AFAIR the actual Moon suits never returned to Earth. They were left in the ascent stage of the Lunar Module after the crew transferred back into the tiny Command Module in Moon orbit. There just wasn't space (so to speak) in the Command Module for them and it wasn't like they were going to be used again.
"The queen has no role in the security status of her government."
She has a face-to-face meeting with the Prime Minister, the senior elected official of Her Majesty's Government once a week to talk about Britain and he Commonwealth, its current state and its place in the world. She receives a box of State papers every day to work through -- as Head of State she signs off on all treaties and international agreements negotiated between the UK and other countries.
She's been doing this for more than fifty years now, day in day out, week in week out. She's heard everything, the good news and the bad. Apart from the anodyne Christmas address to the nation she keeps her mouth shut about it, as it is her duty to do so.
The final choice of sites for the ITER project came down to Japan and France -- the US was about the only major ITER partner which supported Japan. Most of the other partners are European and supported the French option, Cadarache. It was rumoured that the US opposed the choice of Cadarache in part because the French government didn't support the Iraq invasion in 2003. How true this is we don't know but in that period there was a lot of anti-French hysteria in the higher ranks of the US government which may have contributed.
Hospital use -- how do you disinfect it? Will it survive being wiped down with astringent cleaning solutions several times a day?
Warehousing -- drop it several times on a concrete floor from a metre up, landing on all four corners. Does it survive that experience and continue to function OK?
There are portable data-logging devices that will survive that sort of treatment (and worse) and they're available today, but they don't come cheap. Any Apple netbook-type device is not going to find much of a market in those sorts of areas.
The UK has about 70 tonnes of Pu-239 in storage. Using small-ball physics packages, that's enough Pu to make about 7,000 nuclear weapons. Right now we've got about 200 nukes capable of being used in a shooting match (about the same number as Israel is suspected of owning) and we really don't want any more. We don't need more plutonium.
We also don't need more tritium -- there's a couple of warehouses filled with tritium-filled Trimphone dials which are being left to decay through half-life because it's more trouble than it's worth to recover or dispose of the tritium. Some enterprising fellows even made novelty keyrings from tritium capsules -- I have a couple of them hanging on a desklamp next to me even as I type this. We might need lots of tritium in the future to fuel fusion reactors but that's not a certainty, and lithium-breeding in the fusion reactors will probably produce enough for our needs.
The reason to build new PWRs in the UK is that they demonstrably work and they have a long service history behind them; there should be no surprises and gotchas. We also have fuel production and reprocessing lines set up to manufacture and recycle PWR fuel elements. Using less-well-known reactor designs could mean problems down the road, and we've been through that with the Advanced Gas-cooled Reactors (AGRs) we built in the 1980s. They're great reactors, even more efficient than PWRs but they cost a lot more to build and operate, and their unique fuel rods require special and expensive handling.
You (I presume you are American) already have one. It's called... ITER. The US whined and moaned over the possibility of ITER being built in France, delaying its start by a couple of years. Eventually the US was over-ruled by nearly all the other countries who actually wanted to get on with developing fusion as a possible power source. Cadarache in France was finally chosen as the site and the project is now up and running.
There are a number of other fusion research projects going on around the world, but ITER's job is to figure out how to build a magnetic-containment fusion reactor that will deliver electrical power into a grid, and it's the biggest and best-equipped (and best-funded) of the current fusion projects.
EDIT: I may have spoken too soon, as I'm not sure the US has actually agreed to contribute to funding ITER, at least in this budget cycle.
Philips made radio tubes for the Wehrmacht, Kriegmarine and Luftwaffe during WWII while Holland was occupied by the Nazis. Remarkably the tubes suffered a high failure rate, but only after several hours of flawless operation, enough to get them past inspection and initial fitment but not much longer. Odd that.
"So the government gets to choose who will watch over the government?"
All the parties represented in the Commons put forward candidates for a Peerage in proportion to the number of seats they hold. They are often party politicians but not always -- for example Mary Warnock was made a Baroness and Life Peer in 1984 after her work chairing the Committee of Enquiry into Human Fertilisation and Embryology. As a member of the House of Lords she has been instrumental in shaping the legislation that has helped make the UK a world leader in cloning and embryology research.
"A cop can just take you from the street and lock you up for 42/28 days without anyone asking why(like a judge...)"
Ummm, no. Longest detention without judicial oversight is 24 hours in the UK. After that a suspect's case is brought before a judge who can order an extension of detention without charge once the police have laid out the reasons they want to hold the suspect in this manner. After that the detention order is reviewed regularly by a senior judge who will usually give the police a few days at a time before they have to come back and explain themselves. If it stretches out the judge will want to hear really really good reasons why the detention order should be extended at all.
