[Athen] [EXT]Re: PDF remediation help

[Metzessible]

It never ceases to amaze me how much hate this stupid file format receives.
It came out in 1993 and it was never intended to be used as an accessible
format. In fact, tagging wasn't even available until 2001. I would argue
the purpose of tagging was merely so PDF could be translated into yet
another file format (like Word or HTML), and using it for accessibility
just happened to be a [un]happy accident.

Don't get me wrong. I'm not a fan of it either, and I worked on PDF/UA! But
getting frustrated with this because it doesn't fit your needs just seems
unnecessary. I mean, if you hate it so much, stop making them. I mean, even
saying "make a PDF" is a misnomer. You're converting something from the
original format and making it something completely different. But tell me
another way to get a format that anyone can use that takes an image and
adds a structure to it, is editable without complex packaging, works
cross-platform, and retains the exact look and feel of the original
document's complex layout.

I've got a print disability. If someone converted a school-provided PDF to
be a wall of text, I'd file a 504 complaint. That's essentially equivalent
to "separate but equal". It's also worth noting that if you're not making
your PDFs accessible and your organization receives federal funding, you're
likely violating a number of laws.

Sorry for jumping on the derailment.

Jon Metz




On Fri, May 19, 2023 at 2:25 PM Metzessible <metzessible at gmail.com> wrote:


> I think this auto-tag "feature" is specifically intended for use with

> Adobe Experience Manager, as that's what uses Adobe Sensei the most.

> They're calling it "Smart Tagging (link:

> https://business.adobe.com/products/experience-manager/assets/smart-tagging.html).

> The caption for that photo is for a File photo, so chances are it's in the

> metadata of the file. I haven't inspected it myself.

>

> On Fri, May 19, 2023 at 12:44 PM <info at karlencommunications.com> wrote:

>

>> I just read a post on LinkedIn by Matt May, the Adobe Accessibility lead.

>> He is not well and taking time to recuperate.

>>

>>

>>

>> Not sure what the status of accessibility is in his absence.

>>

>>

>>

>> Cheers, Karen

>>

>>

>>

>> *From:* athen-list <athen-list-bounces at mailman12.u.washington.edu> *On

>> Behalf Of *Philip Kiff

>> *Sent:* Friday, May 19, 2023 11:05 AM

>> *To:* athen-list at u.washington.edu

>> *Subject:* Re: [Athen] [EXT]Re: PDF remediation help

>>

>>

>>

>> Another quick follow-up on this.

>>

>> I just tested the Demo using a two-page PDF newsletter generated by Canva

>> - which generally produces absolutely horrendous PDF files. And I tested

>> the AutoTag output from my version of Adobe Acrobat Pro and compared it to

>> the output from the Demo API.

>>

>> First thing to note is that I am not given the option of using Adobe

>> Cloud for AutoTag when I attempt to use the AutoTag feature in Adobe

>> Acrobat Pro. I selected not to use the cloud because I want to be able to

>> compare the outputs.

>>

>> Next thing to note is that the output from the AutoTag built in to

>> Acrobat has gotten much better recently. Even just a few months ago, it was

>> not worth my time to start remediating a Canva file with Autotag. But

>> today, the file it output on its own was a good starting point.

>>

>> Comparing the output from Acrobat Pro to the Cloud Demo, I notice a few

>> differences. One significant difference in this file is that the built-in

>> Acrobat Autotag does not correctly tag any annotations (hyperlinks).

>> Whereas the Adobe AutoTag API Demo does a pretty good job of generating

>> proper Link/Reference tags. Also, the AutoTag API Demo actually creates a

>> root <Document> tag like it is supposed to, instead of putting everything

>> inside a <Section> tag.

>>

>> I would say that if you have a choice, right now, and you don't mind the

>> loss of privacy/security with uploading files to Adobe, then I would

>> recommend using the AutoTag API Demo instead of using the built-in Adobe

>> Acrobat Autotag feature.

>>

>> Phil.

>>

>> On 2023-05-19 08:10, Philip Kiff wrote:

>>

>> A couple corrections and a re-interpretation after re-reading things this

>> morning.

>>

>> The media release was from May 17, the day before GAAD, not on May 18.

