Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Ihor Radchenko]

Arkadiusz Świętnicki  writes:

>> I heard about that, but is the situation so dire? Is there nobody at all
>> to take over the maintenance?

> I would like to, but there are multiple issues with this. Sorry if
> that's out of the scope for this thread.

> 1. I have many health issues which make me unable to work for many days
> of the week.
> 2. I have family, so again some less time.

There might be more than one maintainer, so that the workload is
distributed. Of course, that also requires more than one volunteer.

> 3. I don't know how would I even become the maintainer? As far as I
> know, Raman never gave copyright to the FSF so I can't just take
> Emacspeak and continue developing if as if nothing happened. I am fixing
> bugs for myself currently but hoping for a resolution.
> Its not critical yet, however Emacs 31 already broke somethings, such as
> PGP agent and launching Emacspeak on the Mac.

Is Raman not reachable by email anymore? If that's really the case,
emacspeak is GPLv3, so nothing legally stops, say, a fork. Of course,
that's the last resort. Ideally, he can confirm the new
maintainer/maintainers.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at .
Support Org development at ,
or support my work at 

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Arkadiusz Świętnicki]

Ihor Radchenko  writes:

> Arkadiusz Świętnicki  writes:
>
>> Emacspeak supports Orgmode yet but the original author of Emacspeak had
>> stepped down and thus Emacspeak is no longer maintained. If no new
>> maintainer is found, the support will get worse by the day.
>
> I heard about that, but is the situation so dire? Is there nobody at all
> to take over the maintenance?
I would like to, but there are multiple issues with this. Sorry if
that's out of the scope for this thread.
1. I have many health issues which make me unable to work for many days
of the week.
2. I have family, so again some less time.
3. I don't know how would I even become the maintainer? As far as I
know, Raman never gave copyright to the FSF so I can't just take
Emacspeak and continue developing if as if nothing happened. I am fixing
bugs for myself currently but hoping for a resolution.
Its not critical yet, however Emacs 31 already broke somethings, such as
PGP agent and launching Emacspeak on the Mac.
-- 
With best regards
Arkadiusz Świętnicki
https://swietnicki.dev

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Ihor Radchenko]

Arkadiusz Świętnicki  writes:

> Emacspeak supports Orgmode yet but the original author of Emacspeak had
> stepped down and thus Emacspeak is no longer maintained. If no new
> maintainer is found, the support will get worse by the day.

I heard about that, but is the situation so dire? Is there nobody at all
to take over the maintenance?

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at .
Support Org development at ,
or support my work at 

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Arkadiusz Świętnicki]

Pedro Andres Aranda Gutierrez  writes:

> Hi,
>
> we would need some emacspeak expert here...
>  
> My take would be:
> 1. Parse #+LANGUAGE (and other language related keywords that may eventually 
> exist) to
> help select 
>the reader's intended target audience.
> 2. Ignore the rest of the keywords as comments
> 3. Read/parse the document's body.
>- Identify links, images (using and 'alt' tag for the alternative text)
>- Find an adequate way to verbalise tables
>- Find an adequate way to verbalise org-babel output(s)
>
> Doesn't that sound quite like a new exporter?
>
> My .2 cents,/PA
> ox-latex / ox-beamer maintainer
>
> On Sat, 11 Jul 2026 at 11:09, Paul Bryan  wrote:
>
>  Ihor Radchenko  writes:
>
>  > Paul Bryan  writes:
>  >
>  >> Side question: how does emacspeak interact with org-mode? In
>  >> principle, emacspeak could have access to the parse tree for tight
>  >> integration which could be quite powerful. For a potential separate
>  >> thread: is there work org-mode could do to improve this?
>  >
>  > Last time Raman himself was involved in a similar discussion, he did not
>  > raise anything in particular. See
>  > 
>
>  Yes, I see. That looks to be a discussion mostly about exporting - like this 
> thread in fact
>  :) I do see there was some discussion about the org-mode side, and in 
> particular if
>  embedding the org-mode source into the exported pdf was useful. I suppose it 
> wasn't
>  considered all that useful since there doesn't seem to be have been much 
> follow up on
>  that line of discussion.
>
>  But maybe it was just that people are busy and no one really followed it up, 
> and it could
>  indeed be useful after all. I doubt I'll have a lot of time to invest in it, 
> but the question
>  of whether screen reading from a parse tree is useful or not, seems 
> interesting to me.
>  I'll start a new thread if I ever have anything useful to say on the subject.
>
>  --
>  Paul
>
>  --  
> Fragen sind nicht da, um beantwortet zu werden,
> Fragen sind da um gestellt zu werden
> Georg Kreisler
>
> "Sagen's Paradeiser" (ORF: Als Radiohören gefährlich war) => write BE!
> Year 2 of the New Koprocracy
>
> This was produced by a human (implied virtues and weaknesses acknowledged)
> I'd hate this being fed to any form of AS (sorry AI)...
>
hi Pedro,
what you describe here is already handled by speech servers. Emacspeak
scans the text for properties and ity sends text to the speech server
for synthesis. There's no need for a separate exporter (fortunately).
-- 
With best regards
Arkadiusz Świętnicki
https://swietnicki.dev

