Try this analogy. How would it work if we "deregulated" football and hockey? Trust the players to regulate themselves. No more penalties. No more central authority to impose rules.

What a beautiful straw man! Good job confusing anarchy and zero government (in your analogy the penalties) with a free market. Are you trying to argue that the only choice is binary, between a Soviet-style "central control" and "no-government anarchy"? The very fact we have a regulated market shows there is a broad middle ground. All I suggest is that maybe we've moved too far in one direction, not that we should immediately shoot to the opposite extreme.

free markets are a good thing but they need some amount of government regulation to balance out greedy dishonest a-holes.

OK, how about something simple, straightforward and universal, like ending the idea of corporate personhood, or perhaps "corporate fraud automatically 'pierces the veil' allowing lawsuits against executives and stockholders"?

OK, I'll try: domain registrars do not operate in a free market.... If they were in a truly free market, GoDaddy could (and almost certainly would) simply refuse to transfer any domains away from themselves.

You are absolutely, 100% correct in your premise, but you arrive at the wrong conclusion; this is unfortunately all too common because very few people understand what "free market" really means.

Your argument is phrased like this:
* Domain registrars do not operate in a free market.
Correct: a "free market" is one where buyers may select freely from vendors offering different services. Regulation prevents this differentiation, and thus closes off the market from being "free".

* GoDaddy, a registrar, could refuse to transfer names in an unregulated market.
Correct, but you overlook the competition from other registrars -- your conclusion:...

* Therefore, we can expect an unregulated market in domain names to provide fewer services [transfers] than a regulated one.
...is thereby invalid. In a free market, it would be an advertising advantage to allow transfers to your company; furthermore, we can expect each business' interests to force any such agreements to be reciprocal and thus forcing businesses to allow transfers away as well.

I'm sorry, you seem to be conflating corporations -- who have a position within a market, and long term goals -- with the market as a whole. Let me demonstrate by removing the pronouns from your statement:

...still believe that free markets work. Free markets [note: only plural antecedent] have repeatedly proven that free markets need to be regulated or free markets will take advantage of their position.

I can't make head or tails of what you mean by "free markets will take advantage of their position". Markets, as a whole, don't have a position -- that only can make sense within a market, which requires your use of "they" to referring to a smaller unit, not previously mentioned. I'd assume you mean corporations, in which case you're confused as to what "free market" means -- it is when buyers have freedom to select from among multiple sellers, and refers to corporations not one whit.

For the next year (or so), this will be my counter-example when I debate politics with people who argue that a centrally regulated economy is better than the free market -- as in, "I will happily agree with you, if first you explain this one annoying fact please."

* Constituents and businesses pleaded with Congress [the regulatory body of US "central economic planning"] not to pass SOPA. Congress did anyway.
* People threaten to boycott GoDaddy (direct financial loss) due to supporting SOPA and they reverse course immediately.

I feel the answer is clear, obvious, and simple: businesses are more responsive to their "constituents" then politicians are. Therefore, we should discard [most of] the business regulations -- by which I mean things like minimum wage or union laws, not universal "regulatory" laws like EPA pollution controls -- and go back to a free market.

(Oh, and before people asks: EPA regulations are "universal" because private individuals can violate them just like big business does, for example by developing protected land, or burning waste material. Wage and hiring laws are not "universal" because private individuals cannot be in violation of those laws, only businesses.)

I have long described both MMO gaming and Facebook social games as being a "well-padded Skinner box" for their staggered/random reward system. Do you see any possibility for anything else to eventually replace this model?

Is this is the end stage for the ownership of ideas: that just as there is no longer "public land" in the sense that every piece of land has an owner (even if it is the government), every idea will need an owner?

But when you buy software, you aren't usually buying the software, but rather a license to use the software

What dirty, dirty, lies! I can tell you for a fact* that I have never once personally licensed software despite purchasing several valid copies. I have four reasons for saying this:

— First, having a debugger installed on my PC means I can modify the installer [while running] to show no EULA at all, so further proof is required I have agreed to whatever blatherskite the software publisher is trying to push.
— Second, I have never been presented the EULA terms before paying for software, so all claims the transaction was not a sale occur ex post facto.
— Third, I have never given my money directly to the software publisher, but to a retailer such as GameStop, Amazon, or Best Buy, so the publisher becomes a "third party" to this transaction and lacks any legal standing to decide its' status.
— Finally, because US copyright law allows me to make copies as "an essential step in the utilization of the computer program" and one copy as a backup (see Title 17 section 117), I have a pre-existing right to install the software; the typical EULA therefore grants no quid pro quo to me and therefore is contractually void.

