> 1) It
> typically presupposes high levels of expertise in users. Cf. Joanna's
> example of her daughter surfing the web within 10 minutes of her
> taking it out of the box.
This level of user-friendliness is very recent even in proprietary systems. My dad, with a new Windows box, needed helpdesk support to get connected -- no big deal, but isn't that what most people have to do with Windows, right?
Please explain, using *current* examples, how Linux is particularly demanding on a user. Mind you, most nonskilled people I know have never installed or even reinstalled Windows on their own, either. Remember that Firefox/Thunderbird is free-libre software. Also know that the past several years when I've installed Linux, I wound up at a GUI without having touch a command line, just some text-based menus. That may be more than most people want to do, but then again, most people never install *any* OS, even on pre-specified hardware -- a privilege that is rare for Linux users.
> 2) FOSS relies heavily on free riding on
> the resources of employers (and parents, as has, nonjokingly, been
> pointed out). People cannot pay the rent trading code, nor can they
> pay for the computers and Internet connections necessary to develop
> and distribute it.
Can you distinguish the bit about parents from hearsay?
When you say "free riding" on employers, do you mean that employees are working on free software on an employer's time and employer's resources without employer's permission? IIRC, the FSF actually strongly advises against doing that because that creates legal problems around ownership. In fact, I think that's why emacs development forked, at least according to Stallman.
I'm not sure what you mean by people can't pay the rent by trading code. Vanishingly few people (compared to the users) actually contribute any code at all, and I know for a fact that many live quite well on it. Yes, they get paid by employers and grants to do so. Does that not count?
3) Software is a highly unusual product -
> infinitely reproducible and near-zero cost. It is the nonrival good
> par excellence. How can the model be translated into the production
> and exchange of peaches, health care, or transportation? People in
> the computer world have a really exaggerated idea of the size of the
> sector. It's actually rather small.
Yes, in terms of production it's a best-case scenario. I don't suggest that the same model can be extended to more physically-based objects (and I don't know of any notable arguments otherwise). But is it completely useless to have a good example of non-private ownership in a best-case (though not at all ideal!) scenario?
As far as the importance of computer world, I'm not sure if sector size (however measured) is a fair metric. Water and sewage are pretty small too, I bet. Plus the low margin reproduction aspect extends to a great deal of culture, writing, etc. I'm not suggesting that the exact same principles get applied to those as well (and neither do the FSF people), but I don't understand how the whole matter gets dismissed so easily.
-- Andy