The problem is that "any flammable liquid" would cost more than gasoline (absent tax). They demonstrated the Chrysler turbine car running on perfume, for example. Back in the 50s gas was well under a dollar a gallon -- how much would a gallon of perfume cost in comparison? Kerosone (aka JP-1)? About the same price but with less calorific energy per gallon so it would provide less mileage than gasoline. Coal gas may be cheaper per mile but storage and distribution and tankage in the vehicle would be problematic and it's a dirty fuel to produce. Natural gas is getting pricier by the day so that doesn't help much.
The Japanese superconducting maglev system is being designed to run in service at 550km/hr. The test vehicles (which often carry passengers) can achieve this speed pretty much at will, it does not require special versions of the trains. The test area in Japan has two side-by-side tracks about 40km long (longer than the Chinese Beijing airport maglev) and they have successfully run two trains past each other at a closing speed of 1100km/hr, something that may cause problems with super-high-speed steel-wheel trains given their existing rights-of-way have no blast dividers between the tracks.
There's a lot of folks like yourself that figure third-world manufacturers work with stoop labour in candle-lit shacks to outproduce the super-efficient West. It's just not so.
Any new production technology, any new machines or processes that permit manufacturers to make stuff cheaper or better can be bought and used in China or elsewhere in the world for the same ticket price as it would cost in Western countries. For an example, steel-making plants in places like Indonesia and Malaysia were built in the 1970s with the most modern process control systems available at the time allowing them to replace imports of Western steel with a locally-made cheaper product. The end result was the death of steel-making in the West, in part because there was little iron ore left to be sourced locally there. Since it costs the same to make the steel in either place the price was set by transport-to-market and it's cheaper to ship finished steel than it is to ship iron ore and process it in the West. A side-effect of this was a decrease in shipbuilding in the West as Far Eastern shipyards blossomed using modern shipbuilding techniques and design methods building lower-cost ships with cheap locally produced steel.
The same thing is true pretty much in other industries such as electronics or car manufacturing. The production lines in places like China or Taiwan use the same equipment, the same components, the same quality control processes as any business in the West does. These machines and processes weren't stolen, they weren't secret, they were commodities just like a box of computer memory, sold to a customer with cash to pay for them.
The US has been using up its existing stockpiles of Pu-238 to build RTGs for a mixture of civilian deep space projects and black intel operations such as non-solar-powered stealth spy satellites and seabed-emplaced submarine monitoring stations. The Russians agreed to sell the US some Pu-238 under a licence that prevented it being used for military functions but they shut that down when it became obvious the US was reallocating most if not all of its home-grown stockpile to the military side of things. Like oil Pu-238 is fungible and the Russian supply of Pu-238 was effectively enhancing US military capabilities.
The T221 display had crap colour definition, low contrast especially off-axis, slow refresh and it required four video channels to drive it. Video and photo reproduction wasn't its forte though, what it was great at doing was putting a lot of numbers and graphs in front of someone's eyeballs. It mainly sold into scientific and engineering environments (CAD, PCB and chip layout etc.) and financial trading houses but it was never aimed at Joe Public.
You see them turn up on Ebay occasionally, still commanding decent resale prices. If you do buy one make sure you get all the bits like the special cable harnesses etc.
The US has several off-the-shelf medium/heavy lifters such as the Delta 4 Heavy that can put up to 20 tonnes into orbit similar to the Ariane V. What they don't have (and nobody else has) is a superheavy lifter capable of carrying a 70-tonnes plus payload which is needed to perform the one-shot-to-the-Moon mission envisaged for Constellation (with a separate crew flight). However there are problems man-rating an existing lifter; the flight profile needs to be configured so that the maximum acceleration at any point in the flight is tolerable to the Spam-in-a-can plus a lot of other factors such as safety and abort flight modes and hardware mods.
ESA is preparing to buy and fly Soyuz spacecraft from their Guiana spaceport, initially to carry unmanned payloads such as the Progress ISS supply capsule. The Soyuz design is already man-rated and well-proven (over 1700 flights) and it wouldn't take much upgrading to add a manned spaceflight capability to the ESA catalogue based on the Soyuz.
The Falcon 9 did not nail its first flight. The upper stage ended up in an elliptical 240x280km orbit rather than the planned circular 250km one. The upper stage was also yawing and rolling noticeably after orbital insertion -- I watched the web broadcast of the launch and I wondered if that was deliberate or not, it turns out it was not. Hopefully they'll fix those problems in the next few test launches.