Disclaimer: This is not legal advice, I am not your lawyer, and in fact I am not a lawyer at all! I do play one on ... I mean, work with them every day, and several of them have agreed that the above arguments are strong enough that certain other people would lose sleep at night should I ever find cause to argue them in a courtroom.

[[*: This is hyperbole, really, although I suppose you could say it is "pending legal review". I don't feel the need to go beyond what copyright law permits... and that's pretty much what a typical EULA asks anyway. Until there's sufficient harm in the divergence between my interpretation and the demands in the EULA for any software I have purchased, I have no reason for me to actually test these arguments in a court of law. So far I've skirted this by avoiding really problematic publishers, but I fear the day of testing is fast approaching.]]

No, job security is programming in Lisp:

1. Your code will never "fail" when you cover all bases with restarts; and
2. It's fast to write from specification, so you get big props as a get-'er-done type.
3. Third, nobody else can debug it because they can't figure out what is being done in even a simple one-liner like "(eq TODAY (cdr (car *LIFE*)))".
4. Finally, nobody can be hired to replace you because all the schools don't teach the basics of data structure and code flow analysis but the grammar of the latest "l33t"ness in this year's dialect of VB-dot-net-sharp-plus-plus.

Another option would be to hold all bonuses in escrow for some defined period (I would suggest at least 5 years). At the end of that period, the bonus may be claimed.

This would work at least as well for stock based bonuses for CxO level officers; now they have a direct financial incentive to ensure the company will shine just as well after they leave as they do shine up the numbers for next quarter.

The problem with those averages is that it divided on a completely arbitrary line (every year ending in a 0). Either single very large spikes in either direction, or events occurring on 7 or 14 year cycles, could pull the "average" numbers used for each 10 year span. A quick analysis appears to show some sort of ~30 year period cycles occurring (1880-1919 cluster around a small range, as do 1930-1979), and both longer or shorter cycles are practically guaranteed. What I do not see adequately discussed is the prospect of very long, multi-century cycles longer than our temperature records: a cycle of (say) volcanic activity which extends for 800-1000 years could easily fit the "small" deviations in the Holocene paleoclimate records, and would coincide with the current warming period. (Even closer historically, the "little ice age" - the last big dip in the thick black line on that chart - occurs right at the start of our temperature records. Even cycles of a small as 300 years would be completely occluded by the brevity of our direct records.)

The fact the "average" global temperature for each of the past 10 years top the records list does not change the temperatures over that time span. I reference that list because it is the most convenient source for the numbers, and because it shows I am not "cherry picking" favorable data. Certainly, the peak temperatures list will prove global warming is occuring, but does not by itself prove anthropogenic causes.

The 1990 IPCC reports on climate predicted a 0.3 degree change per decade with increasing emissions. The latest report (2007) predicts a similar trend line (stated as 2-6 degrees per century), so I have to wonder: how long would we need to deviate from these predictions before it becomes significant? 10 years (as we have)? 15? 20? 50?

1/15th of the total historical measurement period is certainly statistically significant. However, as we entered the 1990s (as we approached these peak temperatures), the IPCC "consensus" predicted a 0.3 degree change per decade. The latest report (2007) predicts a similar minimum (stated as 2-6 degrees per century), so my question is: how long would we need to deviate from these predictions before it becomes statistically significant? 1/10th the predictive period (10 years) or 1/15th (another 5)?

Yes, I would cite a 23-year-old model, when it only makes 20 to 30 year predictions. After all, wouldn't it be intellectually dishonest to keep changing the "accurate" model [the one needing to be challenged] before any of your predictions can be tested?

What's your source on say there has been no temperature increase in the last 10 years?

Would you believe... NOAA?

Wikipedia's list of hottest years is penultimately sourced from the NOAA. 2001-2010 all make that list, with a maximum difference over this span of only 0.13 degrees Celsius (2005 / 2008 are listed as hottest and coolest, respectively). In order for this to be "increasing", your measuring equipment must have a error rate below 0.13 degree Celsius. A search on their page finds that the NOAA's official equipment manuals gives an optimum [minimum] error range of 0.166 degrees Celsius [listed as 0.3 degrees Fahrenheit]. I don't challenge their numbers, but their accuracy figures says the last decade has seen — within the accuracy of the measuring equipment — no change in average temperature.