PDF Readers: Getting judgemental
During FSFE’s recent campaign to stop governments advertising proprietary PDF readers, a common question was raised by the public in emails and comments. “Can Free Software PDF readers match up to Adobe Acrobat Reader?” they ask. The implication of this question is that promoting alternatives to Adobe, whether they be Free Software (FS) or not, is only reasonable if they are ‘as good as’ (e.g. file compatible and feature feature identical with) Acrobat Reader.
This view undermines the PDF standard however, and falsely defines what PDF is in terms Adobe’s implementation.
Although PDF is an ISO standard, how PDF is used is not standardised. Thanks to non-conforming proprietary PDF editing software, the Internet is awash with PDF files which do not conform to any version of the ISO PDF standard. When FS PDF readers cannot open these non-standard files, they are perceived as inferior.
This is particularly so in the case of PDF documents supplied by government organisations. Citizens are understandably frustrated when they cannot properly view important information from their state, submit tax, or registration information using the PDF forms that they have been provided with. If switching to Adobe’s reader solves their problems, then FS readers are blamed. Nobody asks about whether the file is valid PDF.
So how can FSFE encourage wider use of FS readers in good faith when the expectation is that PDF documents will work more reliably with Acrobat Reader?
The acronym PDF means something specific. It is a file format, and it is amongst a tiny number that has been formally standardised. PDF means the PDF standard. When companies release PDF software that intentionally generates files that do not conform to that standard, they are distancing themselves from PDF. This is exactly what Adobe is doing each time it releases a new version of Acrobat with extended functionality not supported by the PDF standard.
Corrupting standards is a long established practice, and popular amongst big software players. It is a shame however when governments unintentionally support the corruption of standards by providing documents with a .pdf extension which do not conform to PDF standards. When a state employee creates a new file in Acrobat Pro, no warning messages will appear to alert them to the fact that they are about to break the internationally agreed definition of PDF.
When a user downloads such a file, there is no immediate way of telling whether it conforms to standards either. Many naturally assume that because the file is .pdf, it is a valid document in PDF format, and that any reader that does not open it correctly must therefore not be a good PDF reader.
A far more accurate way to assess the success or failure of a PDF reader however would be based on the extent its implementation of PDF standards. It isn’t reasonable to judge an application based upon how well it works with something that it wasn’t designed to accommodate – namely a non-standards compliant PDF file.
Even Adobe’s reader fails when measured by such unfair criteria – Acrobat can’t open non-compliant files any better than FS readers can when they don’t have the advantage of having generated them themselves. In such cases the company ironically becomes the victim of their own game: on more than one occasion I’ve had to assist a friend in opening a PDF document using Free Software because Acrobat was unable to do so.
What is less obvious is that we should take a similar approach to PDF files. When FS readers are criticised because of incompatibility with non-compliant PDF documents, the fault of the file should be highlighted, both to the user, and to the organisation from which the publication originated.
Therefore the question should not be “Can Free Software PDF readers match up to Adobe Acrobat Reader?” but rather “do any Free Software PDF readers fully implement ISO PDF standards?”, and in cases of specific documents: “does this file conform to ISO PDF standards? If not, why not?”.
Adobe’s implementation of PDF is far from perfect. If we judge FS implementations of PDF according to their similarity to Acrobat Reader then they will always come up short. Instead of chasing Adobe’s definition, we should all be chasing ISO’s definition, and this includes the governments who publish documents in PDF.