I can't see NASA or any of the other participants in the ISS programme letting a Spacex vehicle make an approach to their fragile space station unless they are sure it's not going to crash into it.
Cheap ROM has a read-cycle time of about 40-50nS and works OK with CPUs running at under 100MHz. There's a lot of embedded kit out there for which that is a good solution, although today most low-end embedded kit uses EEPROM or flash rather than mask ROM for production-line flexibility. Modern DRAM in the DDR2/DDR3 class has read cycle times in bursts of less than 1 nS. For today's console or PC based games that sort of speed is required to keep the gameplay moving. ROM cartridges would work OK for simple sprite-based TV resolution graphics as they have done in the past. For more modern visually-realistic games they are a non-starter.
I'm just back from three weeks in Japan, travelling around and staying in salaryman hotel rooms (well, apart from the suite upgrade at a resort hotel in Atami when they didn't have any single rooms left). Nearly all these small business hotels have free cabled internet which is pretty fast and reliable, a few have rather ropy low-signal wifi with frequent dropouts and reconnections required. In one case I found myself on an 11MBps connection, probably ancient 802.11b hardware.
I much prefer cabled connections to wifi when staying in hotels. Many hotels supply wifi by sticking a router on every other floor in the building as a minimal option. Doing it right with a hotspot in the ceiling over every room is going to be a lot more expensive for them to implement and most of them will not bother.
There are other advantages to air launch -- no large pad structures required, blast protection systems etc. as the motor only ignites in free air. The payload rocket can be built and handled horizontally throughout its entire lifespan until it is actually fired in flight etc. Adding the extra fuel tankerage to give the rocket that extra 10-12% is not a trivial matter; it needs extra strengthening to carry the extra mass on the pad and under acceleration which adds to the all-up launch weight requiring more fuel and oxidiser to carry it. That extra structure all costs money that is thrown away on every launch, never to be seen again.
As for building a gargantuan aircraft, it's not difficult to do with today's engineering knowledge and materials. It wouldn't even be that expensive to do since it's not meant to be a cost-competitive commercial airliner design built in the hundreds to carry fare-conscious passengers and air freight. A brute-force design would do the job -- big strong wings plus lots of engines for thrust. It doesn't need endurance to fly for ten or twelve hours at a time, it is not fine-tuned to carry passengers and freight at minimum operating costs, its fuel consumption per kilometre flown doesn't need to be cut to the bone because of marketing constraints. There would be at most three aircraft in the fleet, two minimum to maintain a schedule of launches but the project could start with just one so no need for a production line and the expensive engineering optimisations that requires.
There are efficiency advantages to launching rockets into space from aircraft.
A: Rocket motors work best when there is little or no air pressure for them to burn into, increasing the jet speed and the Isp figure. Solid boosters benefit more than liquid fuelled rockets as the exhaust pressure is less. At 35-40,000 feet the air pressure is about 30% of that on the ground, making rockets a few percent more efficient at that altitude compared to sea-level.
B: Existing land-based launchers have to spend a significant chunk of their fuel and oxidiser getting to the velocity and altitude an aircraft launcher has already reached using conventional Jet-A fuel and atmospheric oxygen. This is why there are no single-stage-to-orbit launchers, they all have to lift off heavy and throw away sections of their airframe on the way up so the remaining parts can make it to orbit.
C: The addition of 1000km/hr horizontal velocity from the launcher is, as you say very much less than the orbital velocity of 25,000 km/hr required for LEO, but it's 4% which is worth having.
A rough BOTE calculation suggests an air launch to LEO can be done for about 10-12% less fuel load than a similar launch from the ground. That parlays into less parasitic airframe, tankerage etc. that has to be flightworthy and capable of withstanding extended multiple-G accelerations and which requires fuel and oxidiser to get it off the pad.
As for size limitations on air launches that's due to a lack of really big off-the-shelf aircraft capable of being modded for use as launchers. If there was a determination to build an air-launch system for getting, say, 20 tonnes of payload into LEO per launch then designing and building a couple of dedicated carrier aircraft to do the job would be easy. They don't have to be efficient fuel-misers the way modern airliners or even military aircraft are, they can use multiple off-the-shelf aero engines to make up for the lack of efficient design and excess weight as their fuel costs are going to be a minor part of the operating TCO. My own thoughts on what they would look like tend towards a land-launched catamaran high-wing seaplane with the orbital vehicle suspended between the hulls (a bit like the Virgin Eve carrier aircraft).
