Linux has an inferior desktop and inferior general application integration to either Windows or the Mac. But it is similar to the Mac in that it has an open source Unix-clone base with a GUI layer riding on top (actually more than one). Linux's GUI layers just aren't nearly as good.

One reason is that 100's of geographically distributed developers just can't do as good a job on creating a cohesive design as people under one roof can.

Another reason was very clearly described by Simon writing on the Chimera mail list: "The open source model, in and of itself, does not encourage good UE [user experience]... in most cases, good UE designers and good coders are different groups of people. Since good UE designers usually don't contribute code, their contributions to open source are ignored."

However, IBM is investing something like $1 billion in Linux. Other companies are investing significantly too, mostly in the form of making their apps Linux-compatible (such as Oracle). All of those companies would like to break Microsoft's stranglehold on the computing world. They don't make their profits off of the OS, so they can back an open-source OS. But they can't take over the desktop as long as their desktop environment is not as good as Microsoft's (or Apple's).

But think of the semiconductor industry's Sematech, which "is a unique endeavor of 12 semiconductor manufacturing companies from seven countries. Located in Austin, Texas, USA..." The key phrase here is "Located in Austin, Texas". They are getting a bunch of people under one roof with funds contributed by 12 large companies to work on technical projects of interest to all the members.

Suppose the companies who are investing heavily in Linux did exactly the same thing. They could put together a team of comparable size to the one working on Mac OS X desktop environment in one location if they chose to, and probably create something as good given enough time. There is no reason the resulting code couldn't be open source.

To address Simon's point, this group can have the discipline in to make sure the designers and coders work togehter to implement the best design ideas.

The participating companies companies would tend to be hardware companies or application companies that want to break Microsoft's monopoly.

Is exactly that happening with the Gnome development effort? According to an employment notice originally on Ximian's web site but no longer online, this effort has "concentrations in Boston and Chicago, with QA in Mexico City." And it is supported by such big-name companies as IBM and Sun. And on August 21, 2002, they released v1.0 of their Human Interface Guidelines.

Maybe, just maybe, against all expectations they are doing the right things and have a chance.

Update: I just recieved this in an email from Joel:

Indeed much of the Gnome UI work is going on in Boston at Ximian, but unless they get a lot more funding I don't know how long this will last.

At least on the surface, it's obvious that the current contributors such as IBM should invest a lot more and bring it up to an OS X-like pace so that they can create a true open-source competitor to Windows and Apple. Joel's info makes it sound like they won't. Actually I would expect them not to do the right thing in this, so if they don't it's no surprise. But hopefully they will come around.

10/17/02: Doc Searls' SuitWatch discusses Linux desktop progress.

1/23/03: Sun promises a Linux desktop environment by summer 2003.

2/5/03: Group Hopes to Give New Life to Desktop Linux [Datamation]

2/7/03: Gnome 2.2: "An example is a new Gnome 2.2 feature called Startup Notification. This signals that an application is in the process of opening, so that the PC user doesn't feel the need to click on it again." This is very basic stuff which, of course, "real" application launching UI's such as Mac OS X's have always had. But it shows Gnome is slowly moving in the right direction.

If you have more info on this please email me.

		© Copyright 2006 Gary Robinson. Last update: 1/30/06; 2:48:04 PM.