example, she can
put any tag code in the headers and footers, can plug into output
routine, can expressly forbid loading packages incompatible with the
accessibility code, etc. Thus an accessibility package with the
interface for class authors might be simpler to make than an
accessibility package with the interface for document authors.
I would suggest releasing such package first, so document class
authors could use and test the code. What do you think?
--
Good luck
-Boris
Numeric stability is probably not all that important when you're guessing.
Begin forwarded message:
Date: 30 October 2016 3:08:40 PM AEDT
To: Volker Sorge <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>>
Cc: Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>>, Boris Veytsman <borisv at lk.net<mailto:borisv at lk.net>>, "Karl Berry" <karl at freefriends.org<mailto:karl at freefriends.org>>
Subject: Re: Making PDFs produced from Latex accessible
Hi Volker, Jeff, Boris, Karl
On 28/10/2016, at 6:40 AM, Volker Sorge wrote:
Hi Ross,
Thanks for getting back to us. I agree with most of your points. In particular, it will be difficult to get authors to voluntarily do extra work to be able to generate accessible or even semantically rich PDF documents in all cases. But this is not what we want to do anyway.
I'm adding in Boris and Karl to this conversation.
Both should be interested in what you have to say,
and to the way I'm responding to it.
Instead we want to concentrate on the most important use case, which I would crudely define as follows:
Author X has to submit an accessible paper to conference Y.
- X writes paper in LaTeX
- generates a PDF file for the CRC.
- moves PDF to Adobe Acrobat X Pro (possibly change platform, e.g. from Linux to Mac/Windows, have to buy Acrobat or use the limited trial version).
- run accessibility audit and manually take care of all the problems such as reading order, header nesting, alt text tags etc.
- save paper and sends it to publisher for Y.
- AND MOST IMPORTANTLY: Never touch the PDF again!
There are two major problems with this workflow:
1) If X wants to correct a mistake in the paper they have to go through the entire process again, which is very time consuming.
2) If there is a problem with paper publisher of Y will not touch/rerun the LaTeX, while they can generate a basic accessible PDF from a word doc directly. This might eventually force Y to only allow Word submissions.
Yes, OK.
This should spur an urgency within the TeX world, to fully implement
tagged PDF. But it has not done so.
I claim author X faced with this extra work will certainly be prepared to include a bespoke accessibility package and put enough additional information into the paper to ease the above workflow.
So having a package that would already take care of the most common accessibility issues in PDF (and is at least as good as what MS word can produce) would be very helpful. Some issues I would identify off the top of my head are:
- Automatically generate the correct reading order (at least for sections/sub-sections etc.)
- Generate the correct nesting of headers
- Fill in the title, author information.
- Allow to manually create some reading order exceptions (e.g. for thank you or copyright notes)
- Allow to specify alt-text for graphics, formulas, etc.
- [Optional] Provide a way to automatically fill in Alt Text for maths formulas adhering to Maths accessibility conventions.
This all sounds simple enough, in principle only.
In practice it is not that easy.
I'll attempt to explain why.
My 2009 paper was meant to indicate the extra technical difficulty
and what extra information needs to be included within a tagged PDF
document, over and above what is present in the paged view
that TeX software currently produces.
Now for a document to be assessed, in some automatic way, as being
accessible, it will need to pass validation tests.
These tests might be done using Acrobat Pro, or other software
validation tools; e.g. as produced by Callas software
https://www.callassoftware.com/en
This will require proper adherence to the published PDF/UA standard,
and correct implementation of the PDF tagging structures.
(Essentially this means adherence the forth-coming PDF 2.0.)
Here I'll discuss just some of these requirements.
Requirement #1:
*all* content on the page must be tagged, whether included
via the structure tree or not.
The "or not" here refers to "Artifacts", such as headers, footers,
page numbering, background images and other decorative features
that may be on the page.
Things like horizontal rules don't count, since this does not use
characters from fonts; so fraction bars normally don't count.
Arrows and arrowheads *can* count as content, if placed using
font characters, as with Xy-pic --- another of my projects.
Multiple Dingbats, used to create decorative borders using font
characters, must also be tagged as "Artifact".
The upshot of this is that every package that is used to include
specialised content, must be checked for compatibility with tagging.
It may be that the package simply provides an easy front-end to
internal LaTeX commands, in which case the tagging can be done
by patching the LaTeX commands. Or it may be that the package
calls TeX primitives directly, in which case it is the package
that should also handle the tagging.
