In the recent accusations that the GNOME Foundation has been supporting Microsoft's OOXML format at the expense of ODF, KDE has been presented as a counter-example. Based on a KDE News article, Richard Stallman suggested that "major KDE developers" had announced "their rejection of OOXML" and urged GNOME to do the same. More recently, a widely linked story on ITWire used the same article to declare that KDE has taken a "principled stand" against OOXML. However, if you go the source, the story is more nuanced than these claims suggest.
True, KOffice -- not KDE, which is a separate project -- will not be implementing OOXML in a hurry, and part of the reason is political. However, the decision is also a practical one, and KOffice has not rejected the possibility of supporting OOXML in the future.
David Faure has been a KDE developer for nine years, and a KOffice developer almost as long. He has worked on Konqueror, KDesktop, and many of the KDE core libraries. In KOffice, he has focused mainly on the KWord word processor. In addition, Faure has sat on the OASIS technical committee for ODF since 2003, where he represents KOffice's concerns, and helped steer through modifications required by KWord, such as support for frames repeated on each page and a desktop publishing mode.
Few developers, you may safely say, are stronger advocates of ODF than Faure. Speaking unofficially for his fellow KOffice developers, Faure says, "What we like about it is that it was designed from the start to be independent of the design of the application, like HTML or XML. It is also based on many existing standards."
By contrast, Faure is scathing about OOXML. "Microsoft is pretending to everyone that they have an open format. What they want is to convince companies and governments that their format is open, so people will use the format in the future. But in reality, as you can see in many postings on the Web and on blogs about the technical problems with that format, OOXML is just as closed as [Microsoft's previous binary] format." He says that the complexities of the format mean that only Microsoft Office is ever likely to be able to implement it completely, adding bluntly, "It's all just make-believe."
Yet, despite these strong opinions, in talking about the decision not to support OOXML, Faure says, "The main reason is not political. The main reason is lack of resources."
Faure explains that, for the past year, all of the project's energies have been devoted to KOffice 2.0, a massive rewriting of the entire office suite. "That means that support for other formats comes later," he says. "We need first to have a working application before we can import files."
Under these circumstances, trying to develop support for OOXML would be almost impossible. "It is a very complex format," Faure says. "Its type of XML is almost as complex as the old binary format. And with the old format, the old .DOC format, we only ever managed to get to the point where we could import text and its formatting, but not images." Besides, he says that he has seen "a large number of deficiencies" in OOXML, leading him to conclude that "it will be very difficult to implement. We might start at some point, but who knows when we would finish?"
That said, contrary to the impression left by many media reports, KOffice has not ruled out eventually supporting OOXML. "But it has to come from user demand," he says, noting that user demand was why KOffice tried to support earlier Microsoft formats. "I haven't seen many user requests for OOXML just yet. It would be completely illogical for us to work on that format right now just because Microsoft says it is the future and before users demand it."
At the same time, Faure admits that these practical concerns have "a good political side effect. If we had good support for [OOXML] now, it would be an impetus for people to switch to that format. And that's exactly what we don't want to happen. So you could say that it's a political game for us. By not starting that format just yet, we try to make people switch to ODF, because that's a much better format for the future.
"If Microsoft really wanted to do something open, something good for people, it would switch to ODF and help extend it for whatever they need instead of trying to do their own thing, which defeats the purpose of having one standard. The whole idea failed when Microsoft decided to stay in its own game instead of joining us. That's my main regret in all of this, and I hope that more and more people support ODF instead."
Faure is only one developer, and does not officially speak for the entire KOffice team. However, he is a prominent one, and a recent blog entry by Cyrille Berger, another KOffice developer, shows that that he is not alone in his views.
KOffice developers like Faure and Berger deserve credit for balancing practical and political concerns in the middle of the OOXML controversy. Moreover, Faure in particular, has worked five years to improve the ODF standard. In both these ways, KOffice in general stands in marked contrast to the GNOME Foundation, which has ignored the politics of the issue in favor of the practical by rushing to support OOXML after only recently become involved with the development of office format standards.
However, neither of these comparisons justifies the idea that KOffice -- let alone KDE -- has shown a high-mindedness that GNOME lacks. In fact, the idea seems a distortion of at least some of the opinions prevalent among KOffice developers.
Both KOffice and GNOME have been navigating the same shoal-ridden waters. The difference is that KOffice -- perhaps because of greater knowledge due to Faure's involvement with OASIS -- has managed to avoid running aground, the way that GNOME appears to have done.