The state of the art?

By Rick Jelliffe
September 11, 2008 | Comments: 1

An interesting comment from document interop specialist Dr Peter Sefton, who knows his stuff:

As I've covered here many times ODF interoperability between applications is basically non-existent except between Microsoft Office and and its derivatives where some things work quite well. Bottom line is, ODF doesn't have any formal notion of what's conformant - it's up to application developers to implement the bits they feel like implementing.

From various other items on the web, I would say that you would be able to say exactly the same sentence but substitute OOXML for ODF.

The problem is that the bottom line for document interoperability is not the format, but the feature match of the applications. The only way ever to get reliable, bottom-line interchange (enough fidelity that no semantics are lost, with graceful degradation) is by restricting feature use.

An old trick we used to use with text processing in the pre-XML days was to deliberately save the document in some known-to-be-horrible format, then re-open, check, then save again in the high quality format. This helps with feature matching.

In modern times, you might consider first saving the document as HTML (or RTF! or even structured PDF!) and sending that *as well as* your ODF or OOXML version, if it is mission critical to have maximum reach.

The tide is certainly rising, though but unless we legislate (I am not serious here) for a global new-feature freeze on all office applications, the feature-match issue will never go away. It cannot.

Far from being signs of failure, the horrible dialog boxes that tell us, when we save, "Some features may be lost" are the way it should be.

Maybe mandating any base-line format, unless it is a horrible kitchen sink format which will have extreme coverage problems, necessarily involves saying "We are happy with some kinds of information loss."

You might also be interested in:

1 Comment

Up down translation is one approach. The other is instead of blind exchange, give them vision and use an object model as the conformance test. Then true interoperability instead of portable data becomes the goal.

It seems after this much time, we shouldn't have to say that anymore. I think at the technical heart of the ODF/OOXML bitter butter battle is knowledge by the experienced that format is not an effective means to achieve interoperability, that it is just applications sending each other mail in standard envelopes. Thus there is not only no realistic chance that ODF and OOXML will converge soon, there is no realistic requirement.

So all the frustrated are left with is bats to hit each other with because the technical truth, the real elephant in the room is, these are two different applications for the same job and the minority application is attempting to legislate dominance over the majority application.

But reality has never been a requirement for the web. It's been a Panglossian affair from its inception.

News Topics

Recommended for You

Got a Question?