If you got an enum, you can declare a set of that enum, for example:

type
    TParam = (pAaa, pBbb, pCcc);
    TParams = set of TParam;

You can then easily check for membership: if pBbb in Params then ...

However the real power comes from being able to do set operations using + (union), - (difference), * (intersection) and the less than/greater than operators for testing for sub-/supersets, see the documentation for details.

For example, if you want to check if neither pAaa nor pCcc is in Params, you can do

if (Params * [pAaa, pCcc]) = [] then ...

[] is an empty set.

Have you seen Mr Plinkett[1] pick the originals apart? While the presentation is a bit weird, though funny if you like that kind of thing, his points are spot on and overall does a very good job of explaining why the originals were considerably better than the prequels.

[1] http://redlettermedia.com/plin...

Pretty much yes. The Red have better vibration tolerance, and the firmware is tweaked to fit a NAS workload better. For example, a Green will park the head as quickly as it can which for always-on machines can lead to a Green disk reaching its "Load/Unload Cycle" tolerance in months and die prematurely. The Red will not do this.

There's also a difference in how they handle unreadable sectors and such errors which makes the Red play nicer with hardware RAID controllers. An unrecoverable read error in a Green can cause the whole array to go down.

It's the same in Python, yet I haven't noticed it killing any Python programmers. Perhaps functional language designers are more fragile creatures.

I really have tried, but it's just too terse for me.

At this stage I'm pretty certain I won't grok a functional language until someone invents a functional language that doesn't look like some variation of Brainfuck or similar.

Indeed. And what's with this obsession amongst the functional language developers to try to use the least amount of characters possible? We're no longer in the 1960's where verbose source code took a lot of valuable space.

For example, when I read

fun double n = 2 * n

I see a function called double which implicitly takes a parameter n and returns true if n = 0, and false otherwise. Would it have killed them to separate the parameter from the "body", and used proper names? For example:

function double(n) = 2 * n

No you don't. The vast majority of people who watch "pirated" movies are online 24/7 with good enough pipes for streaming.

I've got Netflix and HBO (Nordic). Last sunday I wanted to watch the new XMen movie with my gf. Brick and mortar stores aren't open here on Sundays, so no chance there. And of course neither Netflix nor any other similar streaming service available to me carries it yet. But the pirated Bluray is out there, and it would have taken me just 15 minutes or so to download a rip and start playing it on my TV.

No, what I think they need to do is to make a "movie-Steam" Netflix hybrid. That is, a rich catalog of movies for a monthly fee. Then allow me to pay money for early access to new releases. Once I pay I get to keep access to that movie.

And by early access I mean when the first Bluray hits the street. Because that's when people can get the pirated version.

In that case one might be in for a disappointment. The article[1] seems to only contain about a page worth of equations.

[1]: http://arxiv.org/abs/1409.0917

It's also significantly more difficult to operate than a quadcopter, as well as more expensive.

The one scientists use today: invariant mass aka rest mass.

http://math.ucr.edu/home/baez/physics/ParticleAndNuclear/photon_mass.html

As I understand it, the problem is that the batteries must be at 0 degrees C to accept charge. The limited sun it's getting now isn't enough to heat the batteries (surface temperatures are about -70C IIRC).

That would depend heavily on the radioactive material used, no? For example, Wikipedia lists Pu238 to have just a ~16% drop in output after 20 years.

I'm guessing cost and weight were the key factors for picking solar over an RTG.

I think you guys should play some Kerbal Space Program and see just how much more fuel you need at every stage just to put an extra kg of equipment on that lander.

There are several things at play here.

One is the latency as mentioned, which is very important for VR. Heck even playing with a mouse and a regular monitor I can feel the difference between 60 Hz and 120 Hz, not to mention 30 Hz. At 30 it feels like my mouse is submerged in honey. At 60 it's decent but if you switch suddenly to 120 it you do notice that 120 is quite responsive in comparison.

Then there's also motion blur. Due to the way most LCDs currently operates, they introduce a lot of motion blur. This is beause the image is shining for almost the entire frame ("sample and hold"). The eyes tracks the motion, expects to find the object has moved but get's conflicting data because the monitor is still displaying the same old frame. This causes a perceived motion blur.

This is unlike "modern" CRTs for example, where the image faded quickly (within 2ms or so). For the rest of the frame the monitor was effectively black. That is much better for the eyes, and results in smoother perceived motion at the same framerate.

Even at 120Hz motion isn't completely smooth if you use sample and hold. Newer LCD monitors can strobe the backlight to get an effect similar to CRTs and thus reducing motion blur.

The LHC has about 115 billion protons in each bunch, and was designed for almost 3000 bunches at a time (I forget how many they ran before the shutdown).

Only a fraction of these protons collide, so there'll be plenty left when the beam quality is low enough for them to dump the beam and get a fresh one.