De Icaza is best known for setting up the Mono project in a bid to develop a clone of the .NET programming environment for Linux. He styled it as an attempt to draw Windows programmers over to Linux. There is no evidence to show that it has achieved this aim.
De Icaza, who also spent many man-hours developing Moonlight, a clone for Silverlight which Microsoft hoped in vain to make a Flash killer, claims the reason for the Linux desktop not gaining marketshare was because developers behind the toolkits used to build graphical applications for Linux did not ensure enough backwards compatibility between different versions of their APIs. (Moonlight has now been abandoned.)
And so, he says, frustrated developers started shifting to OSX and to the web. The article detailing his claims did not go into the history of GNOME and examine some well-known truths about the development of the Linux desktop.
At that time, KDE, the other big Linux desktop project, was in existence. There was one reason cited for starting GNOME: the need for a free desktop. This was an emotional appeal to users - KDE, which had kicked off in October 1996, was using a non-free library called Qt which was owned by a company called Trolltech. GNOME claimed to be "free" software. (Qt has since been released under an open source licence as well.)
De Icaza had already interviewed for a job at Microsoft in 1997 which seems peculiar for one who claims to be devoted to free software. Along with Nat Friedman, he set up a company in 1999. It was initially called International Gnome Support, then Helix Code and later Ximian. The aim of the company was to market the GNOME desktop commercially. It was bought by Novell in August 2003.
Both de Icaza and Friedman had one thing in common: they were acquainted before they met, at Microsoft in 1997, where Friedman was an intern on the IIS team. De Icaza interviewed for a job at Redmond, to join the team that was porting the company's Java VM to the Sparc, but was unsuccessful.
But De Icaza did not mention anything of this history when he was asked about the Linux desktop. It is not unreasonable to postulate that if there had been just one major desktop project, it would have made more progress; countless hours have been wasted in internecine desktop wars, many of them sparked by De Icaza himself.
One example: in 2001 Ximian quietly used terms like "kde", "konqueror", "dcop" (the KDE "Desktop Communications Protocol"), and kparts (the KDE component model) as Google adwords for its own ads. When KDE developers Kurt Granroth and Andreas Pour published an article titled "'Business Ethics' in the Open Source Community?" which chastised Ximian for this deceitful practice, Friedman had sufficient chutzpah to say: "We knew what we were doing," noting that Ximian's goal was to ensure "as many users for Ximian GNOME as possible," within the context of "friendly competition."
De Icaza's Mono and Moonlight projects have been extremely divisive, splitting the free and open source community many ways. After Novell was bought by Attachmate and taken private last year, Mono was jettisoned by the company.
KDE had one disadvantage from the beginning - it was an European project. It had rather simple aims - to clone every function which was available to Windows users so Linux users would have all that functionality at their disposal. It has done much more than this and today is eminently more usable than GNOME. KDE never bothered much about marketing itself, the team simply produced software which was good.
GNOME was always viewed as an American initiative. There was more marketing, more buzz words (a trait that continues even today) and many structures of so-called leadership. Until at least release 2.0, GNOME was a clunky, heavy desktop, which was buggy in the extreme, Ximian was something of a nightmare; I personally had to reinstall Linux twice during the time when I was experimenting with it after Ximian installs went terribly wrong.
GNOME 3.0, released last year, hasn't set the Thames on fire either. It has resulted in a litany of complaints.
It is thus surprising to find De Icaza, who has caused a considerable degree of division in the free and open source software community and among Linux desktop developers, not telling the full story when asked why the Linux desktop has not gained more traction.