Any talk of Linux brings with it talk of what it will take to get Linux on the desktop in big numbers.  Much of the talk in this vein revolves around distribution X versus desktop Y, or something of that nature.  The real issue, though, may not be a particular distribution or package model, but the mind-set of the creators.

Rob Enderle, no stranger to the business, weighed in recently on what he thought would be needed to bring Linux to a broader audience.  His big point of comparison was Apple: Unix underpinnings (actually, BSD), but with a polished UI that is the envy of most everyone else.

Arguments about what constitute a good UI could go on forever.  The real issue here is how to produce one, and if we go by Apple's model, that means spending a great deal of money to hire, cultivate, and retain a development team capable of creating and polishing such a thing.

And so arises the question: Do you need to throw tons of money around to produce a good UI?  Or is it more about the development culture, where (as with Apple) the whole project is tightly managed from the inside out?

One associate of mine was of the opinion that good designers can be had cheap enough -- it's just that most FOSS development teams look down their noses at designers and don't see what they have to contribute as being just as important as the backend logic.  As he put it, the problem with FOSS UIs is that everyone thinks they're a UI designer, and most people -- most programmers, especially -- are just not.  They don't understand that what may be counterintuitive to them as a programmer is not counterintuitive to someone who is just trying to get through their day's work.

And so arises an even bigger question: Can you foster the right attitudes about programming vs. UI design without needing a closed-ended, autocratic environment to do it?  Or to put it another way: do we have to choose between a closed but reliable ecosystem, and an open but also scattershot one?

I'd like to think it's not a Hobson's choice between the two, and that an open ecosystem will yield more genuinely useful fruit than something with a one-way door.  But I also don't think it will happen spontaneously.  It will require people who can see Linux as a starting point and not a destination.