Can We Avoid the Great Schism?

Choice is an important element of free software, so it's perhaps no surprise that even at the level of the desktop environment there is more than one offering. But the main alternatives – KDE and GNOME – represent more than just a way of placing icons on a screen. Nowhere is that more evident than in their respective views on Microsoft's OOXML document standard, which are very far apart – perhaps dangerously so.

Here's what the KDE side says:

Recently, the ISO standardisation process of OfficeOpenXML ("OOXML") has gained a lot of public attention. What are the implications of OpenXML as ISO standard next to ODF for Free Software applications?

The standardisation process of OfficeOpenXML has turned sour, not in the least because Microsoft couldn't resist the temptation to cheat. Right now we're seeing evidence of a concerted campaign at discrediting OpenDocument vis-a-vis OfficeOpen XML. That's unfortunate, to say the least.

If OfficeOpen XML becomes an ISO standard, we will, in all likely hood, still not spend time on supporting it. The standard is enormous, very complex and to a large extent so badly specified that a full implementation is probably even harder than implementing the old Microsoft binary file formats. Add to that patent encumbrances and problems with copyrighted elements -- and our conclusion is that we prefer to concentrate on making KOffice a great set of applications that are satisfying to use and satisfying to develop.

And here are the views of Miguel de Icaza, one of the key people in the GNOME world:

OOXML is a superb standard and yet, it has been FUDed so badly by its competitors that serious people believe that there is something fundamentally wrong with it. This is at a time when OOXML as a spec is in much better shape than any other spec on that space.

This has led to a rather crude characterisation of the situation as KDE on the side of the angels and GNOME on that of the devil. Whether or not you agree with that position, there is a deep historical irony that things are being framed by some in this way.

When the K Desktop Environment was first announced in October 1996, it was not greeted with the universal approval that its creator, Matthias Ettrich, had hoped for. Alongside traditionalists who thought that any kind of graphical user interface was “too Windows-like” or just downright “sissy”, there was a deeper concern over the licensing of the underlying toolkit, Trolltech's Qt, which was free as in beer to hackers, but not free as in freedom. As Ettrich told me in 2000:

Everybody joining looked at alternatives [to Qt], and we had a long discussion: Shall we go with Qt? And the result was [we decided] it's the best technical solution if we want to reach the goal that we have.

Since Trolltech refused to adopt the GNU GPL for Qt (at that point: it did later), and since the KDE project refused to drop Qt, many hackers decided that they had to start a rival desktop project that would be truly free. One of the people thinking along these lines was Miguel de Icaza, who ended up leading a global team to create a desktop environment – although that was hardly his original intention:

Initially we were hoping that the existence of the project would make [Trolltech] change their minds, but they didn't. So we just kept working and working until we actually had something to use.

That “something to use” grew into GNOME, a rich, full-featured desktop environment, just as KDE had done, until the free software world found itself with the luxury – some would say liability – of two approaches.

The details of their rivalry are not relevant here; what matters is that in the beginning GNOME was clearly perceived as the saviour of the free software movement, with de Icaza as its knight in shining armour, which is rather at odds with a current widely-held view on his place in the hacker pantheon.

The point here is not to take sides on this question, but to demonstrate the amazing and pernicious effects of Microsoft's recent engagement with the open source world. The growing tensions between the KDE and GNOME camps are just part of that: another facet is the split of companies into those who believe that intellectual monopoly deals with Microsoft are a good idea (Novell, Xandros, Turbolinux) and those who do not (Red Hat, Ubuntu, etc.).

The result is a growing schism that can only serve Microsoft's interests. Unfortunately, this is one area where we don't have a choice: we need to heal the rift. The question is, How?

Glyn Moody writes about open source at opendotdotdot.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Try moving files created on

Janet Kellman's picture

Try moving files created on MS office between Windows and Mac/Linux platforms. It is a complete nightmare. MS office has its own set of mightmarish cross platform and backward compatibility problems.
So I use google docs :)

Janet Kellman, software reviews manager

the problem with all these

Vodka's picture

the problem with all these format wars is rather serious: lots of folks need assurances that a format will be in use for another say 50 years, without having to search for and buy some supporting software in the future. (it may no longer exist.)
Libraries want to convert graphical and scientific documents into a digital format once, not every 3 years to accommodate proprietary word processor changes. similarly i maintain private records which i hope to be able to read in some 10 years from now. consider vital medical and government records; a truly open format with supporting software is of fundamental importance to the society and individuals alike.
now all you gotta do is to ask yourself whether you can trust Microsoft products to store your documents? will you be able to open in fifty years from now your local library digital documents?
it is easy to realize the importance of ASCII standard (you can read most computer typings thanks to it) ODF from has an equally important role to play in the future. i do very much like the idea of standards and having one standard - already approved one - seems to be enough. No second ASCII please. Microsoft can implement a kicking editor based on ODF and still make loads of money; proposing two standards was a dumb and counter-productive move.