>>

>> Adobe isn't claiming they are going to release a stand-alone version of

>> the PDF Accessibility Checker . I must have heard that somewhere else, but

>> wherever I heard it, it doesn't sound like that's true based on the media

>> release. I'm not quite sure what the release is saying about the checker -

>> I think Adobe needs someone with clear language training to help write

>> their media releases!:

>> "The PDF Accessibility Checker, available in Acrobat via an API, will

>> enable organizations to quickly and easily evaluate the accessibility of

>> existing PDFs at scale."

>>

>> So the PDF Accessibility Checker will probably also be an API connecting

>> to a cloud-based function? And when you use it in Acrobat, it will use the

>> cloud-based API, instead of running locally on your computer using code run

>> by Acrobat. It's too bad that they are calling it "PDF Accessibility

>> Checker", since that will mean it gets confused with the per-existing free

>> PAC (PDF Accessibility Checker) tool.

>>

>> I guess we'll see.

>>

>> Phil.

>>

>> On 2023-05-19 01:31, Philip Kiff wrote:

>>

>> So I found Adobe's media release [1] from yesterday (GAAD, May 18) that

>> sparked the Forbes and other article (links at end of emal).

>>

>> Adobe has a page in their Developer documentation area [2] with more info.

>>

>> They have a free demo [3] (requires login) where you can test it out.

>>

>> I think that what Adobe are announcing is the new API aimed at

>> developers, not just a new iteration of the Auto-Tag feature that has been

>> built into Adobe Acrobat Pro for many years.

>>

>> Adobe seems to be suggesting that this API is partially driven by their

>> so-called "AI" branded as "Sensei". It is not clear to me that the AutoTag

>> feature in Acrobat Pro runs  the exact same code as that in their (new?)

>> API. Leaving aside the misleading marketing language around the phrase

>> "artificial intelligence", I have the impression that Acrobat Pro's AutoTag

>> feature depends entirely on code run locally on your computer, whereas the

>> AutoTag API will run somewhere in Adobe's Cloud. Maybe the results will be

>> similar. Who knows.

>>

>> Is that different than the AutoTag in the "new Acrobat" that you're

>> talking about Bevi?

>>

>> Like Bevi and Susan, I find the AutoTag feature currently included in

>> Adobe Acrobat Pro to be mediocre. Though I often still use it as a starting

>> point when I am remediating files.

>>

>> But I can see a value in having an API that developers can incorporate

>> into their own PDF generation processes or when trying to automate

>> large-scale remediation efforts on collections of files. There are already

>> some other document accessibility companies that offer automated solutions

>> for remediation of PDF files in bulk, I think, so I expect this signals

>> that Adobe has decided to enter into this business market more aggressively.

>>

>> There are two other items worth noting that appear towards the end of

>> Adobe's press release. This fall, they plan to release a stand-alone PDF

>> accessibility checker and they plan to include the AutoTag feature in the

>> free Adobe Reader instead of it requiring the purchase of the expensive and

>> overpriced Adobe Acrobat Pro. I'll be curious if the accessibility checker

>> will be a new product that checks against a full set of WCAG or PDF/UA

>> criteria to compete with offerings from CommonLook and PAC, or if it will

>> simply be a stand-alone version of the rudimentary Accessibility Check

>> feature currently built into Adobe Acrobat Pro.

>>

>> [1] Media Release:

>>

>> https://news.adobe.com/news/news-details/2023/Media-Alert-Adobe-Scales-PDF-Accessibility-With-Adobe-Sensei-AI/default.aspx

>>

>> [2] Developer AutoTag API page:

>>

>> https://developer.adobe.com/document-services/apis/pdf-accessibility-auto-tag/

>>

>> [3] Demo Sign Up:

>> https://acrobatservices.adobe.com/dc-accessibility-playground/main.html

>>

>> Phil.

>>

>> Philip Kiff

>> D4K Communications

>>

>> On 2023-05-18 18:45, chagnon at pubcom.com wrote:

>>

>> We’ve been testing the new Acrobat for a year or so, and keep testing the

>> AutoTag feature.

>>

>> It gets better with every new edition, but it’s been around for a few

>> years and we still can’t depend on it for much.

>>

>>

>>

>> It did do one PDF just ok, sort of: it was a one-page document with one

>> heading and several paragraphs of body text. Tagged the title at the top of

>> the page as <H5>, which from our testing seems to be the default heading

>> tag AutoTag uses for the first heading.