Sometimes it may be OK to patch at the level of the TeX primitives;
but this could be problematic. There are different types of Artifact,
and the semantic information needed to determine this may no longer
be available at the level of the particular primitives.
As an example, consider how (La)TeX identifies a paragraph,
for applying the line-breaking algorithm.
Usually just a blank line at the end of text calls up \par .
The start can be triggered simply by encountering text characters,
which internally trigger a macro called \everypar .
These macros \par and \everypar now need to be *overloaded* to handle
tagging, both of structure and content. This is as well as whatever
they would normally do within LaTeX, or within the specialised content
of particular environments defined by packages.
Thus our *package* for tagging must identify all the situations in which
\par and \everypar are used, and make appropriate adjustments.
A quick grep search through the tex-live ../texmf/tex/latex/ tree
reveals 553 instances of \everypar usage within package .sty files,
and the like. Only 18 of these are within the base LaTeX .ltx files.
And of course a user might specify \everypar{} at any stage within
a LaTeX document. Normally this would be quite innocuous; but if \everypar
has been overloaded to handle tagging, then this will be destroyed,
at least within a level of grouping by {...} or within an environment.
Yes, this is likely to be rare; but certainly possible — and it is precisely
what is done within many package-defined macros.
Another special case occurs when detailed MathML tagging is used
for math content, with the use of parentheses and other delimiters.
Depending on just how the MathML stream handles the structure,
the characters to display the delimiters could be viewed as content,
e.g.
<mrow><mo>(</mo><mi>a</mi><mo>+</mo><mi>b</mi><mo>)</mo></mrow>
or as artifacts
e.g. <mfenced><mi>a</mi><mo>+</mo><mi>b</mi></mfenced>
Spacing within mathematical content also has big issues.
This kind of detailed level of mathematics tagging is pretty rare.
So far it may be only myself who can actually do this, inside a PDF.
But it is exactly what Neil Soiffer asked me to do, so that there
is a way to provide fully tagged mathematics for his MathPlayer
Acrobat plug-in.
https://www.dessci.com/en/products/mathplayer/
This is ideally the way mathematical content be presented within PDF files.
But think of the consequences of having such tagging.
The variables 'a' and 'b' should be mapped to upper-plane Unicode
characters 𝑎 and 𝑏 and any greek letters or operator symbols in TeX fonts
or AMS fonts mapped similarly into the proper Unicode code-points.
/ToUnicode maps can do this. But for accessibility, you want more;
since these symbols need to be read aloud in a sensible way.
e.g. you don't want "mathematical small italic b" to be vocalised;
"variable named b" or "variable b" are preferable, or even just "b".
Think further about calligraphic, script or fraktur characters.
Then the best is to vocalise what the character is being used for,
rather than what it is.
(This could apply also to user-defined variable names.)
Requirement #2:
word boundaries must be identifiable to PDF readers/renderers.
Usually this means that space characters should appear in the textual
content of the PDF. TeX normally does not do this!!!
We've all seen how copy-paste from a LaTeX-produced PDF can capture
all the letters, but abutted together, without the inter-word spaces.
I say "normally" here, because there is now a "fake-space" option that
does allow the input to contain invisible zero-width space characters.
Now Copy/Paste works much better; but "Reflow" does not.
Reflow is about having the textual content reflow to fit the window size,
as the content is scaled up for better readability; e.g. by a person
with some (but not total) visual impairment.
Although not actually part of PDF/UA, the concept of Reflow is clearly
meant as an aide to Accessibility. So it's a desirable feature for a PDF.
And it is listed here:
http://www.adobe.com/content/dam/Adobe/en/accessibility/products/acrobat/pdfs/acrobat-x-pdf-accessibility-overview.pdf
If TeX-produced documents are ever going to allow Reflow, then there are
fundamental changes needed to the way the text-strings are written into
the PDF file. Not doing this is a real game-breaker for TeX.
Requirement #3:
Each page needs to be able to "stand-alone" as a tagged PDF document.
Think about extracting a page from a document, and possibly copying,
along with other single pages, into another tagged document.
This creates a big problem for tagging a (La)TeX-produced document,
if all the tagging was to be done entirely by macro-expansion,
using commands from a package.
Where does a page end? Usually somewhere in the middle of a paragraph.