"1. LED backlit screen. This seems strange given the existence of PixelQi and mirasol type LCD displays."
The PixelQi panel is not available today in quantity ten million unlike the commodity low-resolution display designed into the iPad which is shipping now and has been for the past few years. The PixelQi panel is in high demand which means it costs a lot and that would cut into Apple's profit margins. Apple can't sell iPads at a thousnad-buck pricepoint hence no OLED displays.
"2. No camera? What happened? This is very disappointing."
Power and cost. The iPad has a 25Whr battery and it needs longer runtimes than a laptop to differentiate it from the existing market in laptops and especially netbooks. Adding a camera eats into the power budget. It also adds five or ten bucks to the BOM and that cuts into the profit margins.
The Japanese superconducting maglev trains which are currently under development hold the manned record for such trains of 580km/hr and they regularly achieve over 500km/hr, the target for production trains during tests. More impressively the test route (which is longer than the Shangai airport maglev link) is double-tracked and they have run a pair of trains past each other at a closing speed of over 1000km/hr.
Actually the max resolution on a 22" LCD available to the general public is 3840 x 2400. This is the IBM T220 and it was first sold in 2001, later replaced by the T221 which used the same LCD panel but had better electronics support. Viewsonic and iiyama rebadged and sold them too. It was eye-wateringly expensive (more than ten thousand bucks) and frankly not very good with a low refresh rate, crap viewing angles, low contrast and brightness compared to other lower-res LCD panels on sale even back then.
Their main market was financial companies and day-traders as the ability to put a lot of numbers and graphs on a single screen was considered more important than colour fidelity or gamer-twitch refresh speeds.
I flew from the UK to Montreal recently. My local airport has no direct flights to Canada but there is a daily flight to Newark and I could have easily bought a through ticket to Montreal, changing planes in Newark. But... that would mean jumping through all the Immigration hoops in the US, including registering on a US government website at least a week in advance to pre-clear my arrival. It was entirely possible that on my arrival at Newark I could be refused entry to the United States and repatriated back to the UK on a whim even if I had spent hours filling in all the details on the required forms.
Instead I found another flight to Montreal via Schiphol in the Netherlands. On arrival in Schiphol I stayed airside and passed through security only when embarking on the second leg of my flight. I did not have to pass through Immigration or Customs.
The last time I was in the US was 2004. I don't see myself going back again any time soon although I really enjoy my visits to America and would like to go again. It's the getting there that hurts.
The capital of Brazil is Brazilia, a purpose-built capital city somewhat like Washington DC. Rio is just a large Brazilian city, not even the biggest by most methods of measurement (population, incorporated land area etc.)
"USB2 won't even come close to cutting it for a GbE replacement, which is why you don't see any USB GbE dongles, only the 10/100 ones."
That's funny, I've got a GigE USB2.0 dongle sitting on my shelf... It won't do 100MiB/sec but it will let me transfer files over my home network at about 20MiB/sec, noticeably faster than a 100Mib/sec Ethernet connection which usually tops out at about 8MiB/sec data transfers. From what I recall it uses the AX88178 chip which you'll find in a lot of mains wiring networking kit such as Homeplug.
This is a prototype, not a production machine and there isn't anyone making very long thin touchscreens. Intel put three standard screens (I presume something like 800x480 resolution, the sort most high-end smartphones use) into a laptop chassis. On a production machine there would be a purpose-built touchscreen (2400 x 480 pixels perhaps) occupying the same space as the three displays do in the prototype.
I thought you said "ex-boyfriend" in the original posting? Nice piece of revenge for him, blame it on his idiot friend who took his comments the wrong way while his ex is writhing in agony in a hospital gurney, tubes down her throat. I do hope you would have owned up to the "joke" if the worst had come to the worst; a quick medical response to violent food allergies is a great help in preventing deaths in such cases -- an acquaintance of mine found out in a restaurant he was allergic to shellfish the hard way but luckily the people he was with called in paramedics who got a tracheal tube into him before his throat closed up completely. No long-term damage was done, thankfully.
I do hope you had the ER on speed dial when you carried out this experiment on her, you know, just in the unlikely chance that she was in fact allergic to chicken and would in fact react to the proteins involved. If she had claimed to have a nut allergy would you have smeared the steak in peanut butter, you know, just for a laugh?