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Arkadiusz Świętnicki]

Paul Bryan  writes:

> Sébastien Hinderer  writes:
>
>>
>> Understood. Just for the sake of completeness, let me say that as a
>> blind user my preference would be to have access to the original
>> document as I expect it to be muhc more readable than many derived
>> formats. I realise this may not be true for everybody, and also that
>> authors may not want to share their sources. I still wanted to bring
>> this to the conversation so that peoplealso have this possibility in
>> mind.  For researhc papers and books for instance, I way prefer reading
>> LaTeX sources than anything else.
>
> I'm curious how reading LaTeX source compares with reading org-mode
> source for blind users. Naively I would expect org-mode to be more
> suited to screen readers as it a lot less verbose, with much less
> boilerplate to parse or even to ignore.
>
> In the video conversation between Prot and Arkadiusz
> , I recall some
> discussion about how emacspeak can use text properties to add
> emphasis. I presume that sort of thing would be easier in org-mode
> than in LaTeX - or at least that it would be easier to code.
>
> Of course, most academic articles will be written in LaTeX, so the
> above question is somewhat limited in scope.
>
> Side question: how does emacspeak interact with org-mode? In
> principle, emacspeak could have access to the parse tree for tight
> integration which could be quite powerful. For a potential separate
> thread: is there work org-mode could do to improve this?
>
> --
> Paul
>
hi Paul,
Emacspeak supports Orgmode yet but the original author of Emacspeak had
stepped down and thus Emacspeak is no longer maintained. If no new
maintainer is found, the support will get worse by the day.
-- 
With best regards
Arkadiusz Świętnicki
https://swietnicki.dev

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Pedro Andres Aranda Gutierrez]

Hi,

we would need some emacspeak expert here...

My take would be:
1. Parse #+LANGUAGE (and other language related keywords that may
eventually exist) to help select
   the reader's intended target audience.
2. Ignore the rest of the keywords as comments
3. Read/parse the document's body.
   - Identify links, images (using and 'alt' tag for the alternative text)
   - Find an adequate way to verbalise tables
   - Find an adequate way to verbalise org-babel output(s)

Doesn't that sound quite like a new exporter?

My .2 cents,/PA
ox-latex / ox-beamer maintainer

On Sat, 11 Jul 2026 at 11:09, Paul Bryan  wrote:

> Ihor Radchenko  writes:
>
> > Paul Bryan  writes:
> >
> >> Side question: how does emacspeak interact with org-mode? In
> >> principle, emacspeak could have access to the parse tree for tight
> >> integration which could be quite powerful. For a potential separate
> >> thread: is there work org-mode could do to improve this?
> >
> > Last time Raman himself was involved in a similar discussion, he did not
> > raise anything in particular. See
> > 
>
> Yes, I see. That looks to be a discussion mostly about exporting - like
> this thread in fact :) I do see there was some discussion about the
> org-mode side, and in particular if embedding the org-mode source into the
> exported pdf was useful. I suppose it wasn't considered all that useful
> since there doesn't seem to be have been much follow up on that line of
> discussion.
>
> But maybe it was just that people are busy and no one really followed it
> up, and it could indeed be useful after all. I doubt I'll have a lot of
> time to invest in it, but the question of whether screen reading from a
> parse tree is useful or not, seems interesting to me. I'll start a new
> thread if I ever have anything useful to say on the subject.
>
> --
> Paul
>


-- 
Fragen sind nicht da, um beantwortet zu werden,
Fragen sind da um gestellt zu werden
Georg Kreisler

"Sagen's Paradeiser" (ORF: Als Radiohören gefährlich war) => write BE!
Year 2 of the New Koprocracy

This was produced by a human (implied virtues and weaknesses acknowledged)
I'd hate this being fed to any form of AS (sorry AI)...