But this isn't determined until the output routine, after all macro-expansion
has occurred, and the vertical list is being examined to determine suitable
break-points.
Not only do we need end-points for the tagging of content on the page
being output, but also we need to start the tagging, of both structure
and content, of the subsequent page in a way that is consistent with
the structure tagging of the document overall.
This handling of tagging at page-boundaries is best done internally
by TeX, rather than at the level of macros from a package.
It is possible for something like ConTeXt, and perhaps also by LuaTeX,
using Lua scripting to adjust the PDF output.
For pdfTeX or dvips, it is currently not possible — without extra
internal programming structures, induced by extra primitive commands.
My 2009 paper, and subsequent examples, used a version of pdfTeX
extended with such primitives.
I intend creating more examples using this extension of pdfTeX.
The problem is that one has to use these extra primitives in exactly
the correct way, else one can easily create broken PDFs, without
any feedback from the TeX processor to warn of the broken document.
Indeed, if the TeX previewer is one that is unaware of tagging
(e.g., Apple's Preview) then you may not have any inkling that there
is a problem, until viewing with Adobe's Reader or Acrobat Pro.
Then the error can be due to malformed PDF, which may cause the
software to refuse to open the file, or even crash completely.
Or it can be a problem in the structure tree.
In either case there is no easy way for anyone except the document
author to determine what might be going wrong, or what was really
intended. And even then, the author is unlikely to be aware of
what they did wrong.
This is the main reason why I still do not release the work that I've
done so far on tagging, to be used (or rather abused) by others.
It is far too easy to make mistakes, for reasons that only an experienced
TeX guru would understand. Most users expect more, and cannot cope when they
don't get what they think they asked for --- when in reality they did not.
This would just lead to a greater flow of users away from TeX and LaTeX;
which is precisely the opposite of what we want.
Visit the page;
http://rutherglen.science.mq.edu.au/~maths/CICM/
Download my paper:
http://rutherglen.science.mq.edu.au/~maths/CICM/mooreDML2014-postprint.pdf
Look at the attachments to this PDF.
You should find both a fully-tagged PDF titled:
2013-Assign2-soln-3u.pdf (or similar)
and the .tex source used to generate it,
using the specially extended pdTeX .
Examine that source, to locate bits that are not standard LaTeX.
You should not be able to find much, since I've done a pretty good
job of adjusting macro-expansions within an include'd file.
(named tagstart.tex or similar)
This is an example of what you want; essentially including a package
to obtain good tagging. But remember that the processing engine has
been modified, and there is a TeX guru behind it all, taking care of
detailed aspects of the processing, then hiding it away in re-definitions
(and over-loadings) of LaTeX internal commands.
Proof of concept? — maybe.
Ready-to-use package? — certainly not.
I feel that a package of this nature should be do-able. I am not certain how much interaction with other packages would be necessary, but if necessary in a first iteration, we could restrict it to work with certain packages only (e.g., AMS math).
I've tried to explain above, why this is not as practical as it sounds.
Asking users to restrict the packages they use can really only result
in frustration, as people then try to recreate the effects that they
know they can get easily using some other package.
We should schedule a Skype call or Google hangout on this in the not too distant future.
Jeff is on the East Coast, I am still in the Bay Area until 21 Nov, which might make it easier to schedule a call. After that I'll be back in the UK.
Sure.
Time-zones make it hard to find a good time, suitable to everyone.
But that should not stop us trying. :-)
Best,
Volker
On 26 October 2016 at 16:10, Ross Moore <ross.moore at mq.edu.au<mailto:ross.moore at mq.edu.au>> wrote:
Hi Volker, Jeff
On 27 Oct 2016, at 03:10, Volker Sorge <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>> wrote:
Thanks, Jeff, for bringing me in to this conversation.
Hi Ross, you might recall we had a chat about this a couple of years
back in Washington. I agree with Jeff, that if we can't produce
accessible PDFs from LaTeX directly, it will only be a matter of time
before publishers force authors to move away from LaTeX. But I can
think of a number of organisations that this won't happen and
therefore would be interested in a bespoke LaTeX package for
accessibility (e.g., AMS, APS).
The problem is that it will require more than just a package.
Even with a good package and a revised LaTeX format, authors will
need to adjust there coding habits to be aware of the extra information
required to construct good, consistent tagging structures and many
commonly-used packages will need to be updated to be aware of tagging,
and what is required for proper tagging of structures like mathematical
content, tables and figures.