The standardisation process

Song's picture

The standardisation process of OfficeOpenXML has turned sour, not
in the least because Microsoft couldn't resist the temptation to cheat.
Right now we're seeing evidence of a concerted campaign at discrediting
OpenDocument vis-a-vis OfficeOpen XML. That's unfortunate, to say the
download song


Nedd Ludd's picture

I'm confused. What does KDE and GNOME have to do with ODF and OOXML?

An application written using QT is able to write an ODF file.
An application written using GTK is able to write an ODF file.
Both toolkits support ODF.

An application written using QT is able to write an OOXML file.
An application written using GTK is able to write an OOXML file.
Both toolkits support OOXML.

The attitude of the organization that develops the desktop environment toward a particular file format is irrelevant. If anything, the window manager should be file format neutral and so should the organization that develops it.

BTW -- if you need an API for manipulating documents then your best bet is to use a tool that is designed for that. OpenOffice's UNO rocks!

Miguel de Icaza

Anonymous's picture

The rift is easily healed. Let him go away into the wilderness until he realizes that OOXML cannot be implemented by anyone but M$ or one of its pets. He will then return to the fold enlightened. If he manages to get M$ entwined with GNOME, we can fork it.

The fundamental flaw in the quote from Icaza is that the proposed "standard" references secret software from M$. Hence, it is impossible to implement without reverse engineering that other OS. Additionally, the complexity of the standard makes it usable only by corporations with very deep pockets. This is nothing like FLOSS. The EU has spent years trying to get M$ just to talk to other computers systems. What makes Icaza believe he can do better?

More choice, more flexibility

Mark Reid's picture

Is it really a case of a divide? There are two different approaches, but is this a bad thing? It will only become a bad thing if KDE or GNOME go the Microsoft way of making sure that their "rights" are protected by making their system as incompatible as possible with their compeditors. I don't think that either the KDE or GNOME camps are ready or willing to go down that line. In the meantime the existence of the other will keep each on their toes and generate imaginative programming which will be copied and proliferated because of open source.

The real beauty of the current system is that you don't have to have one or the other, you can mix and match packages because in the end it's all Linux. So I've got Ubuntu with it's GNOME desktop and I run a mixture of GNOME stuff, KDE stuff and other packages which are neither. It's that ability to be a la carte about what we are able to do with the Linux platform that is it's real strength (apart form all the security and stability stuff). Hey, that's the beauty of Linux and it's that which we all need to fight to keep.

I would not trust any

Anonymous's picture

I would not trust any standard from MS or anyone in the FOSS community pushing it. They have probably studied companies such as MySQL and their dual license model. They will most likely get everone on one page with great features but then turn around and add the Windows/Office only extensions. They'll then say Open Office etc. is great but to unlock the must haves you must buy Office. The FOSS community will be right back to chasing MS in order to implement those features and provide a compatible product. Rather than taking part in this endless cycle the FOSS community should now be working on its own must have features and tight integration across projects. The most beautiful thing about open source is the the openess you have to integrate at all levels that you don't have with MS. I say use that to make good software and forget about MS. When tightly integrated must have apps that are linux/BSD centric are made then people will come.

I think the fact there are competing alternatives is overblown

Brian W. Masinick's picture

Too bad people have to argue about having more than one approach. I don't know about you, but I love the fact that there is a Vi and an Emacs, a FVWM and a Fluxbox, a Firefox and an Opera, a KDE and a GNOME and so on for each kind of application. I only mentioned two alternatives in each category. Clearly there are many more.

On one hand, when you commercialize, certain camps need to be standardized, but look at the sale of cars. In certain markets, you get all kinds of accessories as standard equipment. In other markets you get an ala carte menu of choices.

I think in software, you need to have the ability to go either direction. The good news is that we can do so. The bad news is that you have to effectively manage when you go one way or another and why you choose to do so. This is good marketing and also good project management. These are two areas where there is plenty of room for much improvement. But let's not suggest that this is all bad. I think it is great that we have the flexibility. What is needed is more maturity in the manner in which we handle the flexibility and choice.

The so-called 'split' is over hyped

Paul Cooper's picture

While there is a nice symmetry in the story, the fact is that Miguel (by his own admission) hasn't been directly involved in GNOME for some years.

The truth is that within each project there are folks with a wide spectrum of views on OOXML, and by and large development and collaboration, both intra and inter project, carries on the same regardless of the OOXML debate.

That projects under the GNOME umbrella, such as Gnumeric and Abiword, may implement OOXML compatibility is not really contentious (as they already support .xls and .doc). The only contentious event was the GNOME Foundations decision to support Jody Goldberg in taking part in the OOXML committee (Jody is a Gnumeric developer).

This could be seen and used as tacit support for the standardisation process of OOXML, however the Foundation has been taking steps to distance itself and make clear it does not support OOXML as an ISO standard. One thing to note is that ECMA does not allow individuals on to it's standards committees, only representatives from companies or organisations, and this was one of the factors in the Foundations decision to support Jody - to allow him to continue his work in the ECMA committee after he left Novell.