jnw2

Model S makes use of SAE J1772 only via an adapter, and it's not clear that the average Model S owner is going to make frequent use of J1772.

Tesla is recommending that Model S owners either install an RV style 50A 240V outlet in their homes, or else install a charging station that has the Tesla connector that is currently only used on Model S. The Superchargers also do not make use of J1772.

If hotels end up installing J1772 and not Tesla charging stations, then perhaps Model S owners will use the J1772 adapters at hotels.

If you're driving 500 miles a day two days in a row to cover a total of 1000 miles, can the early adopters really expect to find a suitable 240 V, 30-50 A charging station at their hotel, or are they going to resort to stopping at a supercharger and not charging at the hotel?

There are some interesting questions about economies of scale, etc, though.

If we figure an average American has a lead acid car battery and maybe a laptop battery pack, we currently have more lead acid batteries than lithium ion, certainly by weight, but probably also by total watt hours.

Tesla and Nissan might end up inverting that ratio in a market where they merely have to get the battery to beat the cost of gas; and if we get to the point where every American has an 85 kwH battery pack in their car, that's multiple days of average US electrical consumption (I believe average per capita electrical consumption is about 1.5 kw, and average per capita total energy consumption around 6 kw in the US).

Meanwhile, there's no path to pushing up volumes of lead acid battery production significantly. Maybe there will be a few old lead acid car batteries getting recycled after Tesla conquers the world, but if that's all we're relying on, those recycled batteries won't power very much in the grand scheme of things.

IIRC, solar prices have been dropping at a rate of 7% per year. If that keeps up, and if the grid price remains constant, solar will eventually become cheaper than grid.

In 1992, would you have told me today's smartphones would be impossible?

If you only lose a few solar panels in a storm, and most of the panels on your roof survive, you may just use a bit less electricity for a while.

Several strings of solar panels + lithium ion batteries might very well be cheaper than paying your local electric monopoly for transmission line capacity in 20 years. And if the non-redundant parts (the inverter, perhaps) fail, it might not be all that different from your water heater failing today.

(Although if we had that technology in cheap enough form, some of the major loads in your house may switch to DC to avoid conversion losses to AC, since solar panels and batteries are both inherently DC technologies, and that might make the inverter less important.)

Mass transit has some advantages with predictable trip lengths: no transit agency is going to insist that they need a bus that can go 200 miles on a charge if their average bus only travels 50 miles each morning. Meanwhile, the average consumer ``needs'' car with a battery that is several times what their 99th percentile trip length is.

I bet batteries don't tend to like going from full to empty in the time it takes to launch a plane.

I wouldn't be surprised if later in the Ford class's life, we do see supercapacitors taking over.

http://en.wikipedia.org/wiki/A1B_reactor indicates that there's 300 MW per reactor on the Ford class, and two reactors, so maybe 600 MW total, although I'm not sure if that's the output of the steam powered electrical generators, or if it's the theoretical amount of energy in the steam coming out of the reactor.

http://en.wikipedia.org/wiki/Miles_per_gallon_gasoline_equiv... says the EPA figures 33.7 kwH is the same as a gallon of gasoline, and there's enough slop in estimating how much power a carrier will have left over for making jet fuel that the error in pretending that jet fuel is the same amount of energy per gallon as gasoline is probably noise.

Other commenters are saying that synthetic fuel is going to take 2-4 times as much energy as the synthetic fuel stores, so let's assume roughly 100 kwH to make a gallon of jet fuel.

That suggests that 600 MW total power might be able to make as much as 6000 gallons per hour if people are happy to leave the carrier drifting with no lights and all its defensive equipment turned off. If the carrier is carrying 75+ planes, that suggests it can make less than 100 gallons an hour per plane. The Google Search summary of http://www.google.com/url?q=http://wiki.answers.com/Q/What_i... says a Gulfstream III consumes 568 gallons per hour. A supersonic fighter jet probably consumes somewhat more, and that leaves me wondering if a Ford class reactor is going to be able to produce enough jet fuel for an active fleet of fighter jets. Certainly, the planes don't fly 24 hours a day, but this estimate suggests that the carrier's reactors might not even have enough left over power to make enough jet fuel to have the average plane on board flying one hour out of 24.

http://en.wikipedia.org/wiki/USS_Gerald_R._Ford_(CVN-78) notes that ``Unfortunately the power limitations for the Nimitz class make the installation of the recently developed [Electromagnetic Aircraft Launch System] impossible.'', which seems to indicate that existing nuclear powered aircraft carriers don't always have large amounts of electricity to spare. However, more electricity will be available on future aircraft carriers.

