The awful Accessibility of PDF Documents

Are PDF's accessible? One heretical question and my desktop publishing friends would smack their expensive Macs in my face. In my opinion, however, the question is justified.

Article Content

PDF is not designed for accessibility

From its foundation, the web is designed to be accessible. The fact that this is not always implemented is another matter. If you haven't locked the right mouse button, anyone can look into the source code of a website, adapt the code to their requirements and thus fix many errors that disturb accessibility. And it can be that if you use the appropriate content management systems or libraries, you create an accessible website almost by accident and without intention.

PDF is just the opposite: you can't see the code without special tools. In terms of its basic principle, it is not accessible. And it is not possible to create accessible PDFs on the side. If no one does the work or is aware of the need, the resulting PDF will not be accessible.

Semantics must be grafted on

Historically, there is no semantics in PDF: text structures, image descriptions, machine-readable formatting do not exist in PDF and consequently in most authoring tools. This also makes it difficult to integrate accessible PDFs into the publication process. To date, the accessibility features in InDesign or Acrobat Professional are separate from the actual editing tools. If you create an ePub with InDesign and have no idea about accessibility, you will get a non-accessible document.

Too expensive, too complicated, too closed

It is comparatively expensive, time-consuming and demanding to create accessible PDFs. Let's just compare the procedure with a comparable format: ePub.

ePub can be made accessible with any text editor. Those who like it more graphically can work with Calibre or Sigil. Adobe Acrobat Pro costs around €600, the ePub programs cost nothing.

Adobe Acrobat and InDesign requires a significant learning curve for desktop publishers. We're talking about persons familiar with these tools. For persons who are not DTP professionals, the effort is significantly higher. And hat is for something that word processors or ePub editors do almost casually: structuring texts, creating image descriptions, a reasonable text flow... I could train everyone in creating accessible ePub files - but nobody is interested in it.

In addition, the creation process, especially with Adobe Acrobat itself, is very error-prone. Even on poorly programmed websites, I don't see as much crap as I've seen in accessible PDFs made by professionals: There are spaces in the middle of words, tables are a mess, graphic elements are not sufficiently hidden from screen readers, and so on.

As an aside, there is one oddity: PDF explicitly allows blocking assistive technology access to the content. Take a look in Adobe Reader under File - Properties - in the Security tab. This certainly has something to do with some DRM bullshit. But Adobe and the PDF Association should be ashamed of allowing such crap.

PDF is now an ISO standard. But it is not open because it is a binary format. On the other hand, even the Microsoft formats are a model example of openness. ePub is completely open and maybe that's why it's underestimated. I doubt it will be possible to open a PDF-A file in the distant future.

Accessible PDF is a money-burning machine

If you look at the woman power, the time and money that is invested in accessible PDFs, you have to come to one conclusion: Accessible PDFs are a huge waste of resources. That would be justified if there would be significant added value. But it usually isn't.

This money is then no longer available for other measures such as sign language or plain language. In an optimal world, we would have enough resources for everyone. In our imperfect world, resources are as finite as my patience.

Adobe has done too little

Adobe makes a lot of money from the boom in accessible PDFs. But it has done nothing itself to give the topic more impetus.

So far there has been no cost-effective solution for integrating PDFs into automatic document workflows. A library that could be integrated into technical processes would actually make the process easier. Or server-based routines that automatically convert websites into accessible PDFs. Or a plugin for the most common editorial system WordPress. Or an AI that makes it easier to create accessible PDFs with Acrobat. Absolutely nothing comes from Adobe, and why should they, they make a lot of money from Acrobat and the licenses.

PDF-UA – a closed society

As great as the accessible web scene is, the accessible PDF community is ugly. It starts with the fact that one cannot really speak of a community, but of dozens of individual fighters.

This is why little or no know-how about accessible PDFs is freely available on the Internet and free of charge. This also reflects the spirit of the PDF community: anyone who has found a good solution keeps it to themselves. This makes it difficult for beginners to familiarize themselves with the system and to search for errors.

Adobe is to blame

For once, the culprit for this situation can be clearly identified: his name is Adobe and the PDF UA Foundation.

The Adobe company has overslept the topic of accessible PDFs for years, picked it up too late, implemented it half-heartedly and miserably. Even professionals do not use Adobe's pure functions, but use additional tools and extensions. Anything Adobe says about accessibility and inclusion can safely be dismissed as advertising babble . When it comes to digital accessibility and accessible documents, Adobe is the stumbling block and we would be better off without them.

The failure of the PDF Association

The PDF Association appears to be an Adobe appendage. The information they provide about accessible PDFs is confusing and actually irrelevant in practice.

An example is the company PDFlib. It praises themselves for their tool, which supposedly produces accessible PDFs. Neither the PDF UA whitepaper nor the documentation for the tool is accessible. So the tool can't be that great.

PDF is a historical misstep

PDF owes its raison d'être mainly to the fact that it is platform-independent. As a template for print products, it definitely has its advantages.

For other purposes, it's no doubt fine. Office documents harbor a security risk, ePub viewers are not integrated in all systems. And many documents are actually printed out and only briefly viewed on the screen for this purpose.

However, PDFs are not suitable for the Internet. Brochures are put online as PDF because they are there anyway as a waste product of the print process. Only a few apps and programs can optimize multi-column documents in such a way that they are displayed reasonably and look good on a smartphone. I myself have not yet managed to read an accessible PDF on a smartphone. But these are the devices that are primarily used today in the private and also in the professional sector.

Incidentally, it remains a mystery to me why the bad habit is still maintained. The multi-column structure typical of large-format brochures may have its advantages in the print sector. On digital devices, multi-columns offer no advantage, even on large screens. How wide a text column is displayed can be set by the width of the program window or with the reading program. The main advantage of digital reading is actually that you can adjust all factors such as font, size, contrast and so on to your preferences. Exactly these advantages should be undermined by PDF in the sense of the inventor. Conversely, nobody would think of printing out a long text on a DIN A4 page and only ever printing on the left half of the paper.

So PDF is almost as dead as Flash, or at least it should be as far as the web is concerned.

Other articles