06 juin 2018 By Dietrich von Seggern
Coming back from the recent PDF Days (in my opinion the best one ever), I was thinking about the post presentation discussion to my overview on PDF based ISO standards (PDF/X, PDF/A, PDF/UA...). I covered the 6 subset standards using PDF/ in their names, PDF itself and some more standards that are closely related such as "Processing Steps" or "CxF". Some of the attendees seem to have been confused by the amount of different standards. Now you can blame the presenter, but in any case I guess it is a legitimate question why we have so many different PDF based standards and how difficult it would be to create a PDF that complies with more than just one of them.
We heard of this requirment first when PDF/A-1 came out in 2005 when users approached us about whether and how we can make it possible to convert a PDF to PDF/X (for professional print workflows) and PDF/A (for archival) at the same time. During the PDF Days, people where interested in converting to PDF/A and PDF/UA (for accessibility).
ISO definitely did not introduce that many standards to confuse people. It is rather so that requirements of the print industry are quite different from archival. E.g. the print color space (CMYK) uses four values while the - therefore smaller - definitions in RGB only use three. Accessibility has demanding requirements on the creation process which should not be applied to every PDF/A file. PDF/A on the other hand has certain requirements on metadata that you would not want to apply on PDF/X or PDF/UA files (since there is no need for easy identification and retrieval). So while it is understandable that people are asking for a less confusing structure, there are good reasons why there are different standards for differnt purposes. In the end it boils down to the fact that ISO wants to make it as easy as possible to create PDF files that comply to a certain standard and therefore defines specific requirements for specific use cases.
What the ISO commitees, however, are about to do is to harmonize the standard texts. E.g. the provisions on fonts are in essence the same in all subset standards. In the current versions the wording and some details are different, which makes their implementation unneccessarily difficult (if you want to support more than one standard). To be honest the goal of ISO is to make their maintenance of the standards easier, but that will at the same time make it easier to create PDFs that comply to several such subset standards; or I should rather say to implement applications and workflows that create PDFs...
The good message from me, representing callas software, is that in our pdfaPilot und pdfToolbox products as well as in Acrobat Preflight it is easy to create a Profile that validates against several standards or that converts input files into more than just one subset standard at the same time . This can be done by setting checkboxes and selecting the respective standards in pop-ups and here in our online documentation is an article that describes that in detail. The products also internally put requirements into right order where needed: E.g. if you convert to PDF/X and PDF/A you want the more demanding standard to have priority (e.g. to keep any CMYK color definitions in the original) or for PDF/A and PDF/UA to not invalidate the tagging structure. Our products are smart enough to make proper decisions in such cases: to keep as much as possible from the original PDF and in case of conflicts to use the requirements of the more demanding standard.