What’s Next for PDF
A couple of months ago I started hearing rumors about a new version of PDF, often mysteriously referred to as PDF “next.” As it happened, the PDF Days Europe 2017 conference in Berlin on May 15 had an entire track dedicated to this topic. Could PDF.next be the easy-to-output-to, easy-to-distribute, low-cost digital publishing format that so many of us are seeking? I decided to go to Berlin to find out.
As soon as I arrived, I was quickly reminded of the diverse market that PDF must serve. This was a hardcore technical conference with sessions on accessibility, PDF as a service, PDF/X, PDF/A, PDF/UA, and much more. But I (and many others) were there to learn about PDF.next!
Happy quarter-century, PDF
First, it’s worth remembering that the PDF format was invented 25 years ago, when laptops were in their infancy, and cell phones didn’t even have screens. PDF was all about reliable printing and exchanging documents between Macintosh and Windows computers.
Today, people want to access their information “from watch to wall”—on screens large and small. Mobile phones now generate over 50% of web traffic. There are many people whose only computing device is a phone. I think we can all agree: for small screens, PDF is not a good format. At best on small screens PDF forces us to pinch and zoom in order to read. Furthermore, PDF is “where data goes to die.” The original document may contain all kinds of rich tables, charts, and spreadsheets, but once it is exported to PDF, only the visual appearance remains. There is no way for readers to sort table columns, view alternate graph types, or change data ranges.
PDF Days 2017 was organized by the PDF Association, which exists to “promote the adoption and implementation of International Standards for PDF technology.” After nine years of work, they are about to submit the specifications for PDF 2.0 to the ISO.
PDF 2.0 is an evolutionary, not revolutionary update to the PDF 1.7 specification that has some new capabilities that will eventually be useful for printing, color management, and accessibility. You can learn more about PDF 2.0 here.
But the real juicy talk at the conference was about PDF.next, which leverages some of the new features in this updated specification, but lies “beyond” PDF 2.0.
PDF.next is all about making PDFs relevant in the era of mobile devices with small screens. What if a PDF could reflow and adapt to these screens, presenting alternate views depending on the type of device or size of screen the viewer is using? In other words, what if a PDF could be “responsive”?
You’ve probably heard of “Tagged PDF,” and you may know that with enough patience and skill you can create a tagged PDF from InDesign. Today, the primary reason for doing so is to make the file accessible to people with vision impairment, since screen-reading devices need this tag information to specify reading order, information hierarchy, and alternate text that describes images.
The idea behind PDF.next is to add more complete, accurate tagging (often referred to as “semantic information”) to PDFs so they can be transformed on-the-fly into reflowable, responsive HTML for reading on small screens.
Maybe someday in the future, when a user opens a PDF.next document with an old reader or browser, they will see the pixel-perfect page rendition they are used to now. But when a user opens the same PDF in a modern PDF reader on a mobile device, the reader software will use the tags and structure in the PDF to transform that information into HTML. CSS will be supplied either by the document creator or by the PDF reader software, so that the HTML reflows and looks great on the smaller screen.
The benefits of PDF.next
In this scenario, the user would get all the benefits of a traditional PDF: an easy-to-distribute, document-centric, single file that is backward compatible, easy to print, and can be validated, signed, and protected if desired. These would be combined with the read-anywhere, accessible, responsive benefits of HTML and CSS.
Another interesting idea is embedding the data for a chart or table in a PDF. Someday, in PDF.next, that data could be referenced and called upon when the HTML/CSS representation of the file is viewed. So a chart that appears static in the PDF page rendition could come alive in the HTML rendition with sortable table rows, range sliders, or interactive charts.
Adobe and the future of PDF
Keep in mind that Adobe no longer “owns” the PDF format. As of 2008, PDF is an ISO standard. So while Adobe is certainly interested in the development of PDF, they are no longer the sole driver of future development. They are only one of many players. Leonard Rosenthal from Adobe is on the PDF Association board, and he represents Adobe in the ISO. (Leonard will be on hand at PePcon: The Print + ePublishing Conference 2017 to discuss the future of PDF.)
When are PDF 2.0 and PDF.next coming?
PDF 2.0 should become an ISO standard within the next few weeks. (Remember that took about nine years to ratify!) But the rest of this stuff (the PDF.next stuff) is a long way off. There is a lot that needs to happen before we see the benefits. At the very least, we need robust tools that help us accurately tag InDesign files with the proper tagging structure. Then, we need PDF reader software that can transcode those tags into HTML and display it on small screens. Many of us will need education in creating structured documents from InDesign and editing CSS.
All of these things may or may not happen. Leonard Rosenthal says, “we are taking our time and doing this right.” If all these pieces fall into place, we could have the best of both worlds: robust PDFs that adapt to any type of reading situation. Stay tuned!