>>

>>

>>

>> Duh.

>>

>>

>>

>> It seems to be sales hoopla from Adobe in honor of GAAD day. Time will

>> tell if we actually get anything of substance from Acrobat.

>>

>>

>>

>> Sure wish Adobe would focus on making a more accurate PDF from MS Office

>> and InDesign so that we don’t have to remediate the PDF after it’s made,

>> but maybe that’s asking too much!

>>

>>

>>

>> *— — —*

>>

>> Bevi Chagnon  *| *Designer, Accessibility Technician* |*

>> Chagnon at PubCom.com

>>

>> *— — —*

>>

>> *PubCom: Technologists for Accessible Design + Publishing*

>>

>> consulting • training • development • design • sec. 508 services

>>

>> *Upcoming classes* at www.PubCom.com/*classes*

>> <http://www.pubcom.com/classes>

>>

>> *— — —*

>>

>> Latest blog-newsletter

>> <https://mailchi.mp/e694edcdfadd/class-discount-3266574> – *Simple Guide

>> to Writing Alt-Text

>> <https://www.pubcom.com/blog/2020_07-20/alt-text_part-1.shtml>*

>>

>>

>>

>> *From:* athen-list <athen-list-bounces at mailman12.u.washington.edu>

>> <athen-list-bounces at mailman12.u.washington.edu> *On Behalf Of *Susan

>> Kelmer

>> *Sent:* Thursday, May 18, 2023 4:53 PM

>> *To:* enews at toptechtidbits.com; Access Technology Higher Education

>> Network <athen-list at u.washington.edu> <athen-list at u.washington.edu>

>> *Subject:* Re: [Athen] [EXT]Re: PDF remediation help

>>

>>

>>

>> To quote from the article:

>>

>>

>>

>> “For example, before using the new Adobe API, a global financial firm

>> spent more than 9 hours to make each individual presentation slide deck

>> accessible. With Adobe’s new API, the firm has automated 70–80% of that

>> process.”

>>

>>

>>

>> Should we tell them?  Should we tell them that had they just created the

>> file correctly in the first place, it would have taken only a few minutes,

>> and no 9 hour remediation would have been required?

>>

>>

>>

>> NINE HOURS?   What the heck were they doing?  Oh yeah, they were using

>> Adobe to fix an Adobe file.  That works.

>>

>>

>>

>> /sarcasm

>>

>>

>>

>> *Susan Kelmer *

>>

>> Alternate Format Production Program Manager

>>

>> Disability Services

>>

>> Health and Wellness Services

>>

>> *T* 303 735 4836

>>

>> *www.colorado.edu/disabilityservices

>> <http://www.colorado.edu/disabilityservices> *

>>

>>

>>

>>

>>

>>

>>

>> *Due to the nature of electronic communication, the security of this

>> message cannot be guaranteed. If you’ve received this email in error please

>> notify the sender immediately and delete this message. *

>>

>>

>>

>>

>>

>>

>>

>>

>>

>> *From:* athen-list <athen-list-bounces at mailman12.u.washington.edu> *On

>> Behalf Of *Top Tech Tidbits

>> *Sent:* Thursday, May 18, 2023 2:49 PM

>> *To:* chagnon at pubcom.com; 'Access Technology Higher Education Network' <

>> athen-list at u.washington.edu>

>> *Subject:* Re: [Athen] [EXT]Re: PDF remediation help

>>

>>

>>

>> Here’s some additional information released by Forbes:

>>

>>

>>

>> Adobe Wants To Help Make PDF Documents More Accessible With AI-Based

>> Auto-Tag API | May 17, 2023 | Steven Aquino | Forbes

>>

>>

>> https://www.forbes.com/sites/stevenaquino/2023/05/17/adobe-wants-to-help-make-pdf-documents-more-accessible-with-ai-based-auto-tag-api/?sh=5d0d4855b24c

>>

>>

>>

>> *Aaron Di Blasi, PMP*

>>

>>

>>

>> Publisher (2020-Present)

>>

>> *Top Tech Tidbits*

>>

>> The Week's News in Adaptive Technology

>>

>> https://toptechtidbits.com 🌐

>>

>>

>>

>>

>>

>> *From:* chagnon at pubcom.com <chagnon at pubcom.com>

>> *Sent:* Thursday, May 18, 2023 2:09 AM

>> *To:* 'Access Technology Higher Education Network' <

>> athen-list at u.washington.edu>

>> *Cc:* enews at toptechtidbits.com

>> *Subject:* RE: [Athen] [EXT]Re: PDF remediation help

>>

>>

>>

>> Hi Aaron,

>>

>>

>>

>> I don’t know why the article is being posted as news. It’s grossly out of

>> date and is from the Microsoft news network portal.

