Accessible PDF

24 views
Skip to first unread message

Rob Beezer

unread,
Jun 11, 2026, 11:24:05 PMJun 11
to pretex...@googlegroups.com
Accessibility experts,

My tools suggest this PDF is "UA-1". And I think the math is backed by speech
strings that a screen reader might read.

If anybody is interested, can you take it for a drive?

Thanks,
Rob
pdf-fo-development.pdf

Joseph DiMuro

unread,
Jun 12, 2026, 4:27:35 PMJun 12
to bee...@privacyport.com, pretex...@googlegroups.com
Hi Rob. Adobe Acrobat Pro found one issue: table headers. But the issue isn't with the table; it's with the side-by-side. Apparently, the side-by-side is treated as a table in the PDF, so Adobe is barking about a table with no header.

I don't know if this is something to be fixed or ignored. Adobe isn't 100% to be trusted, after all.

-Joseph

--
You received this message because you are subscribed to the Google Groups "PreTeXt accessibility" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pretext-a11y...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pretext-a11y/MTAwMDA0MS5iZWV6ZXI.1781234640%40pnsh.

David Austin

unread,
Jun 12, 2026, 4:46:50 PMJun 12
to Joseph DiMuro, bee...@privacyport.com, pretex...@googlegroups.com
I put the PDF into Blackboard, our LMS, and the accessibility checker said: "100%.  Perfect!  Keep up the good work!"

David

Rob Beezer

unread,
Jun 13, 2026, 12:07:54 PMJun 13
to pretex...@googlegroups.com
Thanks very much, Joseph. Just what I needed, since I don't own Adobe Acrobat.
The development loop has been using veraPDF as a checker, which maybe didn't
flag this?

I saw that #sidebyside were going into tables, which I figured might be a bad
shortcut. (I'm not seeing everything my assistant has done, I need to do a
*lot* of reviewing). Apache FOP seems pretty flexible, so perhaps we can get
back to something more like "panels".

Thanks again!

Rob

On 6/12/26 13:27, Joseph DiMuro wrote:
> Hi Rob. Adobe Acrobat Pro found one issue: table headers. But the issue isn't
> with the table; it's with the *side-by-side*. Apparently, the side-by-side is
> treated as a table in the PDF, so Adobe is barking about a table with no header.
>
> I don't know if this is something to be fixed or ignored. Adobe isn't 100% to be
> trusted, after all.
>
> -Joseph
>
> On Thu, Jun 11, 2026 at 8:24 PM 'Rob Beezer' via PreTeXt accessibility <pretext-
> a1...@googlegroups.com <mailto:pretex...@googlegroups.com>> wrote:
>
> Accessibility experts,
>
> My tools suggest this PDF is "UA-1".  And I think the math is backed by speech
> strings that a screen reader might read.
>
> If anybody is interested, can you take it for a drive?
>
> Thanks,
> Rob
>
> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt accessibility" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to pretext-a11y...@googlegroups.com <mailto:pretext-
> a11y%2Bunsu...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pretext-
> a11y/MTAwMDA0MS5iZWV6ZXI.1781234640%40pnsh <https://groups.google.com/d/
> msgid/pretext-a11y/MTAwMDA0MS5iZWV6ZXI.1781234640%40pnsh>.
>
> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt accessibility" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> to pretext-a11y...@googlegroups.com <mailto:pretext-
> a11y+uns...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pretext-a11y/
> CAJdq7mKzphjRhhYMRtZ_dncRHDahx2o%2BiaCxcJUQWjzNdJHoXQ%40mail.gmail.com <https://
> groups.google.com/d/msgid/pretext-a11y/
> CAJdq7mKzphjRhhYMRtZ_dncRHDahx2o%2BiaCxcJUQWjzNdJHoXQ%40mail.gmail.com?
> utm_medium=email&utm_source=footer>.