I'm not in the bay area.

Given your constraints, I wonder if what you actually want is one switch for each upstream, and a gigabit link from each of your physical Xen servers to each switch that goes to an upstream. If your switches have to route at all, they can each have a default route to the one respective upstream connected to that switch. Then, have a VM on each physical Xen server with full routing tables to control how the outbound packets from all the VMs on that physical host get routed. And hopefully your switches can deal with a /32 route for each VM or something, if they just have a route per VM plus a default route to the one ISP they're connected to.

There's also the variant where you have a /24 block (or something) which has one IP address which is your ISP's router, and then one IP address per physical Xen server, and if your ISP is willing to take individual /32 announcements to your servers (I think this is technically feasible, but may require more mental flexibility than some ISPs have; on the other hand, I think you're talking Cogent, and mentioned that Cogent is willing to peer directly with your customers, which may imply that you're dealing with a flexible ISP), you can have the ISP do the layer 3 routing to the right physical host for traffic going to each VM. (The idea is that the ISP would not propagate the /32s to their peers; you'd separately send them a broader announcement that they could pass along to their peers.)

One other issue in this scheme is that if a single gigabit link between one physical Xen host and the switch going to one ISP breaks, inbound traffic that happens to come to the ISP attached to the switch with the broken link ought to have some mechanism to find its way over to the other switch for the one physical Xen host that has the broken link. I think there are ways that OSPF and or iBGP with the non-default next-hop-self setting can be made to do this.

Elon has five children. Getting Elon + 5 children into a Tesla is a key goal, and presumably his smallest child is going to outgrow the optional rear facing seats in the Model S at some point...

A Xeon E3-1220 has an 8MB L3 cache and costs about $210. If you deaggregate the entire IPv4 routing table to /24s to support fast lookups and store one nibble of destination data per route (probably just a destination interface number, if you can live with only 15 or 16 possible destinations, but you probably don't have that many upstreams), doesn't that mean that the entire routing table can fit in the L3 cache on the CPU (with a bit of the L3 not used by class E space, which will leave a bit of space for code, assuming you aren't doing anything but routing on this CPU)?

Now, it's quite possible that the Linux kernel's routing isn't dimensioned to provide this sort of performance, but if you want to avoid CAM in the IPv4 world, I don't see why you'd really need CAM...

(Also, I haven't looked carefully at how the L3 cache actually works on the E3s to make sure that it's shared between code and data, etc etc etc. Also, if you're actually implementing this, you presumably do want to include class E in the table, and then hope/assume that when class E doesn't get used, the CPU will be smart enough to cache the code instead.)

How well would the average US city cope with providing enough tap water for its population suddenly tripling?

Improving medical technology with a goal of curing currently-uncurable diseases, making health care cheap enough to make Medicare sustainable (perhaps specifically by looking for ways to eliminate 90% of the costs associated with the half dozen or so families of diseases that Medicare finds most expensive), and improve quality of life with better treatments.

Better transportation infrastructure, especially high speed rail. (Notice how the French have better track than the Germans because the Germans put off investment while playing with maglev; Elon Musk might be at risk for repeating the same mistake in the US with Hyperloop.)

Fiber optic cable to just about every building in the world, owned by people who charge at a rate that isn't much higher than the construction cost and don't try to keep charging more and more for higher data rates.

Possibly some more good parks.

We need cheap solar and batteries too, but I think those things are likely inevitable given enough time at this point.

Did Carter show anywhere near as much leadership in pushing the limits of technology as Admiral Rickover did, though?

