E-mail List Archives

Search E-mail List Archives for

Number of posts in this thread: 5 (In chronological order)

From: Christine Hogenkamp
Date: Tue, Jun 14 2022 10:45AM
Subject: PDF/UA Matterhorn Protocols - how to use?
No previous message | Next message →

Hello everyone,

I am working on creating a how-to doc for my organization for how to create
a PDF that passes PDF/UA, which includes a section on how to check the PDF
to ensure it has passed. I have downloaded a copy of the Matterhorn
Protocols which was recommended as a way to help users interpret the PDF/UA
standards doc which is, in itself, a fairly complex technical document. The
problem I am having is that I can't seem to find instructions on how this
checklist should be used.

For example, this article:
https://www.pdfa.org/climbing-the-matterhorn-an-introduction-to-the-definitive-algorithm-for-pdfua-conformance/

"How to Use the Matterhorn Protocol
The basic approach for implementing the Matterhorn Protocol is to map the
Failure Conditions to the various tasks implied by the specific
word-processing, content extraction or other context."

Perhaps I am just obtuse, but I am having a hard time parsing what this
actually means in real life terms. How does one "map the failure
conditions"? Am I meant to open the accessible PDF in question in Acrobat,
then read through the Matterhorn Protocols while I go through my PDF to
manually check each Human checklist item to see if I can find any instances
of failure in the PDF, using the Tags panel? Then use PAC3 to check the
Machine checklist items?

It would be very helpful if anyone could point me towards more specific
step by step instructions for using the Matterhorn Protocols.

Thank you in advance!

*x*
*Christine Hogenkamp (She, Her)*
Front-end Developer
ContextCreative.com <http://contextcreative.com/>;

*20 years of purposeful creativity*
* ˟**˟*

From: L Snider
Date: Tue, Jun 14 2022 2:45PM
Subject: Re: PDF/UA Matterhorn Protocols - how to use?
← Previous message | Next message →

Hi Christine,

I find WCAG and PDF/UA hard to understand in terms of wording even after
all these years! WCAG got a bit better the last 3-4 years, but I still find
it not that usable for many...

Basically, download this PDF/UA checker (it was PAC 3). It will guide you
in terms of errors, and then look up the errors online and figure out how
to solve them (there is one Section 508 site that does PAC 2 errors, it has
handy videos)

https://pdfua.foundation/en/pdf-accessibility-checker-pac

Cheers

Lisa

On Tue, Jun 14, 2022 at 1:46 PM Christine Hogenkamp <
= EMAIL ADDRESS REMOVED = > wrote:

> Hello everyone,
>
> I am working on creating a how-to doc for my organization for how to create
> a PDF that passes PDF/UA, which includes a section on how to check the PDF
> to ensure it has passed. I have downloaded a copy of the Matterhorn
> Protocols which was recommended as a way to help users interpret the PDF/UA
> standards doc which is, in itself, a fairly complex technical document. The
> problem I am having is that I can't seem to find instructions on how this
> checklist should be used.
>
> For example, this article:
>
> https://www.pdfa.org/climbing-the-matterhorn-an-introduction-to-the-definitive-algorithm-for-pdfua-conformance/
>
> "How to Use the Matterhorn Protocol
> The basic approach for implementing the Matterhorn Protocol is to map the
> Failure Conditions to the various tasks implied by the specific
> word-processing, content extraction or other context."
>
> Perhaps I am just obtuse, but I am having a hard time parsing what this
> actually means in real life terms. How does one "map the failure
> conditions"? Am I meant to open the accessible PDF in question in Acrobat,
> then read through the Matterhorn Protocols while I go through my PDF to
> manually check each Human checklist item to see if I can find any instances
> of failure in the PDF, using the Tags panel? Then use PAC3 to check the
> Machine checklist items?
>
> It would be very helpful if anyone could point me towards more specific
> step by step instructions for using the Matterhorn Protocols.
>
> Thank you in advance!
>
> *x*
> *Christine Hogenkamp (She, Her)*
> Front-end Developer
> ContextCreative.com <http://contextcreative.com/>;
>
> *20 years of purposeful creativity*
> * ˟**˟*
> > > > >

From: Duff Johnson
Date: Tue, Jun 14 2022 2:58PM
Subject: Re: PDF/UA Matterhorn Protocols - how to use?
← Previous message | Next message →

Hi Christine,

The design of the Matterhorn Protocol is intentionally open-ended to allow for all manner of implementations. Most use it to underpin software development or categorize test procedures. It's not really intended for end users in a remediation context…. except at a systematic, workflow sort of level. For example, you could use Matterhorn to verify that your software is capable of addressing each type of check necessary to a file's conformance with PDF/UA.