Ross, if you are interested in this, maybe we could have a face to
face one of these days.
I’m certainly interested.
I’ve been very busy with teaching commitments the past few years.
Hopefully soon I’ll get more time to devote to implementing
structures that are useful for Accessibility, and produce more
example documents. We have a new Head of Department, who is
aware of the need to support Accessibility, so I’m expecting to get
a bit more actual support at my university than I’ve had previously.
Best,
Volker
On 26 October 2016 at 06:12, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
As luck would have it, I'm attending a conference with Volker Sorge (CC'd)
who is also interested in this, so I'm adding him to the conversation.
Thanks,
Jeff
On Tue, Oct 25, 2016 at 9:25 AM, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
I ran across your paper here:
https://www.tug.org/TUGboat/tb30-2/tb95moore.pdf
That describes work that I presented back in 2009.
The best place to find more recent work is here:
https://www.tug.org/twg/accessibility/
Scroll down to the list of papers, with links, near the bottom.
We've been trying to make the conference papers we produce in computer
science accessible to people with disabilities. One of the challenges we
have is that so many of the publications are produced as PDFs and many
authors use Latex.
Do you have any updates on the state-of-the-art on producing
accessible/tagged PDFs directly from Latex? Do you know anyone who would
have the experience to create a Latex package for tagging PDFs? I could
contribute real resources to this project, but I would need to find someone
with the necessary expertise.
Yes, there are not many people with the proper expertise.
In the TeX User Group community, there are people such as:
Boris Veytsman and Bruno Le Floch
who have shown some interest in Accessibility.
Some of the CICM/MKM community are sort-of interested,
via semantic enrichment.
And every so often I get an email from a student somewhere,
wanting a quick and easy solution, to produce an accessible PDF
for their PhD thesis.
I’m in contact with the ISO/PDF working group that develops the PDF/UA
and other standards. This includes engineers from Adobe.
But these guys know of TeX/LaTeX, but don’t have expertise in programming
with it.
Thanks!
Jeff
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com/>
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com/>
Hope this helps, at least a bit.
Cheers
Ross
All the best,
Ross
Dr Ross Moore
Mathematics Dept | Level 2, S2.6 AHH
Macquarie University, NSW 2109, Australia
T: +61 2 9850 8955 | F: +61 2 9850 8114<tel:%2B61%202%209850%209695>
M:+61 407 288 255<tel:%2B61%20409%20125%20670> | E: ross.moore at mq.edu.au<mailto:rick.minter at mq.edu.au>
http://www.maths.mq.edu.au<http://mq.edu.au/>
[cid:image001.png at 01D030BE.D37A46F0]<http://mq.edu.au/>
CRICOS Provider Number 00002J. Think before you print.
Please consider the environment before printing this email.<http://mq.edu.au/>
This message is intended for the addressee named and may
contain confidential information. If you are not the intended
recipient, please delete it and notify the sender. Views expressed
in this message are those of the individual sender, and are not
necessarily the views of Macquarie University.<http://mq.edu.au/>
Begin forwarded message:
From: Volker Sorge <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>>
Date: 28 October 2016 6:40:21 AM AEDT
To: Ross Moore <ross.moore at mq.edu.au<mailto:ross.moore at mq.edu.au>>
Cc: Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>>
Subject: Re: Making PDFs produced from Latex accessible
Hi Ross,
Thanks for getting back to us. I agree with most of your points. In particular, it will be difficult to get authors to voluntarily do extra work to be able to generate accessible or even semantically rich PDF documents in all cases. But this is not what we want to do anyway.
Instead we want to concentrate on the most important use case, which I would crudely define as follows:
Author X has to submit an accessible paper to conference Y.
- X writes paper in LaTeX
- generates a PDF file for the CRC.
- moves PDF to Adobe Acrobat X Pro (possibly change platform, e.g. from Linux to Mac/Windows, have to buy Acrobat or use the limited trial version).
- run accessibility audit and manually take care of all the problems such as reading order, header nesting, alt text tags etc.
- save paper and sends it to publisher for Y.
- AND MOST IMPORTANTLY: Never touch the PDF again!
There are two major problems with this workflow:
1) If X wants to correct a mistake in the paper they have to go through the entire process again, which is very time consuming.