Me, I'm allergic to eating duck, chicken, pretty much all types of fowl. It's not much of an allergy and if I eat enough of it (difficult to do now but I did when I was a kid) then a rash, throwing up and diahrrea was about the limit. I've steered clear of fowl since then (chicken eggs are OK, for some reason) as I'm not fond of the smell of cooked chicken, turkey etc. possibly because of aversive conditioning in the past. It's possible I'm not allergic any more but it's not something I'm interested in testing. Life is too short and pork smells great.
The 2408WFP Dell display is fitted with a S-IPS high-gamut panel capable of displaying a lot more colours than the cheaper 2408/2409W screens that use lower-cost DSTN technology panels. Serious graphics users want the extra colour quality of S-IPS or PVA panels and are willing to pay for it, the cheap 6-bit multiplexed DSTN screens are fine for working with spreadsheets and word processing in an office environment. You pays your money and makes your choice.
There are very few 1MT-plus nukes left in anyone's arsenal. Accurate aiming and the need to shrink the physical size of the weapons to provide MIRV capability in sub-based missiles and smaller silo-based launchers meant a move away from the older higher-yield devices. Typical maximum yield from the latest generation of weapons is about 500-600kT. The WE177 freefall bomb deployed by the RAF had a dial on the side that allowed the yield to be adjusted from about 200 to 500kTonnes.
Radius of effectiveness at the target depends on how high up the bomb is triggered. Attacking hard targets such as docks, freeways and railheads (all built low to the ground of concrete and stone with few structral voids) requires a lower detonation point compared to destroying clusters of freestanding buildings. Going after buried structures such as command bunkers requires ground-burst or even earth-penetrator weapon designs.
It is believed the Chinese have some megatonne-range weapons to compensate for their delivery systems probable error on target but the Big 4 (US, Russia, UK, France) all have accurate warhead deployment systems that have negated the need for high-yield weapons.
"Fibre To The Cabinet" is reasonably inexpensive to implement as it involves optical fibre bundle pulls from the local exchange through exiisting ducts to roadside cabinets where the DSLAMs are installed. One bundle pull plus cabinet refits will supply 200-300 homes. Rewiring all 25 million homes in the UK with individual fibre-class connections would cost a lot more, an order of magnitude more at least.
I expect the FTTC system to be improved as it is rolled out, in the same way ADSL has gone from 512kbits/sec with a limit of 3km distance to 16 to 24Mbits/sec over arbitrary lengths of wire. The experimental installs of FTTC in the UK are going to start at 40Mbits/sec.
Experimenters using twisted pairs and DSLAM-style equipment have achieved 500Mbits/sec on short hauls, the sorts of distances that a cabinet-to-home wire link would cover so 100-200Mbits/sec are probably achievable for FTTC. The very few users who need more than that sort of data rates have commercial services available to them but obviously not at UKP25 quid a month.
The MC68000 was not available in production quantities at the time the IBM PC design was being finalised. The chip was late and buggy -- I used a dev board with a pre-production version of the chip clocked at half-speed, 4MHz, in 1982. Attempts to run it at 8MHz (the datasheet spec speed) were a failure.
There were other reasons for IBM to go with the 8086-family chipsets:
1) the 8086/8088's bus could easily drive the 8080-family support chips such as the 8251, 8255, 8259 etc. to build a complete system. The MC68k family support chips were even later than the release of the CPU itself (in some cases like the MMU several years late) and the MC68k bus could not be easily interfaced with the Intel family chips which were cheap and in plentiful supply.
2) the 8086 family's internal data registers and addressing modes were designed to simplify conversion of existing 8080 code to run on the new 16-bit CPUs. The 68k, although a superior CPU in all respects to the 8086 family, had no tools available to make code conversion from the 6800 or other sibling CPU family (6809, 6502 etc.) simple -- all 68k code had to be written from scratch.
3) the 68k was an expensive chip, not suprising as it was complex and required a large die, necessitating a 0.6" wide 68-pin DIL ceramic package. Motorola's target market for the chip was $10,000 workstations, not "toy" desktop computers only costing $2,000. By comparison the 8088 was cheap as chips.
There were a lot of suits used in testing and training, pretty much identical to the ones that actually went to the Moon. Some of these extra suits were loaned out for public exhibits and displays -- I recall hearing about a suit on display at a Star Trek convention which had one of the quick-detachable gloves stolen from it.
The Command Module is about the size of a compact car internally, and it was the only part of the Apollo mission hardware that returned to Earth. The crew wore pressure suits on launch and descent but they weren't the full Moon suits with external backpacks, cooling systems, radios etc.
