I believe the universe was created by an omnipotent, antisocial hedge-maze 3.5 years ago. Science can do nothing to prove me wrong. You actually *can't* know the truth in these types of questions, which is why it's called faith. Please stop confusing the two.

Oh, and when religious types can't define their faith in non-disprovable ways they are rank amateurs, and should be ridiculed for their ignorance. Attack ignorance, not faith. We'll all just get along better.

-Your friendly neighborhood theological non-cognitivist

Can anyone become a programmer?

Apparently.

Yes, this can happen. It's called Horizontal Gene Transfer, and it's been happening for a long time. I figure in another 25 years or so we'll discover that some disease we have newly discovered will have been caused by the introduction of some novel gene and then its horizontal transfer to some vector we hadn't thought about. I'm just hoping that whatever that disease is doesn't outright kill us before we get good enough at gene manipulation to design a counter to it.

In the business world innovation is only ever justified by some cost-benefit analysis. Yes, competition will drive new features, iff the return on those features is great enough. Patent and copyright laws are specifically intended to widen this profit making window with the intent of fostering innovation.

Decreasing the time you can monetize any new idea serves to decrease the monetary value of that innovation. Decreasing the value of innovations doesn't sound like a winning formula for fostering innovation.

When you figure it out, you should be able to apply the same term to any grade school course the subject of which they hand out graduate degrees. Math, English, Biology, Physics, History, etc. These beginner courses provide the lingua franca (programming languages) which we use to *begin* teaching Software Engineering (well the parts that aren't just pure engineering principles). If you want the academic, abstruse term, how about "Fundamentals of Algorithmic Expression".

Exactly. What they should have done is refuse to do any development until they could perfect interplanetary teleportation. Then they could simplify the design down to just the payload!

I believe the complaint here lies in the semantic ambiguity of "the test failed". Does that mean the test 1) was improperly designed such that it could not possibly prove that which it was commissioned to prove (test design was sound/flawed) or 2) was perfectly designed, the proof of which is that a failure mode of the device being tested was demonstrated (test device passed/failed)

As an engineer I take great pains to say what I mean as unambiguously as possible, because I know that failures happen quite often due to miscommunication. I read the summary above and am left in no doubt what happened, without even having seen the test results.

What I find truly interesting is how many people, upon seeing a test that passed all of the designed criteria say "the test passed" rather than either "the test was unnecessary" or "the test was insufficiently comprehensive". There are no flawless systems. If your test didn't result in some kind of failure you should immediately question whether you designed your tests correctly.

We already have <TuringCompleteLanguage>. All other languages are unnecessary.

So in theory I could acquire a good number of anonymous bitcoin and have my shiny new drugs drop shipped to an ex, or maybe some poor politician I disagree with. Or, I could just ship it directly to me and claim I was being targeted. Just, you know, *theoretically*.

'They will become popular when we get away from the word insects and use something like mini-livestock (PDF).'

Hivestock.

The canaries in this coal mine just started dying. It's news because there's about the be a cave in.

Way to totally miss the point. 14,000 Olympic athletes and 500,000 spectators are at the games. If the competitions were the point of the games they'd be spread out (in location and time) so the whole thing would be less of a logistics nightmare. Now whether you think that the Olympics are about cultural exchange or economics is up to you, but it certainly isn't about handing out 2300 medalions. Broaden your perspective.

Creating unusual object structures:
I once played around with a state machine framework that was object oriented c. It had a virtual table at the top of one base class and at the bottom of a different one. Using the same layout in the virtual table allowed multiple inheritance without any special effort and without any wasted space. It's the only object structure I've ever played with that I couldn't implement with C++ classes and inheritance (I guess a *really* good C++ compiler might be able to optimize to that structure)

Virtual Static members and member functions:
There are occasionally times when I need polymorphic behavior, but the behavior itself isn't instance specific. As a contrived example, imagine needing to query an instance of an object for the total number of peer objects that are in existence (I'm probably managing a count during construction/destruction). I need to call some member that is class specific, but that member will only need to use static members to execute. As it is this ends up being declared virtual and the this pointer is (needlessly) passed in.

Those are the only two instances I've run across where I actually wrote code up to the point of noticing that I couldn't do that in C++ (I'm actually still shocked ten years later that there's no such thing as a pure virtual static function).

Yeah, but sometimes you just get tired of typing the same boilerplate into every single function. I program in C most of the time and have dealt with at least three different "object oriented c" implementations. Every single one ends up with a this pointer (this/self/me) as one of the function parameters and every single one uses some variant of GetObjectFromMember for virtual functions:

void ExitSensorArrivalTimer_v_MSecTimer_expiry( MSecTimer *timer_self )
{
      Sensor *self = GetObjectFromMember(Sensor,m_exitSensorArrivalTimer,timer_self); ...
}

Sometimes a little language hand-holding helps limit the carpal-tunnel issues. This macro is also quite error prone (since C provides almost no type safety when casting pointers to and fro).

The tech to do this exists today. Capturing 3D shapes from standard camera/video data: doable: there are Youtube videos of people doing this ( http://www.youtube.com/watch?v=IquKgEyXo2g ). Generating the part files from the 3D model: difficult: there is a bit of art in turning 3D models into (efficiently) printable objects (hollowing out a bust to save material, or handling unsupported areas). Printing the part: moderate-easy: Depending on the type of printer you use (Fused Deposition Modeling is more challenging, Selective Laser Sintering is easier).

Now, it will probably be quite a while before anyone bothers to make it easy to go from one end to the other. I'm just saying that with a billion dollars you could go on your own personal crusade and make it happen this year.