Rob Beezer

unread,
Jun 13, 2026, 12:09:26 PMJun 13
to pretex...@googlegroups.com
Thanks, David. Very much appreciate the testing. And that's the sort of
encouragement I need! ;-)

Rob

On 6/12/26 13:46, David Austin wrote:
> I put the PDF into Blackboard, our LMS, and the accessibility checker said:
> "100%.  Perfect!  Keep up the good work!"
>
> David
>
> On Fri, Jun 12, 2026 at 4:27 PM Joseph DiMuro <joseph...@biola.edu
> <mailto:joseph...@biola.edu>> wrote:
>
> Hi Rob. Adobe Acrobat Pro found one issue: table headers. But the issue
> isn't with the table; it's with the *side-by-side*. Apparently, the side-by-
> side is treated as a table in the PDF, so Adobe is barking about a table
> with no header.
>
> I don't know if this is something to be fixed or ignored. Adobe isn't 100%
> to be trusted, after all.
>
> -Joseph
>
> On Thu, Jun 11, 2026 at 8:24 PM 'Rob Beezer' via PreTeXt accessibility
> <pretex...@googlegroups.com <mailto:pretex...@googlegroups.com>> wrote:
>
> Accessibility experts,
>
> My tools suggest this PDF is "UA-1".  And I think the math is backed by
> speech
> strings that a screen reader might read.
>
> If anybody is interested, can you take it for a drive?
>
> Thanks,
> Rob
>
> --
> You received this message because you are subscribed to the Google
> Groups "PreTeXt accessibility" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pretext-a11y...@googlegroups.com <mailto:pretext-
> a11y%2Bunsu...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pretext-
> a11y/MTAwMDA0MS5iZWV6ZXI.1781234640%40pnsh <https://groups.google.com/d/
> msgid/pretext-a11y/MTAwMDA0MS5iZWV6ZXI.1781234640%40pnsh>.
>
> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt accessibility" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> To view this discussion visit https://groups.google.com/d/msgid/pretext-
> a11y/CAJdq7mKzphjRhhYMRtZ_dncRHDahx2o%2BiaCxcJUQWjzNdJHoXQ%40mail.gmail.com
> <https://groups.google.com/d/msgid/pretext-a11y/
> CAJdq7mKzphjRhhYMRtZ_dncRHDahx2o%2BiaCxcJUQWjzNdJHoXQ%40mail.gmail.com?
> utm_medium=email&utm_source=footer>.
>
> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt accessibility" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> To view this discussion visit https://groups.google.com/d/msgid/pretext-a11y/
> CANXmVMC%2BdKXF6FrJUr-k2pJwQRHgmJ3OyncpqYMbPTOjKje3vg%40mail.gmail.com <https://
> groups.google.com/d/msgid/pretext-a11y/CANXmVMC%2BdKXF6FrJUr-
> k2pJwQRHgmJ3OyncpqYMbPTOjKje3vg%40mail.gmail.com?
> utm_medium=email&utm_source=footer>.

Rob Beezer

unread,
Jun 13, 2026, 2:26:47 PMJun 13
to pretex...@googlegroups.com
There is a more complete posting on "pretext-dev", just now. With updated
samples at the website.

Michael Cantino

unread,
Jun 23, 2026, 7:43:39 PM (12 days ago) Jun 23
to bee...@privacyport.com, pretex...@googlegroups.com
Hi! 
I saw this come through, but I was too busy to do anything about it. Also, I never got around to taking a deep dive into PDF accessibility because it is needlessly obtuse and annoying. I did ask my friend, Jennifer Lucas, to take a look at it though. Jennifer previously worked at PCC and at the Washington State School for the Blind as a Digital Accessibility Specialist. She's spent a lot of time really understanding the ins and outs of PDF accessibility. She took a quick look at the PDF and gave some notes. I didn't really give her any context, other than the group is working on accessible PDF output from PreTeXt. I think this was just a manual review of the tags to see how everything was structured, and she didn't use an automated checker. I know she recommends the PAC PDF checker