Slashdot Mirror
The problem is that "any flammable liquid" would cost more than gasoline (absent tax). They demonstrated the Chrysler turbine car running on perfume, for example. Back in the 50s gas was well under a dollar a gallon -- how much would a gallon of perfume cost in comparison? Kerosone (aka JP-1)? About the same price but with less calorific energy per gallon so it would provide less mileage than gasoline. Coal gas may be cheaper per mile but storage and distribution and tankage in the vehicle would be problematic and it's a dirty fuel to produce. Natural gas is getting pricier by the day so that doesn't help much.
The Japanese superconducting maglev system is being designed to run in service at 550km/hr. The test vehicles (which often carry passengers) can achieve this speed pretty much at will, it does not require special versions of the trains. The test area in Japan has two side-by-side tracks about 40km long (longer than the Chinese Beijing airport maglev) and they have successfully run two trains past each other at a closing speed of 1100km/hr, something that may cause problems with super-high-speed steel-wheel trains given their existing rights-of-way have no blast dividers between the tracks.
There's a lot of folks like yourself that figure third-world manufacturers work with stoop labour in candle-lit shacks to outproduce the super-efficient West. It's just not so.
Any new production technology, any new machines or processes that permit manufacturers to make stuff cheaper or better can be bought and used in China or elsewhere in the world for the same ticket price as it would cost in Western countries. For an example, steel-making plants in places like Indonesia and Malaysia were built in the 1970s with the most modern process control systems available at the time allowing them to replace imports of Western steel with a locally-made cheaper product. The end result was the death of steel-making in the West, in part because there was little iron ore left to be sourced locally there. Since it costs the same to make the steel in either place the price was set by transport-to-market and it's cheaper to ship finished steel than it is to ship iron ore and process it in the West. A side-effect of this was a decrease in shipbuilding in the West as Far Eastern shipyards blossomed using modern shipbuilding techniques and design methods building lower-cost ships with cheap locally produced steel.
The same thing is true pretty much in other industries such as electronics or car manufacturing. The production lines in places like China or Taiwan use the same equipment, the same components, the same quality control processes as any business in the West does. These machines and processes weren't stolen, they weren't secret, they were commodities just like a box of computer memory, sold to a customer with cash to pay for them.
The US has been using up its existing stockpiles of Pu-238 to build RTGs for a mixture of civilian deep space projects and black intel operations such as non-solar-powered stealth spy satellites and seabed-emplaced submarine monitoring stations. The Russians agreed to sell the US some Pu-238 under a licence that prevented it being used for military functions but they shut that down when it became obvious the US was reallocating most if not all of its home-grown stockpile to the military side of things. Like oil Pu-238 is fungible and the Russian supply of Pu-238 was effectively enhancing US military capabilities.
The T221 display had crap colour definition, low contrast especially off-axis, slow refresh and it required four video channels to drive it. Video and photo reproduction wasn't its forte though, what it was great at doing was putting a lot of numbers and graphs in front of someone's eyeballs. It mainly sold into scientific and engineering environments (CAD, PCB and chip layout etc.) and financial trading houses but it was never aimed at Joe Public.
You see them turn up on Ebay occasionally, still commanding decent resale prices. If you do buy one make sure you get all the bits like the special cable harnesses etc.
The US has several off-the-shelf medium/heavy lifters such as the Delta 4 Heavy that can put up to 20 tonnes into orbit similar to the Ariane V. What they don't have (and nobody else has) is a superheavy lifter capable of carrying a 70-tonnes plus payload which is needed to perform the one-shot-to-the-Moon mission envisaged for Constellation (with a separate crew flight). However there are problems man-rating an existing lifter; the flight profile needs to be configured so that the maximum acceleration at any point in the flight is tolerable to the Spam-in-a-can plus a lot of other factors such as safety and abort flight modes and hardware mods.
ESA is preparing to buy and fly Soyuz spacecraft from their Guiana spaceport, initially to carry unmanned payloads such as the Progress ISS supply capsule. The Soyuz design is already man-rated and well-proven (over 1700 flights) and it wouldn't take much upgrading to add a manned spaceflight capability to the ESA catalogue based on the Soyuz.
http://www.esa.int/SPECIALS/Launchers_Home/SEMFFUZO0WF_0.html
The Falcon 9 did not nail its first flight. The upper stage ended up in an elliptical 240x280km orbit rather than the planned circular 250km one. The upper stage was also yawing and rolling noticeably after orbital insertion -- I watched the web broadcast of the launch and I wondered if that was deliberate or not, it turns out it was not. Hopefully they'll fix those problems in the next few test launches.
