A NTR requires the reactor core to be hotter than the exhaust gas stream/propellant in order to transfer heat to it. Anything over 4000 deg K, structures in the core are going to melt and that would be bad, and that limits how hot and how fast the exhaust will be.

Nuclear thermal is more efficient than chemical rockets but not that much more efficient. It can use readily available mass like cometary or asteroidal ice or gases like methane mined from Titan but if you have access to such sources then simple cryogenic fuel/oxidiser combos like LOX/LH2 produced from ice by solar-powered electrolytic plants are going to be easier to manage and less massive than a reactor-based rocket motor. In such a case the vacuum of space works to your advantage to keep the LOX and LH2 from boiling off too fast.

Don't forget the multi-megawatt radiators needed to provide a cold sink for those reactors. Chemical rocket engines dump heat into the exhaust gases but in a vacuum radiators have to be huge and heavy to get rid of significant amounts of heat from something like a nuclear reactor. They also have to be shaded from sunlight to stop them absorbing heat...

Nuclear propulsion reactors are quite a bit smaller than the current generation (no pun intended but I'll take what I can get) of PWRs and BWRs which typically start at 3GWt and can go up to as much as 5GWt (assuming the EPRs ever get finished and/or anyone commissions a full-scale ESBWR from Hitachi). Even the largest propulsion reactors like the Ford class CVNs at ca. 1GWt can be swung a lot faster as there's less fuel and heat to deal with. The very high fuel enrichment levels (RN Astute-class submersible cruisers use 90+% enrichment fuel in their Rolls Royce reactors) also helps as a major problem with swinging a low-enrichment PWR or BWR is the buildup of short-lived Xe-135 fission products in the fuel pellets. This isotope is very neutron-absorbent and causes problems controlling the the swing down and up again. It can be done and is done but it's not as simple as twiddling a dial on a control panel.

GenIIa reactors like the Russian VVER-1200 and the uprated French M310 designs can swing their output by 30% in fifteen minutes or so, given modern control systems and a few decades of experience in running such PWRs and BWRs. It doesn't happen often because nuclear fuel is so cheap and reducing power output doesn't save much money.

It depends if there's a production line for large components and a guaranteed market for future orders. The Chinese are rolling out 1GW reactors from breaking ground to grid connection over a period of about five years or so but they've got predictable orders of the large components needed for a reactor and teams of engineers who move from one site to the next as their particular tasks (pouring the basemat, building the containment, installing the reactor vessel etc.) on a given construction site are completed, they don't have to learn how to do it again from scratch every time. Rosatom is in the same position, building a number of reactors of similar design in Russia and around the world but also leveraging a turnkey operation capability, supplying fuel and taking away spent fuel for reprocessing and waste disposal which is very attractive to countries like Vietnam, Jordan and other Arab nations.

Ningde 3, a 1GW reactor on the central coast of China started construction with first concrete in January 2010 and achieved grid connection a couple of days ago, about 63 months later. Two more Chinese reactors of similar capacity are expected to come on line this year.

Adding landing legs to the first stage and not using all the fuel in the tanks, that could have been payload. Wait, what?

"The Dragon doesn't need any of those things. Its purpose is to take a few people into orbit and back"

Where are they going to go when they get to orbit? The ISS won't be there after 2020 or so. Bigelow is a lot of hot air. The Russian ISS-remnant will be serviced by Soyuz. So what's left for Dragon/Falcon apart from space tourism?

The Dragon alone can't build an ISS Mark 2 or even a Mars Expeditionary vehicle, it needs a workshop vehicle/microstation to dock to for the crew to do anything significant in orbit (and have a shower, use the toilet etc.). That's what the Shuttle was, as well as being the crew vehicle. A Falcon Heavy could launch such an unmanned microstation but they cost a lot of bucks so having it recoverable to be refurbished and reused would be a good idea. A heatshield re-entry system to splashdown on something that large would be cumbersome so tiles or another lightweight heat protection system on a lifting-body or winged vehicle would be preferable, like the X-37 or the new ESA re-entry testbed article flown recently on a Vega. Thus is the Shuttle reinvented, better and shinier than before with the lessons of the past learned.

Every Shuttle flight needed spacewalks, the cargo bay, the spacelab, the manipulator or other features sadly lacking in a people-only spacecraft like the Dragon until the ISS was ready for habitation. The ISS couldn't have been built without the Shuttle though, not without a (non-existent at the time) SLS that could throw a complete space station or large ready-to-go part of it with spacesuit airlock(s), manipulator arm, power systems etc. into orbit in one launch. Even then a "fork-lift truck" spacecraft like the Shuttle would probably have been needed to move shit around as more parts arrived in orbit -- the ISS is over 400 tonnes as is, nothing larger than about 16 tonnes in one piece.

Either a workshop/fork-lift spacecraft would have been recoverable to Earth as the Shuttle was or it would have been disposed of into the upper atmosphere and a new one built, furbished and launched every time a new task needed to be accomplished. The Shuttle was there to do that job, repeatedly.