2) If there is a problem with paper publisher of Y will not touch/rerun the LaTeX, while they can generate a basic accessible PDF from a word doc directly. This might eventually force Y to only allow Word submissions.
I claim author X faced with this extra work will certainly be prepared to include a bespoke accessibility package and put enough additional information into the paper to ease the above workflow.
So having a package that would already take care of the most common accessibility issues in PDF (and is at least as good as what MS word can produce) would be very helpful. Some issues I would identify off the top of my head are:
- Automatically generate the correct reading order (at least for sections/sub-sections etc.)
- Generate the correct nesting of headers
- Fill in the title, author information.
- Allow to manually create some reading order exceptions (e.g. for thank you or copyright notes)
- Allow to specify alt-text for graphics, formulas, etc.
- [Optional] Provide a way to automatically fill in Alt Text for maths formulas adhering to Maths accessibility conventions.
I feel that a package of this nature should be do-able. I am not certain how much interaction with other packages would be necessary, but if necessary in a first iteration, we could restrict it to work with certain packages only (e.g., AMS math).
We should schedule a Skype call or Google hangout on this in the not too distant future.
Jeff is on the East Coast, I am still in the Bay Area until 21 Nov, which might make it easier to schedule a call. After that I'll be back in the UK.
Best,
Volker
On 26 October 2016 at 16:10, Ross Moore <ross.moore at mq.edu.au<mailto:ross.moore at mq.edu.au>> wrote:
Hi Volker, Jeff
On 27 Oct 2016, at 03:10, Volker Sorge <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>> wrote:
Thanks, Jeff, for bringing me in to this conversation.
Hi Ross, you might recall we had a chat about this a couple of years
back in Washington. I agree with Jeff, that if we can't produce
accessible PDFs from LaTeX directly, it will only be a matter of time
before publishers force authors to move away from LaTeX. But I can
think of a number of organisations that this won't happen and
therefore would be interested in a bespoke LaTeX package for
accessibility (e.g., AMS, APS).
The problem is that it will require more than just a package.
Even with a good package and a revised LaTeX format, authors will
need to adjust there coding habits to be aware of the extra information
required to construct good, consistent tagging structures and many
commonly-used packages will need to be updated to be aware of tagging,
and what is required for proper tagging of structures like mathematical
content, tables and figures.
Ross, if you are interested in this, maybe we could have a face to
face one of these days.
I’m certainly interested.
I’ve been very busy with teaching commitments the past few years.
Hopefully soon I’ll get more time to devote to implementing
structures that are useful for Accessibility, and produce more
example documents. We have a new Head of Department, who is
aware of the need to support Accessibility, so I’m expecting to get
a bit more actual support at my university than I’ve had previously.
Best,
Volker
On 26 October 2016 at 06:12, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
As luck would have it, I'm attending a conference with Volker Sorge (CC'd)
who is also interested in this, so I'm adding him to the conversation.
Thanks,
Jeff
On Tue, Oct 25, 2016 at 9:25 AM, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
I ran across your paper here:
https://www.tug.org/TUGboat/tb30-2/tb95moore.pdf
That describes work that I presented back in 2009.
The best place to find more recent work is here:
https://www.tug.org/twg/accessibility/
Scroll down to the list of papers, with links, near the bottom.
We've been trying to make the conference papers we produce in computer
science accessible to people with disabilities. One of the challenges we
have is that so many of the publications are produced as PDFs and many
authors use Latex.
Do you have any updates on the state-of-the-art on producing
accessible/tagged PDFs directly from Latex? Do you know anyone who would
have the experience to create a Latex package for tagging PDFs? I could
contribute real resources to this project, but I would need to find someone
with the necessary expertise.
Yes, there are not many people with the proper expertise.
In the TeX User Group community, there are people such as:
Boris Veytsman and Bruno Le Floch
who have shown some interest in Accessibility.
Some of the CICM/MKM community are sort-of interested,
via semantic enrichment.
And every so often I get an email from a student somewhere,
wanting a quick and easy solution, to produce an accessible PDF
for their PhD thesis.
I’m in contact with the ISO/PDF working group that develops the PDF/UA
and other standards. This includes engineers from Adobe.
But these guys know of TeX/LaTeX, but don’t have expertise in programming
with it.
Thanks!
Jeff
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com/>
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com/>
Hope this helps, at least a bit.