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Paul Bryan]

Ihor Radchenko  writes:

> Paul Bryan  writes:
>
>> Side question: how does emacspeak interact with org-mode? In
>> principle, emacspeak could have access to the parse tree for tight
>> integration which could be quite powerful. For a potential separate
>> thread: is there work org-mode could do to improve this?
>
> Last time Raman himself was involved in a similar discussion, he did not
> raise anything in particular. See
> 

Yes, I see. That looks to be a discussion mostly about exporting - like this 
thread in fact :) I do see there was some discussion about the org-mode side, 
and in particular if embedding the org-mode source into the exported pdf was 
useful. I suppose it wasn't considered all that useful since there doesn't seem 
to be have been much follow up on that line of discussion.

But maybe it was just that people are busy and no one really followed it up, 
and it could indeed be useful after all. I doubt I'll have a lot of time to 
invest in it, but the question of whether screen reading from a parse tree is 
useful or not, seems interesting to me. I'll start a new thread if I ever have 
anything useful to say on the subject.

--
Paul

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Ihor Radchenko]

Paul Bryan  writes:

> Side question: how does emacspeak interact with org-mode? In principle, 
> emacspeak could have access to the parse tree for tight integration which 
> could be quite powerful. For a potential separate thread: is there work 
> org-mode could do to improve this?

Last time Raman himself was involved in a similar discussion, he did not
raise anything in particular. See 
https://list.orgmode.org/orgmode/87czew3w5l.fsf@localhost/

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at .
Support Org development at ,
or support my work at 

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Paul Bryan]

Sébastien Hinderer  writes:

>
> Understood. Just for the sake of completeness, let me say that as a
> blind user my preference would be to have access to the original
> document as I expect it to be muhc more readable than many derived
> formats. I realise this may not be true for everybody, and also that
> authors may not want to share their sources. I still wanted to bring
> this to the conversation so that peoplealso have this possibility in
> mind.  For researhc papers and books for instance, I way prefer reading
> LaTeX sources than anything else.

I'm curious how reading LaTeX source compares with reading org-mode source for 
blind users. Naively I would expect org-mode to be more suited to screen 
readers as it a lot less verbose, with much less boilerplate to parse or even 
to ignore.

In the video conversation between Prot and Arkadiusz  
, I recall some discussion about 
how emacspeak can use text properties to add emphasis. I presume that sort of 
thing would be easier in org-mode than in LaTeX - or at least that it would be 
easier to code.

Of course, most academic articles will be written in LaTeX, so the above 
question is somewhat limited in scope.

Side question: how does emacspeak interact with org-mode? In principle, 
emacspeak could have access to the parse tree for tight integration which could 
be quite powerful. For a potential separate thread: is there work org-mode 
could do to improve this?

--
Paul

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Sébastien Hinderer]

Ihor Radchenko (2026/07/10 17:28 +):
> Sébastien Hinderer  writes:
>
> > One thing that I was wondering: why is this experimented with pdflatex
> > rather than lualatex which seems more future-looking?
>
> It is both.
> LuaLaTeX is easier as there is dedicated effort from LaTeX maintainers
> for LuaLaTeX to support PDF tagging and PDF/UA2. However, the existing
> PDFLaTeX documents may not be easy to convert to LuaLaTeX. So, having a
> way to produce decent tagged PDF with PDFLaTeX is valuable.

Makes sense, thanks.

> During the last meetup Pedro mentioned one simple example - books
> written in Org mode for pdflatex. They often contain all kinds of
> pdflatex-specific typographical tweaks and cannot be converted to
> lualatex. Having a way to make them accessible will be nice.

Understood. Just for the sake of completeness, let me say that as a
blind user my preference would be to have access to the original
document as I expect it to be muhc more readable than many derived
formats. I realise this may not be true for everybody, and also that
authors may not want to share their sources. I still wanted to bring
this to the conversation so that peoplealso have this possibility in
mind.  For researhc papers and books for instance, I way prefer reading
LaTeX sources than anything else.

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Jacob S. Gordon]

Hello,

I have a bit of experience generating UA-2 documents this way, and
working around some of the limitations. Happy to test things out.

