Microsoft and Open Source

[jackofallgeeks]

So there's an article Here
about Microsoftbranishing it's patents at the Open Source community and
saying, "hey, liscencing fees are just how you do business." It's really
all quite dull and I imagine none of you are interested (even *I* didn't
finish the article, and I care about some of this stuff). But somewhere in
there the Microsoft rep has a line about how, "what gives me pause is the
premise that: 'Make the stuff expensive or free, but don't make it cheap' -
I don't buy that. We should price things according to their value and in a
way that will work for the industry and customers as a whole."

The thorn that really sticks me there is that "we should price things
according to their value." Now, I've never actually taken a economics class
(though I'd like to; yeah, I'm that sort of masochist), as I
understand it the driving forces for price ae supply and demand, not value.
There are LOTS of things that have value but little to no cost. The
most-strawman of these is air: it's ultimately valuable but costs nothing
because, well, it's *everywhere*. I think a good book is *significantly*
more valuable than a DVD, but unless you've got a quality collector's
edition the DVD often costs 4x as much. There are other examples out there
of valuable things being priced low and less-valuable stuff being priced
high because of the whole supply/damand thing.

To insist that just because something is valuable is should be "priced to
reflect that" or something isn't just dumb, I'd say it's dangerous. At the
very least it's not sustainable: all it takes is for someone else to sell it
for less and you're out of business. (Which is where patents become a
problem: if someone becomes more successful than you, you sue. This is the
threat that a disk-drive manufacturer is making against sold-state drives.
SSDs are higher-capacity and speed than traditional harddrives, but the
harddrive manufactureres own patents on how drives interface with the rest
of the computer, and so they'll sue for infringment on those patends if SSDs
get big; and they will get big.)

Anyways...

Comments

[dikaiosunh.livejournal.com]

Well, sort of...

1. In Econ 101 world, the comment about pricing would just be irrelevant, not dangerous. In a perfect market, where all prices clear, and where there's a potentially infinite number of producers and consumers, with decreasing returns to scale, no transaction costs, perfect, knowledge, etc. then things will sell at the price where the nice smooth supply and demand curves intersect, and nothing that MS could say would change that. You're right - if they tried to sell above that price, someone would undercut them. And if (for some reason) they tried to sell below that price, there would be shortages of their product (cf. Zimbabwe's attempts at price controls; or, if you assume they could produce as much as they wanted, it'd have to be because their per-unit cost was 0, in which case they'd undercut everyone and sell for free anyway).

Of course, we don't live in an Econ 101 world. In Econ 201, MS extracts monopolistic rents, as do the other major software brands. In Econ 301 and beyond, things get really messy because people aren't perfectly rational, there are information problems, etc. etc.

Over in Soviet Econ 101 they also point out that prices are set neither by the value of the inputs (one version of which would be the labor theory of value, in a nutshell the Marxist theory of "fair" value) nor by perfectly rational decisions in the market. Rather, there's a lot of irrationality, especially in the fact that people value things in part because of the accepted price, and that price is based on perceived value, and so forth. Back in Capitalist Econ 401 they start talking about price stickiness and differential equations and I am no longer of any help. Then down the hall you've got a guy modeling imperfect markets with agent-based computer simulations, and finding all *sorts* of weird stuff.

2. Re: patents. Stop me if you've heard this, but patents - at least in theory - have a perfectly respectable purpose, and it's not just to enforce "irrational" prices. They're there (again, read all this as "in theory") to make initial R&D investments rational. The problem otherwise would be that many bits of IP have a very low cost to produce each unit beyond the first. So, if MS spends $1 million (just to have a round number) creating Windows, but then it only costs $1 to produce a Windows CD, and let's say it costs $100,000 to reverse-engineer Windows once it's created... Imagine that MS and its competitors each produce 1 million Windows CDs. MS' per-unit cost is $2. Its reverse-engineering competitors' per-unit cost is $1.10. Who do you think is going to go out of business? So, if there were no protections for initial investment (effectively, some way of imposing an extra 90 cents of cost on competitors), no one would do R&D and civilization would collapse.

In essence, it's an attempt to simulate for IP what happens with other products that have high initial investment costs. Toyota would go out of business quickly if it had to pay to build its factories, but then anyone could come use them for free to make cars.

I suspect there might be some way you could fix this problem with futures markets, or something, but some form of IP protection is the simplest and most common way to do it.

I'm not an expert on open-source business models, so I'm sure there are creative ways to get around this. But, as I understand it, open-source development relies heavily on altruism and on non-monetary compensation (the glory of being the guy/gal who developed X), both of which are hard to make work on large scales.