Another option: some use Matterhorn to develop a precise means of identifying and categorizing specific types of errors.

It would certainly be possible to iterate through the tag tree of a PDF and validate each-and-every tag for conformance with the human check provisions. In practice, this can be done quite quickly once the user is familiar with the concept of PDF tags in general.

It's up to the user to understand that a list item's bullet must be contained in an <Lbl> tag, and not in the <LBody> tag (to take one common example).

Matterhorn simply allows you to classify an error (In the above case, 01-006 "The structure type and attributes of a structure element are not semantically appropriate for the structure element“) as it relates to PDF/UA-1.

I hope this information is useful.

You may also find the PDF Association's other formal advice in this area, the Tagged PDF Best Practice Guide: Syntax, to be useful.

https://www.pdfa.org/resource/tagged-pdf-best-practice-guide-syntax/

Duff.







> On Jun 14, 2022, at 12:45, Christine Hogenkamp < = EMAIL ADDRESS REMOVED = > wrote:
>
> Hello everyone,
>
> I am working on creating a how-to doc for my organization for how to create
> a PDF that passes PDF/UA, which includes a section on how to check the PDF
> to ensure it has passed. I have downloaded a copy of the Matterhorn
> Protocols which was recommended as a way to help users interpret the PDF/UA
> standards doc which is, in itself, a fairly complex technical document. The
> problem I am having is that I can't seem to find instructions on how this
> checklist should be used.
>
> For example, this article:
> https://www.pdfa.org/climbing-the-matterhorn-an-introduction-to-the-definitive-algorithm-for-pdfua-conformance/
>
> "How to Use the Matterhorn Protocol
> The basic approach for implementing the Matterhorn Protocol is to map the
> Failure Conditions to the various tasks implied by the specific
> word-processing, content extraction or other context."
>
> Perhaps I am just obtuse, but I am having a hard time parsing what this
> actually means in real life terms. How does one "map the failure
> conditions"? Am I meant to open the accessible PDF in question in Acrobat,
> then read through the Matterhorn Protocols while I go through my PDF to
> manually check each Human checklist item to see if I can find any instances
> of failure in the PDF, using the Tags panel? Then use PAC3 to check the
> Machine checklist items?
>
> It would be very helpful if anyone could point me towards more specific
> step by step instructions for using the Matterhorn Protocols.
>
> Thank you in advance!
>
> *x*
> *Christine Hogenkamp (She, Her)*
> Front-end Developer
> ContextCreative.com <http://contextcreative.com/>;
>
> *20 years of purposeful creativity*
> * ˟**˟*
> > > >

From: Alan Zaitchik
Date: Wed, Jun 15 2022 6:13AM
Subject: Re: PDF/UA Matterhorn Protocols - how to use?
← Previous message | Next message →

I use it for documentation purposes only since any real testing will be done with tools, of which there are a number. And with screen readers, of course.
A

> On Jun 14, 2022, at 12:45, Christine Hogenkamp < = EMAIL ADDRESS REMOVED = > wrote:
>
> Hello everyone,
>
> I am working on creating a how-to doc for my organization for how to create
> a PDF that passes PDF/UA, which includes a section on how to check the PDF
> to ensure it has passed. I have downloaded a copy of the Matterhorn
> Protocols which was recommended as a way to help users interpret the PDF/UA
> standards doc which is, in itself, a fairly complex technical document. The
> problem I am having is that I can't seem to find instructions on how this
> checklist should be used.
>
> For example, this article:
> https://www.pdfa.org/climbing-the-matterhorn-an-introduction-to-the-definitive-algorithm-for-pdfua-conformance/
>
> "How to Use the Matterhorn Protocol
> The basic approach for implementing the Matterhorn Protocol is to map the
> Failure Conditions to the various tasks implied by the specific
> word-processing, content extraction or other context."
>
> Perhaps I am just obtuse, but I am having a hard time parsing what this
> actually means in real life terms. How does one "map the failure
> conditions"? Am I meant to open the accessible PDF in question in Acrobat,
> then read through the Matterhorn Protocols while I go through my PDF to
> manually check each Human checklist item to see if I can find any instances
> of failure in the PDF, using the Tags panel? Then use PAC3 to check the
> Machine checklist items?
>
> It would be very helpful if anyone could point me towards more specific
> step by step instructions for using the Matterhorn Protocols.
>
> Thank you in advance!
>
> *x*
> *Christine Hogenkamp (She, Her)*
> Front-end Developer
> ContextCreative.com <http://contextcreative.com/>;
>
> *20 years of purposeful creativity*
> * ˟**˟*
>

