OK, people are "addicted to MOOCs" much as people are "addicted" to TV or to the Internet.

How does this help me teach people to be engineers?

Then you cannot patent some linkages either.

The Bennett mechanism is a highly overconstrained "mobile spatial four-bar linkage." (Yeah, yeah, the Bennett mechanism is 100 year old prior art, but I am saying some new discovery yet to be made of a linkage.) What makes something a Bennett mechanism is a precise relationship between the link lengths and the link "twist angles." So a Bennett mechanism is not the materials nor whether you use roller or journal bearings nor the thickness and shape of the links. A Bennett mechanism is essentially a mathematical relationship.

That the Bennett mechanism is mobile instead of a rigid "tensegrity structure" has to do with some deep mathematical relationships that may yet yield to a simple proof.

What do you have against Ryan Seacrest?

The man doesn't have any pretensions that what he is doing isn't just a job and isn't anything more serious than entertainment.

Even for the home power user, it's a pain if you are forced to buy brick and mortar and on a deadline because the family needs a new machine. Old-style PCI has become a pain to find on cheap towers.

I had to shelf a bunch of cards when our old ~2003 machine died 2 years ago:

• an old $120 PCI video card (the cheapest replacement costs ~$60 these days)
• a Soundblaster card (I miss when they used to have MIDI and joypad ports)
• a potentially secondary NIC that would have helped me play around with work multi-nic setups and software firewalls
• an adapter card for 3 Firewire inputs.
• a modem card (didn't get more than a couple uses after we went to DSL, but can save you a dollar and a trip to the Western Union when your potential new boss requires documents faxed before you're hired)

An old system I used to maintain had 2 native USB port and about 4 more via a PCI card I added. I think my switch from the hand-me down ISA to my self-bought PCI setup had only 1 card that got dumped. We just have so many more gadgets today.

A few cents or dollars per NEW module kills timely standard adoption. We're talking about ISPs, so let's use a well-known evolutionary example with WIFI routers available to users even outside the ISP chain:

First, no wifi at all,
then default / empty passwords all neighbors could steal,
then WEP only because WPA wasn't supported,
then no WPA2...
then (or mixed in with the above):
no support for G,
then no support for N
finally, "support" for N on just 130mbps, but not multiples of it. The unwritten word is also SINGLE band (2.4Ghz)

That is what I remember from a ton of different routers I either got from ISPs, owned, gave away or just troubleshooted. The great fragmentation tells you that it won't be an easy problem to solve. I mean, just check your Wifi now and see how many of the ancient no-nos you can still see from neighbors around you who PAID for their routers --I don't even want to know what they have to settle for at the Modem level.

Providing an upgraded router may not be the same as just "going" up to an IPv6-enabled router. Supply chains take forever (5 years) to provide today's optional features.

If you need more proof that a 2 year cycle for upgrades means nothing, just look at how few top of the line smartphones *refreshed yearly* support 5Ghz bands. Even if you paid through the nose to correct that, you still must leave the 2.4 Ghz band open because your pricy game console [refreshed every 5 years] isn't that lucky or your visitors' gadgets are behind. It's not a pretty picture. Give it 10 or 15 more years

Exactly. This isn't a major, it's a class at most. I'm a Mechanical engineer and took a few CS classes as electives.I was the only one in my class of CS majors that would unit test. I wrote script upon script to beat my projects to death. Consequently I also managed to get one of the highest grades in the class. This was back in 2003 when "CS" meant "I like computers" but there were numerous people in my class that would turn in half assed work.

Even at work where I use Matlab I try to test every single scenario possible in my scripts. Especially the stuff that I put on Matlab File Exchange, for example: https://github.com/jedediahfrey/matlab_saveppt2/blob/master/Test_SavePPT2.m

Even if they did I would be interested in how they are going to backup their data and somewhere down the time track run a disaster recovery scenario. Oh I would love that contract. :)

I dispute your assertion that the auto companies "grade themselves" on gas mileage.

As far as it is known, every model of every car maker has to go to Ann Arbor for the original "EPA" drive cycle -- this is for smog control certification, and you extract the EPA City numbers from that, and I think the old "Highway" test is included, both tests used for CAFE regs.

EPA has a new or extended cycle including the high speed highway driving used for the current window sticker. For most cars, they get the "new" sticker numbers be a formula applied to CAFE, but for selected cars, they or the car makers run a full test on them.

With respect to cheating rather than optimizing for or "teaching to the test", EPA relies on the car makers to conduct coast-down or other tests to come up with the drag coefficients to put into the chassis dyno in Ann Arbor. Folks should look up the Test Car List Data (Google for it) for the raw EPA results along with the drag coefficients for the different car models. I am thinking that if we "crowd source" that task, there may be "interesting" findings.

This business of EPA numbers and whether they can be achieved or not achieved generates such emotions.

This is Slashdot. We are geeks. We embrace the scientific method, which includes experimental control.

One knock on the EPA is that they "do a lab test that is never reproduced out on the highway."

What I am saying or suggesting is that driving 55 MPH on a calm wind day without the A/C is a good proxy for the EPA Highway test. I am not saying you have to drive 55 or should drive 55, it is just that there are some stretches of road with a 55 MPH limit that would be a good place to test a car.

So Question 1, what gas mileage do you get driving at 55 MPH on the highway under calm wind conditions, and how does it compare with the EPA test. It's called "experimental control." That is, you first try to see if a car gets anywhere near the EPA gas mileage when driven under as close to EPA conditions as is reasonable for someone without a chassis dyno and an exhaust gas analyser like they have in Ann Arbor, MI at the EPA testing station. If the C-Max and Fusion Hybrid are not getting their EPA numbers, and I am talking the raw numbers and not the "adjusted" numbers on the window sticker, then there is a case to be made that Ford is cheating.