Sure the Shuttle never succeeded in what it was originally designed for, nor was its launch tempo (50 launches a year was mooted at one time!), costs or other factors met. It did the job though when the ISS was built and when it was laid out in the early 70s nobody knew the ISS was going to exist at all. There's lots of Monday-morning quarterbacking about how things should have been but having the Shuttle to hand made things a lot easier when it was needed.

The Dragon has no airlocks, no space (heh) for spacesuits, no external cargo capacity in the service module (although they're working on it) to carry spare parts, no manipulator arm to tether and position EVA personnel around the Hubble or other large space infrastructure item like, say, an ISS Mark 2. It's a minimal spam-in-a-can meatbag-to-orbit delivery system, not a lineman's truck with a cabover as the Shuttle was.

The Shuttle's OMS fuel load could be maxed out to 18 tonnes if lots of in-orbit manoeuvering was planned at the cost of a reduced payload bay manifest. Most flights it didn't carry that much fuel but it didn't need to be rebuilt to take max fuel/oxidiser if the next flight necessitated it. Any Dragon plus disposable workshop mission is going to cost more and take longer as each workshop will have to be built and individually tailored to the expected mission's requirements. The other option is to build a son-of-Shuttle recoverable workshop/living quarters spacecraft, a bit like the autolanding X-37, but I don't see any budget for that anywhere.

How much would it cost to build and launch an 50-tonne "workshop" spacecraft to do the Shuttle's job and then ditch it into the upper atmosphere after every flight? A lot more than a billion a flight, never mind the extra launch of a manned capsule to dock with the Space Workshop module.

A recoverable and reusable spacecraft with the capabilities to do the same job as the Shuttle would need heat-tiling, some aerodynamic appendages to control re-entry and oh look! it's a Space Shuttle!

Saying that the Shuttle concept was laid out in the days when scheduling launches and docking in space was not as refined as it is today so the workshop was integrated with a manned "capsule" and living space and everything went up in one stack. One use (about the only repeated use I can think of in fact) for Falcon Heavy would be to launch a unmanned son-of-Shuttle "workshop" which would be recoverable to autoland on a runway for refurbishment and repurposing after spending a few months in orbit being visited by Soyuz/Dragon/SLS crews.

At the moment Dragon's only intended purpose is ferrying crews to the ISS but the space station is wearing out and its days are numbered. Once it is decommissioned then what? Dragon can only put meatbags in space, it can't do anything else unless there's somewhere for them to work and live. The Shuttle was an inelegant solution to that problem but it worked for 133 and a half flights. Sure it cost a lot but spaceflight generally costs a lot, thousands of bucks per kilo into LEO.

The "space truck" was actually a complete space station. It had living space for seven people, airlocks for EVAs, a shower and a toilet as well as having 20 tonnes of cargo space in the back of the "truck" and a payload arm/manipulator.

The Shuttle had considerable cross-range capability once in orbit with up to 18 tonnes of manoeuvering fuel (twice the total payload of a current Falcon 9) and could stay in orbit for up to a month if needed with a reduced crew. It did most of the heavy lifting of the construction of the ISS in orbit and carried out multiple Hubble repair and upgrade missions. At the end it came back down to Earth and landed on a runway.

The Dragon capsule is purely for canned monkeys with no toilet, no shower, no airlocks and no EVA capability. It has no cargo capacity, no manipulator arm, limited cross-range capacity in orbit and limited endurance and it certainly can't be used to carry out maintenance flights to the Hubble or its successors.

There's a version of blindfold Go where both players use the same colour of stones. They can see all the stones placed on the board so it's theoretically still a full-information game. They remember who played which stone or they can work it out from the pattern.

"Amazon has close to no overseas presence outside of English speaking countries."

Well, apart from France (75 million people), Germany (80 million), Japan (120 million), South Korea (50 million), Spain, Greece, Italy, the Scandinavian countries, Holland, Belgium etc. etc.

Add them up and you'll find the populations of Amazon's non-English-speaking markets are way larger than the English-speaking nations. The total number of customers and total sales may be lower -- Japan, for example has Rakuten/Tenso as a serious competitor to Amazon.co.jp for online sales -- but they're out there and selling to anyone with a credit card and a keyboard whatever language they speak.

Raw uranium ores are a lot more radioactive than pure uranium oxides like yellowcake (U3O8) because of all the shorter-lived isotopes that have built up in the ore bodies from a billion years or so of decays of U-235 (700 million years) and U-238 (over 4 billion years). The other thing is that solid lumps of uranium are a good shield against radiation and the alpha particles resulting from decay events a millimetre or two under the surface are unlikely to escape the lump of metal and be dangerous.

Don't forget Japan which has delivered cargoes to the ISS using their home-grown launcher. They also launched a spacecraft, Hayabusa deep into the Solar system to rendezvous with a comet and return particle samples back to earth. The Hayabusa-II followup mission launched in December 2014 and it plans to return samples of an asteroid as well as landing three small hopping "rovers" on it for close-up study of the surface.