From: Christine Hogenkamp
Date: Thu, Jun 16 2022 10:37AM
Subject: Re: PDF/UA Matterhorn Protocols - how to use?
← Previous message | No next message

Thanks for the responses! Let me try to consolidate my replies:

@Lisa: Thanks for the link, we do sometimes use PAC3 (at client request)
and we have an ongoing doc for troubleshooting the errors we get; some are
easy to understand and others are more difficult to parse what needs to be
fixed and how. I have tried to find a comprehensive guide for PAC errors,
so far this website covers the most I've seen in one place:
https://taggedpdf.com/508-pdf-help-center/

Right now my research is more of a "future-proofing" for future projects
than "We actively need this for work right now" and I think I am slowly
wrapping my head around the idea that some of PDF/UA is not directly in the
hands of our designers, but more that we make sure we use programs that
will format the PDF correctly. Definitely some of the guidelines of PDF/UA
and the checkpoints of the Matterhorn Protocol seem to describe technical
aspects of PDF that are more software-oriented, for the apps that create
PDFs. Our designers just have to trust that InDesign is doing its job to
format the accessible tagged PDF if we do our part to format the file to
correctly create tags, logical reading order, alt text etc.

I would like for us to also improve our use of PAC 3 (or I think they just
released PAC 2021?) to get fewer errors, since that should mean our
formatting of the source design file has improved.


@Duff: Thank you for the helpful contextualizing of the Matterhorn
Protocol. It makes a lot of sense as a guide for software creation, there
are definitely some technical aspects that are deep in the "gears" of a PDF
that I think of as just the default state of a PDF file. We aren't looking
to reinvent the PDF wheel, just making sure we are using the correct
settings in InDesign (and correct fixes in Acrobat for things that can't be
done in Indesign) so that the tagged PDF is formatted to pass PDF/UA.

re: "It's up to the user to understand that a list item's bullet must be
contained in an <Lbl> tag, and not in the <LBody> tag (to take one common
example)."

We don't currently have a lot of info about correct formatting for PDF
tags, I'm a developer who works in HTML etc and I know PDF tags don't work
exactly like HTML tags do (with *some* overlap/similar structure) so I
would welcome any suggestions for resources/websites that can explain how
these common PDF tags should ideally be created by the app that creates the
PDFs. Even better if the info is Indesign specific :)

Currently we are at the mercy of how InDesign is programmed to export using
the tagged PDF option, for the overall arrangement of the tag tree and
which tags are used. I have noticed some discrepancy between how the tag
tree lists tags in the Structure panel of Indesign and the Tag panel of
Acrobat, the export process seems to create extra tags in a way we don't
have control over. Maybe that's ok? Maybe these defaults are the correct
tags a PDF needs in the overall document structure, such as the use of the
tags <Document> and <Art> etc? I have advised our designers to use the
native tools as much as possible to encourage the proper tags on export,
i.e. using the Insert Table dialogue, instead of creating a fake table with
the line tool or coloured boxes, etc.

re: "Matterhorn simply allows you to classify an error (In the above case,
01-006 "The structure type and attributes of a structure element are not
semantically appropriate for the structure element“) as it relates to
PDF/UA-1."

Oh that makes sense, to have a short form for each type of error. I think
what I will have to do is try to find a good place in our existing
accessible PDF workflow when we review the PDF to add a checklist of the
human-reviewable checkpoints and then the tester will have to use PAC 3 to
test the machine-testable ones. Maybe it's as simple as adding a
conditional step to our existing instructions, i.e. "If PDF needs to pass
PDF/UA, use table of checkpoints in doc 'Title of PDF/UA doc' for the items
that require human review, then use PAC 3"

And thank you for the link to "Tagged PDF Best Practice Guide: Syntax". It
has some great info, I have added it as a resource to our shared folder for
such docs. I have done so much reading lately my brain is swimming ha ha!


@Alan: Thank you for your perspective, do you have any recommendations for
tools that we could use? Currently we use the Acrobat Accessibility Checker
(just to catch basic errors) then we manually review the Tag panel for
correct logical reading order and tags used, and also test with NVDA to
ensure all content is focusable and read out, that the reading experience
seems good and the user can move around the doc in different ways.


*x*
*Christine Hogenkamp (She, Her)*
Front-end Developer
ContextCreative.com <http://contextcreative.com/>;

*20 years of purposeful creativity*
* ˟**˟*

>
>