>>

>>

>>

>> Adobe's New AI Automates PDF Accessibility Tags

>>

>>

>> https://www.msn.com/en-us/money/other/adobe-s-new-ai-automates-pdf-accessibility-tags/ar-AA1biX4V

>>

>>

>>

>>    1. The photo captions reference software releases in 2013, “Adobe

>>    announced Monday, May 6, 20*13*…” and the rest is pure nonsense.

>>    2. The AutoTag feature has been in Acrobat Pro for quite a few years.

>>    It’s had marginal improvement every year, but overall still sucks at

>>    correctly tagging a document, separating text from graphics, and creating a

>>    logical reading order. We teach our clients and students how to use it only

>>    when there’s no better way to remediate a bad PDF.

>>

>>

>>

>> We’re beta testers for Adobe and I don’t see anything substantial on the

>> horizon for Acrobat anytime soon, other than an overhaul of the user

>> interface (which is getting bad reviews). We still have hope for the

>> future, however.

>>

>>

>>

>> I’m chalking up this news release as stale and inaccurate.

>>

>>

>>

>> *— — —*

>>

>> Bevi Chagnon  *| *Designer, Accessibility Technician* |*

>> Chagnon at PubCom.com

>>

>> *— — —*

>>

>> *PubCom: Technologists for Accessible Design + Publishing*

>>

>> consulting • training • development • design • sec. 508 services

>>

>> *Upcoming classes* at www.PubCom.com/*classes*

>> <http://www.pubcom.com/classes>

>>

>> *— — —*

>>

>> Latest blog-newsletter

>> <https://mailchi.mp/e694edcdfadd/class-discount-3266574> – *Simple Guide

>> to Writing Alt-Text

>> <https://www.pubcom.com/blog/2020_07-20/alt-text_part-1.shtml>*

>>

>>

>>

>> *From:* athen-list <athen-list-bounces at mailman12.u.washington.edu> *On

>> Behalf Of *Top Tech Tidbits

>> *Sent:* Wednesday, May 17, 2023 5:52 PM

>> *To:* 'Access Technology Higher Education Network' <

>> athen-list at u.washington.edu>

>> *Subject:* Re: [Athen] [EXT]Re: PDF remediation help

>>

>>

>>

>> Per recent discussions, this article was just released 8 hours ago. I

>> think we should expect a number of updates from Adobe regarding PDF

>> accessibility, remediation and AI in the coming months.

>>

>>

>>

>> Adobe's New AI Automates PDF Accessibility Tags

>>

>>

>> https://www.msn.com/en-us/money/other/adobe-s-new-ai-automates-pdf-accessibility-tags/ar-AA1biX4V

>>

>>

>>

>> *Aaron Di Blasi, PMP*

>>

>>

>>

>> Publisher (2020-Present)

>>

>> *Top Tech Tidbits*

>>

>> The Week's News in Adaptive Technology

>>

>> https://toptechtidbits.com 🌐

>>

>>

>>

>> _______________________________________________

>>

>> athen-list mailing list

>>

>> athen-list at mailman12.u.washington.edu

>>

>> http://mailman12.u.washington.edu/mailman/listinfo/athen-list

>>

>>

>>

>> _______________________________________________

>>

>> athen-list mailing list

>>

>> athen-list at mailman12.u.washington.edu

>>

>> http://mailman12.u.washington.edu/mailman/listinfo/athen-list

>>

>> _______________________________________________

>> athen-list mailing list

>> athen-list at mailman12.u.washington.edu

>> http://mailman12.u.washington.edu/mailman/listinfo/athen-list

>>

>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman12.u.washington.edu/pipermail/athen-list/attachments/20230519/af1417fc/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 8916 bytes
Desc: not available
URL: <http://mailman12.u.washington.edu/pipermail/athen-list/attachments/20230519/af1417fc/attachment.png>