I believe that oxen have been used to provide mechanical power since the Middle Ages, probably much, much earlier. This Irishman would appear to have re-implemented a very old idea.

That's basically what the Australian Feds do. They lie about their age, and wait to see if the suspect wants a meeting. When the perp shows up (often with porn, toys and cameras) they arrest him. Good thing too. I'd be surprised if the US authorities don't do similar, with a few choice moves to ensure the process is legal.

Entropy is reversed by doing work, which is what life does, by transforming energy (pumping heat).

walking around, those that were not out-competed and failed to adapt.

Macroevolution is not a scientific term, and if used in a scientific context muddies the waters with irrelevancies. There is no such thing as macroevolution, nor microevolution. There is natural selection and the wonders of DNA.

Which he did. It is called Smalltalk.

If I'm distributing a commercial product, then I do as you've described. Or I can file-out a class or category (or a change set) and distribute the source for file-in into someone else's image. Not so different from how C works. A very common misconception by the way. Exactly when something is run is also an interesting question, as all objects are live, including those objects called classes.

The thing I hated most about Java is that I can't just create a new class and implement a few messages and use it in existing code, unless I somehow shoehorn it into an existing class hierarchy, or implement an interface. If the existing code was done without interfaces in mind, I have to edit a few hundred methods in dozens (if lucky) of files, and not break anything.

To avoid this I have to over-engineer prototype code, and increase production cost estimates by at least a factor of 4, or else use a generating framework that leaves me with no idea what any of the generated code does, which makes fixing subtle bugs so much fun.

At least smalltalk works. And the smalltalk debugger is possibly the most elegant and efficient ever.

I don't think you've really looked at all the issues with Smalltalk. If you look at squeak, for example, most of the VM is smalltalk, that gets translated to a subset of C, and you only have to implement primitive data types to build a new VM on a new platform.

Gnu Smalltalk is text-file based (and is unusual amongst smalltalks for it).

As for tools, smalltalk provides the tools, they are just different than what is used for other languages. The tools for smalltalk (the browser, monticello, the file list etc) work really well for smalltalk. It's not that you have to learn a new set of tools to use smalltalk, rather that there is no need for any others - nearly 40 years of development has produced a very efficient and intuitive environment.

In my opinion the only place Java beats Smalltalk is in that its threading model is a bit more efficient. In return you are forced to abandon most of the benefits of a dynamic language, and use multiple layers of nearly-C++ to get the job done. Or maybe I just don't understand Java. I don't get why an object has to care about what class a reference dereferences to, and why methods have to have all that type declaration fluff around them. Maybe I'm just old though.

I have to wonder if the Stern Hu case is a factor in this. For those not in the know, Stern Hu is an Australian who is currently on trial in China for accepting bribes and industrial espionage. Mr Hu works for Rio Tinto, the iron ore giant that China failed to buy last year.

Everyone knows that bribery is a large part of doing business in China. Stern has been in court for 3 days now. Even consulate officials have been denied presence at the trial, relying on state reports of what went on each day. Stern has admitted he took bribes, although he challenges the amount claimed. He will be sentenced on Monday - the trial started last Monday.

In the article, Michael Dell talks about how he wants to do business in "places with a legal system". I think that the Hu case, Rio Tinto's experience, and the events with Google would be making any enterprise re-evaluate the cost and risk of dealing with China.

Mod this up. If you haven't read Lockhart's Lament DO IT NOW.

I disagree. Most people are wired for maths, because they can talk and read. Broca's region is heavily involved in processing language, maths and music. Most people can handle language and appreciate music.

Can most people handle, let alone appreciate the theory of algebraic structures, tensor mechanics and multi-variate statistics? No. But in 30 years of working in maths education it's only those with profound issues that can't handle manipulating linear expressions, arithmetic, elementary geometry and mensuration. I'd go on to say that most people can handle the calculus of a single variable. But they have to be taught well, by someone who doesn't make it look hard, because it shouldn't be hard, maths is supposed to make sense, and what makes sense is easy.

I think there is some merit in the Professor's claims, but there has to be caution. Students need to be able to estimate measures, use measuring instruments, read clocks and handle money, all before age 10. These aspects of maths are suited to activity based learning, and can easily be embedded in other subjects.

But what of the kids who have the right brains to cope with more formal material earlier? What of the kids who cannot understand concepts such as zero or fractions without a more formal approach? What about how the retention of number facts is higher if we can get kids to engage with drill and memorisation of tables at early stages rather than later? How do we prevent the kids developing their own unusual understandings of fundamental concepts, because they have found a need in real life, and then we have to unwind their thinking later, because their constructed strategies only work in special cases?

I appreciate a lot of the results in maths education research. But there has to be great caution before we reject those practices that have worked for between 100 and 2000 years in favour of ideas that one or two research projects support. Is everything we do in classes effective? Certainly not. But until we can get class sizes down, better resourcing, attract more mathematicians to the teaching profession and get more individualised strategies working in the classroom we better be careful not to break what we know does work to some extent for the majority of students, even if it's not working optimally.

Note that this quote suggests that Viacom thinks SCO is right...

According to slide on page 22 (sigh), You Tube is a "video Grokster". So You Tube promotes discussion of abuses of the law by greedy litigious bastards whose business model hasn't changed with the times.

The problem with your argument is that in other countries it hasn't happened like your doomsayers predict. Here, as long as you can survive that long, after 12 months a medical insurer cannot knock you back for treatment for a pre-existing condition. If they do, the fines are horrendous, and a large part of that fine goes to you (or your estate if you die of your condition). This has been the case for at least 20 years that I know of, and the health insurers are still making profits, people are still buying their shares, people are still buying insurance, and people are still getting their conditions treated.

As an outsider, the spin that is surrounding this debate is like some kind of sick black comedy. The funniest thing is that so many Americans are buying into the weird agendas that predict the end of civilisation and the resurgence of Stalinism. Look around the rest of the world - these systems can work, which in turn results in a more productive and humane nation.