There's a hard truth to be learned about making most computer users care about open source: they won't care. But that's the beginning of wisdom, not the end of it.
The grim truth is that most people -- including most computer users -- simply do not care about the benefits of open source. They don't care, period. What's worse is that the efforts made to make them care often backfire. Software engineering in the abstract is not something most people are curious about, and is so far removed from any aspect of their work that it's not surprising they do a very good job of not caring about it.
That's because open source isn't a marketing or usability feature. It's a design or software engineering feature. It's not something that was ever meant to be an upfront selling or evangelism point to non-technical users.
It's also because -- and I'm going out on a limb here, but bear with me -- programmers seem to think that human behavior, in the long run, is as pliant as software. Just pass the right arguments (pun intended) and they'll come around to your way of thinking before long.
Except that this never happens in real life, so why should it happen here?
The more forceful open source advocates see it that way. Presumably it's because they've been living for so long inside the whole ship-in-a-bottle atmosphere of open source advocacy, it's all they can see. They assume that a) deep down everyone either already thinks like they do, or b) they can be persuaded to do so if you just make an argument that's sound enough.
It's magical thinking. Verbal arguments only persuade when the problem itself is wholly verbal. When the problem is experiential -- as in, "This stupid program is a pain to use and crashes on me all the time" -- the only thing that persuades is another experience. For most people, that consists of changing programs. Sometimes those programs are not open source. Welcome to life in the big city.
I know that I don't want to be patronized about why a given program is theoretically better because it's open source; I want something that beats the competition, no matter how it was put together. So do 90-something-percent of the rest of the user world. I've ditched Excel for OpenOffice.org's Calc because it's as good for my needs and costs nothing, but I still haven't ditched Outlook because all the alternatives I've looked into simply haven't come close to the same mix of features.
Understand that I do not say these things because I think using open source as a selling point is wholly worthless. It's useful to the people who, again, already care about such things. The rest of them just want to get our work done, and not get bombarded with rhetoric or struggle with amateurish interfaces or boggle at perennially incomplete documentation. (And yes, you can get all of those things from badly-written commercial apps, too -- which is why there's a whole ton of them constantly fighting to improve their lot and win market share away from each other, instead of just taking the passive "if you write it, they will come" approach.)
The argument for using open source to be presented to those folks should be a great program that just plain beats the competition -- all competition -- no matter who made it or why.
Because if you can win over those people, you can win over anyone.
InformationWeek has published an in-depth report on Sun's future under Oracle. Download the report here (registration required).