Invisible Linux: The Details

The other week I theorized about "Invisible Linux" -- what Linux would need to become to really make inroads on the desktop. Since making that post, I've been refining my ideas about what this would be and how it could be created. Read on for more.

Invisible Linux would essentially do for Linux what Apple did with NeXTStep / BSD: take the open source core and wrap it in a layer that may not be open source but which is polished to a degree that makes it irresistible to end users. Such a package would consist of three things: a stable and dependable ABI/API layer for programmers; a professionally designed and use-tested interface for end users; and an organized deployment of all these things for the sake of administrators and commercial software creators.

The first and third could be accomplished any number of ways (pick a stable implementation of each and freeze it), but #2 -- the UI -- is really the toughest and most elusive of the bunch. Getting a UI right for a desktop OS is brutally difficult, and it's clear to me now that it just can't be done as an open-ended community effort. KDE and GNOME suffer from either designs that are good enough but not inspired, or too much design work that doesn't serve any real purpose except being "different" for its own sake.

To make Invisible Linux look good, there needs to be an actual design team with a designer on board -- a graphic artist whose job is to do nothing but create something that we would be happy to spend several hours a day looking at. And, frankly, such people cost money. With their talent, they could have their pick of projects, and I doubt they would work on something this all-consuming and detailed without asking for at least something back.

Everything I see here tells me Invisible Linux would need to be at least partly a commercial software project. One canard that keeps bubbling back to the surface is how any attempt to make Linux fly on the desktop will require the support of third-party, closed-source software makers. The problem is getting them to develop for a platform that commands so little marketshare -- and which is (in their perception) mainly used by people who choose not to pay for software in the first place. That's going to be a tough one to knock down.

In many ways, Linux is already a success, as I've mentioned before. It's been fairly revolutionary as a server platform; nobody needs to be brought back over this ground again. But it's going to remain fragmented on the desktop unless someone comes along, takes it as raw material, and fashions it into a form that even Linux's own creators couldn't imagine -- and does so by harnessing the twin powers of capital and singular design.

But given that Linux is meant to be thus shaped and bent, wouldn't they be happy about that?

InformationWeek has published an in-depth report on Windows 7. Download the report here (registration required).

Follow me and the rest of InformationWeek on Twitter.