The ISO SC 34 meeting here in Korea has been sweetness and light so far. Contrary to Groklaw’s claims, Microsoft has not attempted to prevent ODF by underhand methods AFAICS. Contrary to Gartner, it looks like Open XML will proceed through ISO fast tracking to national vote without incident too AFAICS. ODF has gained a lot in reputation by its ISO standardization and raised the bar. Open XML will similarly gain a lot by reaching or surmounting the same bar. Everyone’s a winner, like the Hot Chocolate song. (I am sure Sun/Microsoft/IBM hate having to go through the mechanics and uncertainties of standardization; I am sure many in SC34 don’t like standardizing schemas at the SC34 level, contrary to long-running policy, especially with the low-review, fast-tracking and PAS loopholes. But it is good for us.)
A nice phrase came up yesterday: ISO standardization of an existing standard represents a second round of openness.
I wish it were always true. Unfortunately, the ISO PAS and Fast-tracking procedures don’t really require much in the way of substantive feedback. ODF, for example, will change in no substantive way in its ISO adoption. National body comments will be added to requests or requirements for future versions. The Ecma Open XML people, so far, are being far more concilliatory in this regard: they know that a Microsoft technology doesn’t have the presumption of innocence that a Sun format does, in the minds of many.
If Microsoft/Ecma/et al manage to demonstrate to the ISO member voters that Open XML had even a first round of openness at Ecma, that it has some different use from ODF, if it supports SC34 specs like RELAX NG, and is scrupulous in its partitioning of Windows-specific hooks to another layer or namespace, I don’t see any national body rejecting Open XML, frankly. Microsoft and Ecma still have work to do in this regard, but it is just the standard kind of technical-level education/discussion/wordsmithing/re-alignment that any specification should have.
ODF and Open XML standardization is an important issue because of the Mass. policy on standardization. More power to Mass! If they push everyone to commit to exposing their file formats, that is great. I am sure Microsoft knows that they have a great business opportunity in making office documents a rich media that integrates into media/information systems.
What should procurement officers do? Well, for a start, remember that ODF is not an ISO standard yet. The final text, rearranged for ISO requirements, will take until August or so to get through the ISO machine. Open XML will presumably start its way into the ISO machinery later this year. ODF still has extra parts to be added, but its main part has been finalized; Open XML is broadly complete but not submitted yet. So even though ODF is first through the gate, if they both might be completed and standardized at around the same time. There is even a possibility, I suppose of ODF adopting Open XML formulas for spreadsheets. ODF currently uses SMIL, a Microsoft-associated technology, which Open XML does not. Funny old world.
I’ve been distracted. What should procurement officers/information architects do for policy in this regard? Well, for a start, mandating that organizations move to an XML or XML-in-ZIP file interchange./archiving policy sooner rather than later. Most importantly, create additional standards (with checking validators) to make sure that everything is not subverted by the use of platform-specific binary formats, embedded objects, macros, and so on. Simply saying only ODF or only Open XML or only ODF and Open XML is not enough, though it is a start. Second, clearly distinguish your different needs: sometimes you need page fidelity (such as for the electronic versions of documents whose paper version has legal status), sometimes you need interchangeability where it you want to guarantee that no platform-specific features are used. It may well be that ODF is appropriate in one situation while Open XML is useful in another, in the same way that HTML is useful precisely because it can be (badly) formatted on lots of different media.
Whatever you do, take statements like “ODF is slow” with a grain of salt. First, you might ask “I thought the MS EULA didn’t allow published benchmarks without permission?” How can I trust any benchmark if I we cannot find out about one-sided results? Then you might say “Surely it is applications that are fast or slow, not files, especially in the early days of a format?”
Remember that for large scale workflows and information systems, the more that information is labelled with computer (and human) accessible names and metadata, the more useful it becomes. ODF and Open XML don’t provide this kind of information. In a spreadsheet, you have to look up elsehwere the information to know what a cell value’s label is: that indirection adds inconvenience to using the information. So ODF and Open XML are progress, but progress in the area of how to make boring, unlabelled information more accessible: easier to access but not to a semantic treasure trove.