Adobe releases PDF standard for ISO certification

On the 29th January Adobe announced it will be seeking ISO certification for the PDF 1.7 standard. Up until this point many vendors have been able to implement the PDF standard based solely on trust that Adobe will not significantly change the format and break their respective implementations. If PDF gains ISO certification then this will ensure any vendor can develop for and use the standard in the knowledge that it will not change and be 100% interoperable with other implementations of the same standard.

This is good news for governments and business as it means once certified PDF will become a permanent, unencumbered format. These characteristics will enable organisations to use PDF as an archiving medium for 2-dimensional digital documents with the confidence that no single company can dictate or control the use of the format. This is important because in the past companies like Microsoft have unduly effected the industry with their monopolistic control over formats (as evidenced by the infamous Halloween Memo).

Microsoft and Autodesk will not be too pleased at this development as both parties have formats which are competing against PDF (XPS and DWF) for user acceptance. What is not so clear from the announcement is whether Adobe's ISO ambitions cover their newly announced 3D extensions to PDF or just the well entrenched 2D format. My guess would be this week's announcement only covers 2D as 3D is no doubt still under constant development. Either way it adds pressure to Autodesk who are really pushing DWF hard as the most important medium for 2D and 3D documentation in the Architecture, Engineering and Construction space.

Deelip Menezes posted the ISO news on his blog and he got an interesting reaction from Brian Mathews of Autodesk Labs. In the comment he argues that DWF holds the advantage because it is based upon Microsoft's XML-based XPS format which is more clearly documented and more 'open' because it is in XML. This "XML good - binary bad" line of thinking ignores the fact that XML by design is significantly slower and heavier than a binary format which limits its applicability. Also it is important to remember that a format's clarity of documentation is entirely dependent on the efforts of the format creator, something Microsoft has proven notoriously bad at.

Brian also raises another point:

"PDF has always been an open specification, but has been based on a very complex binary format that takes 1000's of pages to fully document (with lots of external reference documents)."

Out of interest I downloaded the Adobe PDF specification from their website. Yes it is 1,310 pages long but this includes almost 200 pages of index and 100 pages of examples. That leaves a round about 1,000 pages devoted to the actual standard and from a quick click through I found the sections well laid out and in parts quite an interesting read (well as interesting as a standard can get). Whilst 1,000 pages is a lot there is a lot to cover and I would rather see a format span 10,000 pages and be understood in depth rather than glossed over in 500. Also another yardstick by which a good standard can be judged is how often it is independently implemented and for PDF this number is quite significant, just checkout Wikipedia's rundown.

The other issue Brian Mathews raised was Adobe's XML-based PDF format codenamed Mars:

"Adobe followed recently with an announcement that they too will be developing their own fully XML alternative to PDF. Why? Because PDF isn't open enough since it isn't XML."

Reading through the intentions behind Mars it becomes apparent that it is not so much a "PDF replacement" as Brian implies but rather a complimentary format. Why? Binary formats are difficult to index by search engines or pull data out of without pre-processing. Mars is not a rewrite of the PDF format but rather an implementation of the PDF standard using XML syntax (after all, XML is just a language syntax). This format enables quick, low overhead indexing and data manipulation at the cost of end-user speed. As a consequence Adobe are pushing this as a complimentary format and certainly not as a replacement to binary PDF. Employing complimentary file formats is not unusual, for example the Industry Foundation Classes come in both binary and XML versions to suit specific use cases.

What this all this means in the long term is still uncertain but it certainly looks like the competition is heating up.