If we had one of the surplus Hubble-diameter Naval Reconnaissance Office mirrors orbiting Mars with a suitable camera, would we have better pictures than this?

And could one of those mirrors be mounted in the trunk of a SpaceX Dragon spacecraft?

I'm not sure the EV charging at night thing will continue forever; Germany has already had a Saturday where their solar panels covered 50% of their electricity needs for a few hours, so if they double or triple or quadruple their solar capacity, they may start finding they need EVs to soak up all the excess solar energy in the middle of bright days.

Solar panels seem to keep getting cheaper; distribution via the power grid does not. This may lead to property owners with enough land eventually finding an economic incentive to get most of their power from solar panels on their land instead of using the grid.

On the other hand, Tesla is busy getting luxury car buyers to help subsidize the capital investment in building storage on a bigger scale, and nobody has figured out how to get luxury car buyers to subsidize grid improvements.

In the long term, since the grid is a monopoly that is not eager to drop its price, whereas solar and batteries are likely to keep dropping in price, in part due to competition and in part due to appealing to early adopters, I expect that a few decades from now, the typical homeowner will use mostly solar and batteries instead of paying the monopolies' higher prices.

http://googleblog.blogspot.com/2009/07/should-you-spring-cle... indicates that solar panels mounted flat do indeed need to be cleaned, but panels mounted at a sufficient angle will be rinsed off adequately by the rain and do not seem to benefit measurably from manual cleaning.

Would the founders of a startup who are willing to pivot as needed get something lower than the highest possible grit score?

I was under the impression that Amtrak from Newark Airport to 30th St Station in Philadelphia was available as an airline code share (and that train segment is fast enough that there's no good reason to get on a plane for Newark<->Philadelphia, even if you are connecting to another flight out of Newark), but I haven't figured out how to find such a code share with Google's flight search tools.

(The governor of IL and the Chicago mayor also asked Amtrak to study how to extend some of the Amtrak routes to O'Hare, presumably to replace some ``commuter'' flights, and they wanted the study done by the end of this summer, but I haven't seen any evidence of the study being done yet.)

I've certainly seen virtualization environments where RAM is a much larger fraction of the physical server's cost than the CPUs. How much RAM are you using when you aren't using many CPU cycles?

The Americans with Disabilities Act may not have much sympathy for the inconvenience of providing accessibility, though.

Is there a good argument against Oracle releasing ZFS under the GPL (so that it could be integrated into the Linux kernel)?

IIRC, a post to NANOG a while ago claimed that the cost to run fiber to the home is well over $1k/household. I think the math works out to the conclusion that neither Netflix nor Google has enough money in the bank to run fiber to a large fraction of America.

So the question becomes: if you get fiber to 1% of America, is that going to be enough of a demo project to convince someone to come up with the money for the rest of America?

Did the local electric utility decide it wasn't worth running 120V power to your home either? I don't know of any reason that running fiber is any more expensive than running power lines (if anything, being able to piggyback the fiber on existing poles should make it cheaper).

There are also airplanes without radios (though such planes normally operate at untowered airports).

http://www.avweb.com/news/sayagain/191072-1.html has a (now retired) US air traffic controller's thoughts on a fatal collision involving two of the three planes a European controller was responsible for. It suggests that not all cases of sleep deprived controllers are harmless.

But maybe an electric car needs a gasoline engine as much as a tablet computer needs a keyboard, and maybe we're several years away from seeing that in the electric car case, just as the belief that a tablet didn't need a keyboard didn't become widespread until the iPad.

Even lengthening existing Amtrak trains can take years of time and bureaucratic approval. The design of the Acela trainsets is such that Amtrak could add four more business class cars, which would double the number of business class seats per trainset, while keeping the current power cars.

Transportation improvements over time tend to make some trips possible in much less time than was previously possible; this is true of the automobile, the Interstate highway system, the jet airplane, and the Golden Gate Bridge. Nobody has demonstrated a 150 MPH bus running on the existing Interstate highway network yet. High speed rail has the potential to increase the distance that can be covered in a 60-90 minute commute.