Linux: It's the apps, stupid

My personal theory about the current crop of 'reality' shows is that the viewing public doesn't want to see winners so much as it wants to seelosers. Would anyone be watching that Survivor crud if people weren'tgetting voted off each week? I don't think so.

It's the same depressing public taste for confrontation andhumiliation that seems to be behind much of the recent news about the opensource world's two favorite graphical interfaces, GNOME and KDE.

I originally greeted the news of the GNOME Foundation witha similar sense of indigestion as I experienced when the Open SoftwareFoundation (OSF) was formed about a dozen years ago. The similarities arestaggering: a bunch of industry heavyweights getting together to drawbattle lines over a graphical interface.

Back then it was the OSF -- made up of IBM (ibm), HP (hwp), Digital (now Compaq (cpq)) --and its Motif GUI, which eventually triumphed in a war of attrition withthe Open Look interface championed by Sun (sunw) and AT&T (T). I look at the GNOMEnews and see the same names -- Compaq, IBM, HP, and Sun -- this time allon the same side. And just what do they expect to accomplish here?

What was so broken within GNOME that it needed a Foundation to fix? Howbadly do the GNOME developers want to take the advice of the companies whocollectively championed Motif, and then its mutant offspring CDE, as what theworld should use for a desktop?

I know these companies mean well and all, but let's not kill GNOME withkindness. It was moving in the right direction before all these nicepeople came along and I hope the developers follow their instincts.They're on the right track and I hope they don't get derailed.

While the Foundation's real purpose may be to dispense advice thatcan (and most probably should) be politely ignored, I'm more concernedat the way Sun's intentions were rolled out, and fear that the road to hell is beingpaved by Sun and GNOME's other well-intentioned bandwagon jumpers.

First there was the announcement that Sun would be putting StarOffice under the GPLand porting it to the GNOME Gtk widget set. Fair enough,they had to pick something, and as StarOffice was written in C itprobably made more sense to use Gtk than the KDE/Qt approach whichrequired the C++ language. Besides, if it were open sourced and reallysignificant, StarOffice could also be ported to KDE -- putting codeunder the GPL means you can do that kind of thing if the demand andwill exists.

"We did have to make a choice, and we chose GNOME for a number ofreasons," said John Heard, manager of architecture and strategy for Sun'sWebtop and applications group. "But Sun will support KDE, and we will helpthe KDE team port StarOffice if the need arises." Heard also said Sunwould support any efforts at closer cooperation between KDE and GNOME.

So much for the reality of the situation. Unfortunately, the demand for aloser arose, fittingly in the middle of Survivor season. First we had thatInternet answer to British tabloids, the Register, trying to evoke fighting wordsfrom Sun and a KDErebuttal (headline: "We're not scared") on the same day. Shortly aftercame the GNOME Foundation hype-a-thon at LinuxWorld as well as Sun'sannouncement that GNOME wouldbe its future desktop, and folks all over the place lost no timedeclaring GNOMEvictorious and KDEa loser. Some KDE fans fired loud returnvolleys and others were simply content toplay warcorrespondent. Even Mac commentatorsgot into the fray, but maybe that's to be expected since they'vealways been used to having no choice of GUI.

But what's the point of it all? The sanest heads have advocated calm, andthe KDE team'sofficial comment on the matter said it best:

None of it matters.

Indeed. We don't need to define losers or even winners. These are thebattle cries of the Unix warriors, come to play out the same tiredskirmishes on Linux turf, marshalled by commercially-bred mind sets whichhold that competition must result in victory and defeat. Thatmind set isn't just unhelpful, it's harmful. The open source communityneeds to see beyond the puffed-up wars and the public spectacle of itall. We need not heed the call to arms.

Linux is and always has been about choice. Someone who uses Linuxin a world otherwise dominated by the likes of Microsoft has already made astatement that choice is important. Why should we be in so much of arush to eliminate the choice of Linux GUI? Why must one size fit all?

GNOME and KDE have different philosophical approaches to many issues thatplay out in very different ways. GNOME wants to take a methodical, highlystructured approach that will have all the pieces fall into place in themost elegant fashion. KDE is more interested in being fast and nimble,releasing sooner, and (for instance) choosing the lightweight Kpartscomponent system over GNOME's more standard, more flexible, but also morecomplex and bloated CORBA-based approach.

I'm too greedy. I don't want one to beat the other. I want both to keepdriving each other forward, so that they keep making each other better.Given the development pace of Eazel's Nautilus and KDE's Konqueror,both KDE and GNOME will soon surpass both Windows and Macintosh GUIs insheer usability.

And recently the news got even better. The decision by Trolltech this weekto releaseits Qt library (core to KDE) under the GNU Public License is anincredibly significant and welcome piece of news that helps to deflate theemotional element. Now there are no more issues of licensing religion toseparate GNOME and KDE, Debian has no reason not to ship with KDE, andsoftware purists must now compare the two desktops on their merits.What a concept.

(One could even argue that the Qt license, being under the "full" GPLrather than the "lesser"GPL used by GNOME's Gtk, is now closer to the Free SoftwareFoundation's ideals than the GNOME project it calls its own. But thatwould be splitting hairs, eh?)

As the emotions inevitably calm down, people will realize that what desktopLinux needs, far more than winners and losers, is applications.

Sure, both GUIs need some serious usability improvements. But Ialready see that happening, especially since Eazel joined the scene, andI'm looking forward to the imminent release of KDE release 2. What Linuxreally needs is a broader range of apps across the board, especially thekind of personal software such as Quicken and Maximizer that just don'thave matches in the open source world. Another area in which Linux reallylags is educational stuff and indeed anything targeted at an age levelbelow that of open source programmers.

Perhaps as apps vendors digest the fact that the installed base of Linuxdesktops will soon surpass that ofthe Mac (if it hasn't already), more will come around. Of course thesevendors have to contend with the mentality of some Linux users that allapps should be free since the OS is -- but that will also diminish overtime if the truly unique and hard-to-duplicate apps get ported. Of course,making Wine easier to use -- so thatLinux users could easily install native Windows apps -- would be a goodresource as well. But that's just a stopgap.

To their credit open source developers, by and large, have ignored thebattle analyses and just kept coding -- and that's good news for all ofus. For as Linux gets better, as GNOME and KDE keep pushing each other, asopen source desktops continue to increase in popularity and attract moreapplications, we all win and nobody loses.

Except Microsoft. And perhaps those who look for losers.

Do you think the win/lose game is bad for Linux? Tell Evan in theTalkBack below or in the ZDNetLinux Forum.