“How many of you can honestly say that you’d come out looking good if I trawled through your files, picked the worst piece of code lying around in there, and reviewed it on national TV?”

I’ve written poor quality code. I’ve never written poor quality open source code, because I know that other people might see it. This is an advantage of writing open source code — it improves quality.

“There are plenty of allegations, but no evidence at all that even one single published result is affected by all this.”

There are plenty of questions that would be answered by open sourcing the code. For example, one chap is scrabbling around trying to make sense of station adjustments. Questions like that would go away more quickly if people could see how such adjustments came about.

At the moment it’s plausible to claim that such problems invalidate anything that relies on these adjusted station records.

[The idea of open source on its own does not solve this problem. You have to build a knowledgeable community, and have a way of dealing with deliberate denial of service attacks on scientists' time. People like Watts aren't going to stop making up spurious attacks on the science just because the data is open. He's been trying to find errors in station adjustments for years, and has found nothing wrong yet. That isn't science, it's more like the spanish inquisition. - Steve]

Regarding comment #4, I don’t personally have to understand all the code and methods myself. It’s enough that people I trust are saying good things about it. In the case of open source encryption software, for example, I can be certain that if there was a way to break the encryption people would be shouting about it.

[You can shop elsewhere - there are thousands of climate scientists across the world. If you don't like the CRU folks, go to any one of a large number of climate science labs elsewhere (start here: http://www.realclimate.org/index.php/data-sources/). An analogy: Imagine your doctor told you that you have to change your eating habits, or your heart is unlikely to last out the year. You would go and get a second opinion from another doctor. And maybe a third. But when every qualified doctor tells you the same thing, do you finally accept their advice, or do you go around claiming that all doctors are corrupt? - Steve]

The idea of getting a broader audience to understand what climate models are by participating in building them is a pet project of mine (see: http://www.easterbrook.ca/steve/?p=482). But setting them to work on an existing climate model wouldn’t work; we’d have to build our own simplified model, aimed at elucidation of the principles, rather than doing cutting edge climate science.

Perhaps we should turn over cyber-security to the scientists as well?

Wading through the hay of all the straw men you so enthusiastically knock down (which I admit is fun to read), I am still left wondering. There is a knowledge domain called simulation-based engineering science. It includes a pretty well defined approach to V&V (includes method of manufactured solutions, etc.). IMHO, this “community of programmers” would be able to “take on the job instead” of the scientists. And since confidence is the crucial issue, their independence would be invaluable.

The point that I think you are not directly addressing is the “accreditation” problem. V&V assumes that an accredited authority performs the V&V. If the confidence in some portion of climate science has been called into question, climate science cannot logically use its authority to reestablish its authority. Outside remediation is needed from a trusted authority. For example, for the “hockey-stick” issue, IIRC, outside statisticians were brought in.

[That's an awfully big "if" in your final paragraph. There are plenty of allegations, but no evidence at all that even one single published result is affected by all this. -Steve]