I can't see NASA or any of the other participants in the ISS programme letting a Spacex vehicle make an approach to their fragile space station unless they are sure it's not going to crash into it.
Cheap ROM has a read-cycle time of about 40-50nS and works OK with CPUs running at under 100MHz. There's a lot of embedded kit out there for which that is a good solution, although today most low-end embedded kit uses EEPROM or flash rather than mask ROM for production-line flexibility. Modern DRAM in the DDR2/DDR3 class has read cycle times in bursts of less than 1 nS. For today's console or PC based games that sort of speed is required to keep the gameplay moving. ROM cartridges would work OK for simple sprite-based TV resolution graphics as they have done in the past. For more modern visually-realistic games they are a non-starter.
I'm just back from three weeks in Japan, travelling around and staying in salaryman hotel rooms (well, apart from the suite upgrade at a resort hotel in Atami when they didn't have any single rooms left). Nearly all these small business hotels have free cabled internet which is pretty fast and reliable, a few have rather ropy low-signal wifi with frequent dropouts and reconnections required. In one case I found myself on an 11MBps connection, probably ancient 802.11b hardware.
I much prefer cabled connections to wifi when staying in hotels. Many hotels supply wifi by sticking a router on every other floor in the building as a minimal option. Doing it right with a hotspot in the ceiling over every room is going to be a lot more expensive for them to implement and most of them will not bother.
There are other advantages to air launch -- no large pad structures required, blast protection systems etc. as the motor only ignites in free air. The payload rocket can be built and handled horizontally throughout its entire lifespan until it is actually fired in flight etc. Adding the extra fuel tankerage to give the rocket that extra 10-12% is not a trivial matter; it needs extra strengthening to carry the extra mass on the pad and under acceleration which adds to the all-up launch weight requiring more fuel and oxidiser to carry it. That extra structure all costs money that is thrown away on every launch, never to be seen again.
As for building a gargantuan aircraft, it's not difficult to do with today's engineering knowledge and materials. It wouldn't even be that expensive to do since it's not meant to be a cost-competitive commercial airliner design built in the hundreds to carry fare-conscious passengers and air freight. A brute-force design would do the job -- big strong wings plus lots of engines for thrust. It doesn't need endurance to fly for ten or twelve hours at a time, it is not fine-tuned to carry passengers and freight at minimum operating costs, its fuel consumption per kilometre flown doesn't need to be cut to the bone because of marketing constraints. There would be at most three aircraft in the fleet, two minimum to maintain a schedule of launches but the project could start with just one so no need for a production line and the expensive engineering optimisations that requires.
There are efficiency advantages to launching rockets into space from aircraft.
A: Rocket motors work best when there is little or no air pressure for them to burn into, increasing the jet speed and the Isp figure. Solid boosters benefit more than liquid fuelled rockets as the exhaust pressure is less. At 35-40,000 feet the air pressure is about 30% of that on the ground, making rockets a few percent more efficient at that altitude compared to sea-level.
B: Existing land-based launchers have to spend a significant chunk of their fuel and oxidiser getting to the velocity and altitude an aircraft launcher has already reached using conventional Jet-A fuel and atmospheric oxygen. This is why there are no single-stage-to-orbit launchers, they all have to lift off heavy and throw away sections of their airframe on the way up so the remaining parts can make it to orbit.
C: The addition of 1000km/hr horizontal velocity from the launcher is, as you say very much less than the orbital velocity of 25,000 km/hr required for LEO, but it's 4% which is worth having.
A rough BOTE calculation suggests an air launch to LEO can be done for about 10-12% less fuel load than a similar launch from the ground. That parlays into less parasitic airframe, tankerage etc. that has to be flightworthy and capable of withstanding extended multiple-G accelerations and which requires fuel and oxidiser to get it off the pad.
As for size limitations on air launches that's due to a lack of really big off-the-shelf aircraft capable of being modded for use as launchers. If there was a determination to build an air-launch system for getting, say, 20 tonnes of payload into LEO per launch then designing and building a couple of dedicated carrier aircraft to do the job would be easy. They don't have to be efficient fuel-misers the way modern airliners or even military aircraft are, they can use multiple off-the-shelf aero engines to make up for the lack of efficient design and excess weight as their fuel costs are going to be a minor part of the operating TCO. My own thoughts on what they would look like tend towards a land-launched catamaran high-wing seaplane with the orbital vehicle suspended between the hulls (a bit like the Virgin Eve carrier aircraft).