Then Question 2, what gas mileage do you get, say, driving 65 MPH as in the Consumer Reports test? If the C-Max and Fusion do OK at 55 MPH but are not so thrifty at 65 MPH, maybe Ford isn't cheating, but they have optimized those cars for 55 MPH when most people drive 65 MPH (or faster) on the highway. This is an important distinction to us geeks grounded in Experimental Control and the Scientific Method.

So them my Question 3, why doesn't Consumer Reports conduct a road test that best replicates EPA conditions before they start knocking cars for not "living up to the EPA numbers." I don't care about the song, "Pull my license, and all that jive, I . . . can't . . . drive . . . 55!" For cryin' out loud, doesn't Consumer Reports believe in Experimental Control? Report a highway number at a constant 55 and then report a highway number at a constant 65.

Question 4, so why did Consumer Reports change their "highway" road test mid stream? I think they sped up their highway road test to be more consistent with the way their readers drive. Fine. But then you have an apples and oranges comparison between a car built now and maybe a car you owned 30 years ago. Suppose I want to know how a Prius stacks up against the low-tech non-hybrid 1.6 litre 5-speed manual-shift Nova/Corolla I had purchased back in 1986? Can't compare by going through old stacks of Consumer Reports.

A lot of people are singing to tune to "I can't drive . . . fifty-five!"

So one level of complaint is that "I get nowhere near the EPA mileage because who drives like that" apart from hyper-miling geeks, old persons, and dudes with a passive-aggressive anti-social attitude. I can see the point, however. Suppose I drive (legally) at 70 MPH and the car is optimized for 50 MPH, how do I read the EPA sticker to find a car that gets OK mileage for the way I drive?

The next level of complaint is that "I am a hypermiling geek, and even I don't get the EPA mileage."

So the question is, is the EPA test even accurate on its own terms?

My limited experience with a Taurus and with a Camry and using a Scan Gauge, if you drive about 10 miles on a 70-deg day without the AC at an average speed of 20 MPH in traffic, you will get better than the derated window sticker and will be within 5% of the "raw EPA numbers." If you drive a constant 55 MPH on a calm wind day, you will get the raw EPA Highway numbers.

The Taurus started to not get the EPA numbers. It turned out to have dragging brakes from rusted caliper pins.

I don't doubt there is cheating, however. I look over the EPA "Test Car List Data" and have seen some fishy drag or coast-down time numbers.

But are people experiencing that if you drive like Granny, some cars get the EPA and o/thers don't? Are there variations between cars of the same model and year, that some engines are manufactured "tight" with much friction and other cars have a natural aptitude for MPG?

You know that some cities still haven't timed their lights, right?

I have a HP dv7 (2.5 years old with 8GB RAM) running Fedora 18 with the latest updates as per a week ago and my Firefox is version 20. The demo definitely does run under Firefox although I did have to do two refreshes before the demo started.

I ran the performance test and I got about 17 frames a second in full screen mode which is not that good but it is still viewable although a bit jerky at time. I did like the refection on the tiles in the church and the water effects although I have seen similar and better in other games of this type. The thing that has me worried when viewing my performance monitor is that the CPU temperature showed about 95 deg C which considering that boiling water is 100 deg C means that I will be sending my dv7 in for repair before my warranty runs out.

Still for something like this to run on a browser under Linux IMHO is quite good and may level the playing field between desktop OS's since basically if you are a PC gamer you have to run your "Games for Windows" under a Microsoft OS (I do know about WINE but IMHO it is a kludge for a kludge anyway). Still I actually prefer console gaming to PC gaming and no amount of arguments can convince me otherwise.

On a lighter note where are all the NPC's and there has to be a dragon somewhere or at least some Orc's (maybe they were doing their washing) :)

Interestingly, your post made me think that google glasses might be a foot in the door techwise for something 100 times dorkier and less public: Virtual Reality Googles.

Those guys playing Temple Run today don't yet realize some augmented reality version may be coming to glasses near them

The 3D is cheap
The immersive stereo view is built into the nature of glasses
The GPS and compasses are already in all smartphones and there's several apps that act like smart HUDs.
It's portable and more natural than a headset, and if you [Google] builds it, [someone] will come with implementations.

I just don't want to see people looking like druggies when they extend their hands to touch things that are floating in their HUD. The eerie effect of talking into thin air left Behind by bluetooth is already sad enough

Actually we didn't have a clue has to who the Ferengi "were supposed to be" apart from Lt Cmdr Data calling them "Yankee traders."

And then actor Armin Shimerman came along with his scenery-chewing rendition of Quark on DS-9, portraying a Ferengi as an angst-driven discriminated-against entrepreneur, evoking memories of a character in, dunno, Shakespeare's "Merchant of Venice" ("does not a Jew bleed . . .").

And then the Ferengi became an ethnic stereotype -- superintelligent for having a four-lobed brain, but adhering to an ethical system known as the "Laws of Acquisition" that are studied with an almost religious devotion, well, you kind of get the idea.

Look, just because Jewish people were involved in the creation of Star Trek doesn't mean Star Trek doesn't reinforce certain stereotypes, maybe by Mr. Shimerman having a blast with his campy character.

I find it extremely creepy that while they responded badly to this, he had exactly what we are afraid of: good quality pictures / video of your environment at all times. I don't agree with the violence. It's just going to be harder to paint ourselves into corners to avoid connecting with everyone else's field of [perfectly recorded] vision. If Glass takes off, it will be like walking around trying not to step on anyone's shadow: impossible.
Here's a stray thought: What will having so much visual info do to people's social networking feeds?