Cheers
Ross
Dr Ross Moore
Mathematics Dept | Level 2, S2.638 AHH
Macquarie University, NSW 2109, Australia
T: +61 2 9850 8955<tel:%2B61%202%209850%C2%A08955> | F: +61 2 9850 8114<tel:%2B61%202%209850%208114>
M:+61 407 288 255<tel:%2B61%20407%20288%20255> | E: ross.moore at mq.edu.au<mailto:ross.moore at mq.edu.au>
http://www.maths.mq.edu.au<http://www.maths.mq.edu.au/>
[cid:image001.png at 01D030BE.D37A46F0]
CRICOS Provider Number 00002J. Think before you print.
Please consider the environment before printing this email.
This message is intended for the addressee named and may
contain confidential information. If you are not the intended
recipient, please delete it and notify the sender. Views expressed
in this message are those of the individual sender, and are not
necessarily the views of Macquarie University.
Begin forwarded message:
From: Volker Sorge <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>>
Date: 27 October 2016 3:10:38 AM AEDT
To: Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>>
Cc: Ross Moore <ross at maths.mq.edu.au<mailto:ross at maths.mq.edu.au>>
Subject: Re: Making PDFs produced from Latex accessible
Thanks, Jeff, for bringing me in to this conversation.
Hi Ross, you might recall we had a chat about this a couple of years
back in Washington. I agree with Jeff, that if we can't produce
accessible PDFs from LaTeX directly, it will only be a matter of time
before publishers force authors to move away from LaTeX. But I can
think of a number of organisations that this won't happen and
therefore would be interested in a bespoke LaTeX package for
accessibility (e.g., AMS, APS).
Ross, if you are interested in this, maybe we could have a face to
face one of these days.
Best,
Volker
On 26 October 2016 at 06:12, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
As luck would have it, I'm attending a conference with Volker Sorge (CC'd)
who is also interested in this, so I'm adding him to the conversation.
Thanks,
Jeff
On Tue, Oct 25, 2016 at 9:25 AM, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
I ran across your paper here:
https://www.tug.org/TUGboat/tb30-2/tb95moore.pdf
We've been trying to make the conference papers we produce in computer
science accessible to people with disabilities. One of the challenges we
have is that so many of the publications are produced as PDFs and many
authors use Latex.
Do you have any updates on the state-of-the-art on producing
accessible/tagged PDFs directly from Latex? Do you know anyone who would
have the experience to create a Latex package for tagging PDFs? I could
contribute real resources to this project, but I would need to find someone
with the necessary expertise.
Thanks!
Jeff
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com>
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com>
Begin forwarded message:
From: Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>>
Date: 27 October 2016 12:12:38 AM AEDT
To: Ross Moore <ross at maths.mq.edu.au<mailto:ross at maths.mq.edu.au>>, Volker Sorge <volker.sorge at gmail.com<mailto:volker.sorge at gmail.com>>
Subject: Re: Making PDFs produced from Latex accessible
Hi Ross,
As luck would have it, I'm attending a conference with Volker Sorge (CC'd) who is also interested in this, so I'm adding him to the conversation.
Thanks,
Jeff
On Tue, Oct 25, 2016 at 9:25 AM, Jeffrey Bigham <jbigham at cs.cmu.edu<mailto:jbigham at cs.cmu.edu>> wrote:
Hi Ross,
I ran across your paper here:
https://www.tug.org/TUGboat/tb30-2/tb95moore.pdf
We've been trying to make the conference papers we produce in computer science accessible to people with disabilities. One of the challenges we have is that so many of the publications are produced as PDFs and many authors use Latex.
Do you have any updates on the state-of-the-art on producing accessible/tagged PDFs directly from Latex? Do you know anyone who would have the experience to create a Latex package for tagging PDFs? I could contribute real resources to this project, but I would need to find someone with the necessary expertise.
Thanks!
Jeff
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com/>
--
==
Associate Professor
Human-Computer Interaction Institute
Language Technologies Institute
Carnegie Mellon University
jeffreybigham.com<http://jeffreybigham.com/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tug.org/pipermail/accessibility/attachments/20161209/7f230ece/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 4605 bytes
Desc: image001.png
URL: <http://tug.org/pipermail/accessibility/attachments/20161209/7f230ece/attachment-0001.png>
More information about the accessibility
mailing list