"1. LED backlit screen. This seems strange given the existence of PixelQi and mirasol type LCD displays."
The PixelQi panel is not available today in quantity ten million unlike the commodity low-resolution display designed into the iPad which is shipping now and has been for the past few years. The PixelQi panel is in high demand which means it costs a lot and that would cut into Apple's profit margins. Apple can't sell iPads at a thousnad-buck pricepoint hence no OLED displays.
"2. No camera? What happened? This is very disappointing."
Power and cost. The iPad has a 25Whr battery and it needs longer runtimes than a laptop to differentiate it from the existing market in laptops and especially netbooks. Adding a camera eats into the power budget. It also adds five or ten bucks to the BOM and that cuts into the profit margins.
The Japanese superconducting maglev trains which are currently under development hold the manned record for such trains of 580km/hr and they regularly achieve over 500km/hr, the target for production trains during tests. More impressively the test route (which is longer than the Shangai airport maglev link) is double-tracked and they have run a pair of trains past each other at a closing speed of over 1000km/hr.
Actually the max resolution on a 22" LCD available to the general public is 3840 x 2400. This is the IBM T220 and it was first sold in 2001, later replaced by the T221 which used the same LCD panel but had better electronics support. Viewsonic and iiyama rebadged and sold them too. It was eye-wateringly expensive (more than ten thousand bucks) and frankly not very good with a low refresh rate, crap viewing angles, low contrast and brightness compared to other lower-res LCD panels on sale even back then.
http://en.wikipedia.org/wiki/IBM_T220/T221_LCD_monitors
Their main market was financial companies and day-traders as the ability to put a lot of numbers and graphs on a single screen was considered more important than colour fidelity or gamer-twitch refresh speeds.
I flew from the UK to Montreal recently. My local airport has no direct flights to Canada but there is a daily flight to Newark and I could have easily bought a through ticket to Montreal, changing planes in Newark. But... that would mean jumping through all the Immigration hoops in the US, including registering on a US government website at least a week in advance to pre-clear my arrival. It was entirely possible that on my arrival at Newark I could be refused entry to the United States and repatriated back to the UK on a whim even if I had spent hours filling in all the details on the required forms.
Instead I found another flight to Montreal via Schiphol in the Netherlands. On arrival in Schiphol I stayed airside and passed through security only when embarking on the second leg of my flight. I did not have to pass through Immigration or Customs.
The last time I was in the US was 2004. I don't see myself going back again any time soon although I really enjoy my visits to America and would like to go again. It's the getting there that hurts.
The capital of Brazil is Brazilia, a purpose-built capital city somewhat like Washington DC. Rio is just a large Brazilian city, not even the biggest by most methods of measurement (population, incorporated land area etc.)
"USB2 won't even come close to cutting it for a GbE replacement, which is why you don't see any USB GbE dongles, only the 10/100 ones."
That's funny, I've got a GigE USB2.0 dongle sitting on my shelf... It won't do 100MiB/sec but it will let me transfer files over my home network at about 20MiB/sec, noticeably faster than a 100Mib/sec Ethernet connection which usually tops out at about 8MiB/sec data transfers. From what I recall it uses the AX88178 chip which you'll find in a lot of mains wiring networking kit such as Homeplug.
http://www.amazon.com/TRENDnet-USB-Gigabit-Ethernet-Adapter/dp/tech-data/B000NIX7A4/ref=de_a_smtd
is an off-the-shelf GigE USB2.0 dongle. 35 bucks.
This is a prototype, not a production machine and there isn't anyone making very long thin touchscreens. Intel put three standard screens (I presume something like 800x480 resolution, the sort most high-end smartphones use) into a laptop chassis. On a production machine there would be a purpose-built touchscreen (2400 x 480 pixels perhaps) occupying the same space as the three displays do in the prototype.
I thought you said "ex-boyfriend" in the original posting? Nice piece of revenge for him, blame it on his idiot friend who took his comments the wrong way while his ex is writhing in agony in a hospital gurney, tubes down her throat. I do hope you would have owned up to the "joke" if the worst had come to the worst; a quick medical response to violent food allergies is a great help in preventing deaths in such cases -- an acquaintance of mine found out in a restaurant he was allergic to shellfish the hard way but luckily the people he was with called in paramedics who got a tracheal tube into him before his throat closed up completely. No long-term damage was done, thankfully.