From Jennifer:
  • The link text color is too similar to the body text color to pass WCAG without underlines.
  • The alt text on the links are the target addresses
  • I don't agree with his choice of footnote tagging system but maybe that's something you already discussed and agreed on. It interrupts the sentence to read the text of the footnote immediately. I find that a little more disorienting than the popular option of putting the footnote directly after the paragraph where it's referenced, which I only think is reasonable if the footnote is only referenced once. I'm not going to get into all of the rules of tagging references and notes because it's a broken system that helps nobody. 
  • Those keyboard key symbols should be given alt text so they speak the name of the key instead of a description of the symbol ("Up" or "Up Arrow" instead of "upwards hollow arrow"). You can put alt text on a <span> - just make sure you use the Alt Text field and not the Actual Text field unless you want it to be what people get when they copy and paste out of the document. 
  • On the example table, there's no reason not to mark the first column as header cells for better navigation. The name should probably read before the table even though it's visually underneath. 
  • The side-by-side example is tagged as a table, and it should not be. 
  • The math is provided as alt text on figures. They're well described but not explorable. I think for simple math like a²+b²=c² it might be better to use real text with appropriate unicode symbols instead of images, but it's debatable. 
  • On page 6, 3·4 is read as "3 dot 4"
  • The equation in the section Theorem 7.5 reads in the wrong order (I think) 
  • The blank in the fill-in-the-blank example is not spoken, the sentence just continues uninterrupted. 
  • The references list is tagged as two single-item lists. 
  • All of the page numbers read in sequence at the end of the document (should be artifacted and not read at all)
  • Some of the text content in section 6 verbatim is not contained in appropriate tags (like <p>). 
That's what I found in a quick look, there's also just a lot of unnecessary divs that I would want to step my tags out of if it were mine but that has absolutely no impact on the end-user and is just about the experience of people validating the tag tree. 

----
I took a quick look with NVDA. I found some of the same things Jennifer found. 
  • Alt text on links did read the URL and not the display text
  • When I came across a footnote, it read the full footnote immediately instead of continuing on with the sentence. I haven't experienced a ton of footnotes with a screen reader, so I don't know how typical that is, but it was a little disorienting for me. 
  • The side-by-side was a little funky with a screen reader. It announced it as a table with 4 columns. Columns 1 and 3 are empty, presumably for visual spacing. Not very pleasant for a screen reader.
  • Surprise to no one, math is not as nice in PDF. Reads as alt text, which means no braille math notation. The long alt text descriptions being interrupted with "graphic" in the middle of expressions is distracting and potentially confusing. 
  • I didn't like how the description on the displayed quadratic formula read. It starts with "graphic start layout first row", but it never announces a second row. The description ends with "end layout". I don't know if we need the layout and first row information included in the description.
  • Jennifer said 7.5 might be reading in the incorrect order, but I'm guessing that star is a reference symbol and that's why it's reading first. 
  • Lists read funny. The screen reader would pause on each number or bullet, and I had to arrow down again to have the list item content read aloud. Usually the screen reader will announce the number or bullet and then immediately start reading the list item content. 
  • The list in the References section at the end is formatted as 2 lists with one item each, instead of 1 list with 2 items.
  • Jennifer noted that the multiplication expressions on page 6 announce the multiplication dot as "dot". I was going to ignore this, but all of the periods adjacent to math content on this page are also read as "dot". That probably needs some attention to avoid confusing readers. 
That's about all I found with the screen reader.

I know I owe David Austin info about reading his graphic on the Monarch. It's been a crazy month, but my summer break starts tomorrow!

Michael Cantino

On Sat, Jun 13, 2026 at 11:26 AM 'Rob Beezer' via PreTeXt accessibility <pretex...@googlegroups.com> wrote:
There is a more complete posting on "pretext-dev", just now.  With updated
samples at the website.

--
You received this message because you are subscribed to the Google Groups "PreTeXt accessibility" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pretext-a11y...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pretext-a11y/MTAwMDAxNS5iZWV6ZXI.1781375204%40pnsh.
Reply all
Reply to author
Forward
0 new messages