On 2026-07-10 13:14, Ihor Radchenko wrote:
> 1. Reference LaTeX (lualatex and pdflatex) documents that showcase
>how to generate accessible PDFs. It is especially important for
>pdflatex where the PDF tagging support is limited and special
>tricks may be needed to produce accessible documents.
> 
> 2. General information about various edge cases we need to be aware of.
>In latex in particular and for accessibility in general
> 
> 3. Useful latex packages that can help in this area

The LaTeX project maintains a [status page] of all the classes and
packages, with issues and examples gathered [here]. IME, if something
isn’t working there’s usually some indication there, so I hope we can
avoid duplicate efforts. At the document level you can [check the
tagging status], and enable some other debugging options:

  \DocumentMetadata{
check-tagging-status,
debug={uncompress,tagpdf}
  }

[status page]
[here]
[check the tagging status]

-- 
Jacob S. Gordon
jacob.as.gor...@gmail.com
Please don’t send me HTML emails or MS Office/Apple iWork documents.
https://useplaintext.email/#etiquette
https://www.fsf.org/campaigns/opendocument

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Ihor Radchenko]

Sébastien Hinderer  writes:

> One thing that I was wondering: why is this experimented with pdflatex
> rather than lualatex which seems more future-looking?

It is both.
LuaLaTeX is easier as there is dedicated effort from LaTeX maintainers
for LuaLaTeX to support PDF tagging and PDF/UA2. However, the existing
PDFLaTeX documents may not be easy to convert to LuaLaTeX. So, having a
way to produce decent tagged PDF with PDFLaTeX is valuable.
During the last meetup Pedro mentioned one simple example - books
written in Org mode for pdflatex. They often contain all kinds of
pdflatex-specific typographical tweaks and cannot be converted to
lualatex. Having a way to make them accessible will be nice.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at .
Support Org development at ,
or support my work at 

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Sébastien Hinderer]

Hello,

Many thanks for this initiative.

As a blind user of Org-Mode, I especially appreciate tthe interest for
accessibility.

I am not sure I have anything useful to say on the mentionned areas, but
I am hapy to help as I can. For a start, I can say that as far as I know
there is no completely accessible PDF docment viewer under Linux (the OS
I use). So what visually impaired people I know do on Linux is to open
PDF documents in Firefox which, as a web browser, is accessible.

One thing that I was wondering: why is this experimented with pdflatex
rather than lualatex which seems more future-looking?

Best wishes,

Seb.

Re: [RFC] Help with PDF/UA2 accessibility tests for ox-latex

[Arkadiusz Świętnicki]

Hi Ihor and all,
thank you for recognizing the need to improve accessibility. I am
currently working on various Tex documents, both correct and incorrect,
to showcase correct behavior, but I think showcasing incorrect behavior
is even more important. PDF is a nasty format, full of quirks which can
even surprise experts, as tools do not find it. Only screen readers
users can.

Ihor Radchenko  writes:

> Dear all,
>
> We are looking to improve Org LaTeX export to produce screen
> reader-friendly accessible documents that can stick to PDF/UA2 standard.
> However, we are not experts in the field, and we need help from people
> more familiar with accessibility standards.
>
> In particular, we want:
>
> 1. Reference LaTeX (lualatex and pdflatex) documents that showcase how
>to generate accessible PDFs. It is especially important for pdflatex
>where the PDF tagging support is limited and special tricks may be
>needed to produce accessible documents.
>
> 2. General information about various edge cases we need to be aware of.
>In latex in particular and for accessibility in general
>
> 3. Useful latex packages that can help in this area
>
> Pedro, the ox-latex maintainer, created a repository where we will
> collect various examples. See https://codeberg.org/paaguti/pdf-ua2-tests
>
> Feel free to either write here, on the list, or create issues on
> codeberg with your suggestions.
>
> Best regards,
> Ihor
>
> "Pedro A. Aranda"  writes:
>
>> there's been going on WRT accessibility in the list recently.
>> I have created a repo in my codeberg to collect all our efforts and 
>> tests in this field. The objective would be to integrate as much as 
>> possible in the LaTeX exporter and to feed the discussion with the 
>> pdflatex maintainers, showing concrete examples of what doesn't work as 
>> needed.
>>
>> Currently I have uploaded an extended version of the ltx-talk Test 
>> presentation and a first concept for a pdflatex with a German document.
>> Feel free to clone and contribute.
>>
>> Please send me PRs with your thoughts and tests.

-- 
With best regards
Arkadiusz Świętnicki
https://swietnicki.dev