I do hope you had the ER on speed dial when you carried out this experiment on her, you know, just in the unlikely chance that she was in fact allergic to chicken and would in fact react to the proteins involved. If she had claimed to have a nut allergy would you have smeared the steak in peanut butter, you know, just for a laugh?
Me, I'm allergic to eating duck, chicken, pretty much all types of fowl. It's not much of an allergy and if I eat enough of it (difficult to do now but I did when I was a kid) then a rash, throwing up and diahrrea was about the limit. I've steered clear of fowl since then (chicken eggs are OK, for some reason) as I'm not fond of the smell of cooked chicken, turkey etc. possibly because of aversive conditioning in the past. It's possible I'm not allergic any more but it's not something I'm interested in testing. Life is too short and pork smells great.
The 2408WFP Dell display is fitted with a S-IPS high-gamut panel capable of displaying a lot more colours than the cheaper 2408/2409W screens that use lower-cost DSTN technology panels. Serious graphics users want the extra colour quality of S-IPS or PVA panels and are willing to pay for it, the cheap 6-bit multiplexed DSTN screens are fine for working with spreadsheets and word processing in an office environment. You pays your money and makes your choice.
There are very few 1MT-plus nukes left in anyone's arsenal. Accurate aiming and the need to shrink the physical size of the weapons to provide MIRV capability in sub-based missiles and smaller silo-based launchers meant a move away from the older higher-yield devices. Typical maximum yield from the latest generation of weapons is about 500-600kT. The WE177 freefall bomb deployed by the RAF had a dial on the side that allowed the yield to be adjusted from about 200 to 500kTonnes.
Radius of effectiveness at the target depends on how high up the bomb is triggered. Attacking hard targets such as docks, freeways and railheads (all built low to the ground of concrete and stone with few structral voids) requires a lower detonation point compared to destroying clusters of freestanding buildings. Going after buried structures such as command bunkers requires ground-burst or even earth-penetrator weapon designs.
It is believed the Chinese have some megatonne-range weapons to compensate for their delivery systems probable error on target but the Big 4 (US, Russia, UK, France) all have accurate warhead deployment systems that have negated the need for high-yield weapons.
"Fibre To The Cabinet" is reasonably inexpensive to implement as it involves optical fibre bundle pulls from the local exchange through exiisting ducts to roadside cabinets where the DSLAMs are installed. One bundle pull plus cabinet refits will supply 200-300 homes. Rewiring all 25 million homes in the UK with individual fibre-class connections would cost a lot more, an order of magnitude more at least.
I expect the FTTC system to be improved as it is rolled out, in the same way ADSL has gone from 512kbits/sec with a limit of 3km distance to 16 to 24Mbits/sec over arbitrary lengths of wire. The experimental installs of FTTC in the UK are going to start at 40Mbits/sec. Experimenters using twisted pairs and DSLAM-style equipment have achieved 500Mbits/sec on short hauls, the sorts of distances that a cabinet-to-home wire link would cover so 100-200Mbits/sec are probably achievable for FTTC. The very few users who need more than that sort of data rates have commercial services available to them but obviously not at UKP25 quid a month.
There were other reasons for IBM to go with the 8086-family chipsets:
1) the 8086/8088's bus could easily drive the 8080-family support chips such as the 8251, 8255, 8259 etc. to build a complete system. The MC68k family support chips were even later than the release of the CPU itself (in some cases like the MMU several years late) and the MC68k bus could not be easily interfaced with the Intel family chips which were cheap and in plentiful supply.
2) the 8086 family's internal data registers and addressing modes were designed to simplify conversion of existing 8080 code to run on the new 16-bit CPUs. The 68k, although a superior CPU in all respects to the 8086 family, had no tools available to make code conversion from the 6800 or other sibling CPU family (6809, 6502 etc.) simple -- all 68k code had to be written from scratch.
3) the 68k was an expensive chip, not suprising as it was complex and required a large die, necessitating a 0.6" wide 68-pin DIL ceramic package. Motorola's target market for the chip was $10,000 workstations, not "toy" desktop computers only costing $2,000. By comparison the 8088 was cheap as chips.
There were a lot of suits used in testing and training, pretty much identical to the ones that actually went to the Moon. Some of these extra suits were loaned out for public exhibits and displays -- I recall hearing about a suit on display at a Star Trek convention which had one of the quick-detachable gloves stolen from it.
The Command Module is about the size of a compact car internally, and it was the only part of the Apollo mission hardware that returned to Earth. The crew wore pressure suits on launch and descent but they weren't the full Moon suits with external backpacks, cooling systems, radios etc.