In the very rare case of someone being held for the existing 28-day period without charge the case for detention will be reviewed by the judiciary at least five or six times. It's almost unknown for someone to be held for an extended period and then released without charge. The usual reason for the police to do this is because there are more serious charges being investigated -- e.g. the suspect has been found in possession of explosives but they're trying to track down who supplied them, if they are implicated in other crimes etc. They *will* be charged with something in that circumstance.
They're available in various x86 processor and RAM capacities, including Transmeta CPUs; the older models (Celeron 266, PII 300) are quite cheap today. Fat battery packs are available (Sony branded ones cost serious bucks but 3rd party units are a lot cheaper) that will run to 12 hours or more uptime. Replacing the HDD with a SSD will save you more battery power. Linux is readily ported onto most of the C1 variants and they all have PCMCIA or CardBus slots to support WiFi.
The accessory I regret not getting for my old PB was a ballistic-nylon shoulder holster for carrying it around.
They need to be very reliable -- they're going to be running 24/7/365, handling several kilowatts of power, operating as efficiently as they can so they don't waste energy unnecessarily. If I had to design something like that I'd overrate everything by a factor of 5 or more to ensure reliability and uptime and doing that will add to the pricetag.
There's also other factors governing the cost. These inverters aren't mass-manufactured as such, and the makers need to provide a long-term maintenance opreration with technicians and spares on tap, not just for warranty servicing. There's also liability and type testing -- having a few of these inverters bursting into flames due to a design defect would be bad news, financially speaking.
The reactor at Torness is an Advanced Gas-cooled Reactor (AGR), a British design built in the 1980s and optimised for efficiency. As it turned out, fuel costs for nukes haven't risen particularly over the last couple of decades so fuel efficiency isn't as important as the designers thought it would be. That's why the UK built cheaper GE-designed Pressurised Water Reactors (PWR) after the last of the AGRs was commissioned. They're not as efficient in turning fuel into power as the AGRs but they're cheaper to build and operate and they were available off a production line instead of being hand-built one-offs.
Here's the tale of a visit to the Torness nuclear power station by an old denizen of Slashdot.
I believe the higher capacity drives will force a rethink on how data is stored and accessed on standalone machines like laptops and desktops. I've only got a couple of terabytes of data on this machine and doing a file search over the five (I think, I can't actually remember how many drives I've got fitted in this thing) disks is already pretty time-consuming. The solution will be to add intelligence to the disk interface so that data indexing is done pre-emptively and the results cached on the fly.
The first generation of hybrid drives are already here but they're only at the beginning of their development cycle. HDD recording densities will increase as will flash RAM densities and that will improve access times but only for the most commonly accessed data.
Imagine a 10Tb HDD built in the classic 3.5" wide form factor, with 256Gb of 1024-bit-wide 150MWord/sec flash memory or MRAM on the controller board acting as cache. The spinning disk becomes a backing store for the flash where data is kept "fresh" by a smart algorithm. The drive spins down intelligently when not needed, saving power and reducing heat dissipation.
Higher recording densities are only one part of the future of disk drive technology.
OFC wire has a use in audio, but not for speakers. Oxygen-free copper has almost no copper oxide crystals in its makeup hence there's no piezoelectric noise generated when the wires flex. It makes great cable for microphones and guitars and other instruments that move during a performance and which carry low-amplitude signals. The difference in resistance between regular copper and OFC is almost immeasurable.
Speaker wires shouldn't move when they've been laid into place. Even if they did they carry high voltages and significant current which dwarfs any induced piezo noise from oxide inclusions.
As for the skin effect also mentioned by somebody above, it only comes into play in the multimegahertz frequency range. Radio amateurs buy silver-plated coax for this reason but it only makes a noticable difference in attenuation at VHF frequencies and above. At audio frequencies there's no skin effect worth mentioning. Multicore wire doesn't enhance the skin effect as it's an electromagnetic effect and the skin effect would tend to concenetrate the current flow in the outside of the entire wire bundle, not on each individual core.
If you want really good speaker wires get the heaviest solid twin-core mains wiring you can get your hands on and use the shortest single runs you can lay down between the amplifier and the speakers. Screening the cable from mains pickup may do some good but is usually pretty irrelevant given the currents running in the wires. Solder the speaker wire directly onto the speaker connections or the crossover unit in the speaker cabinet, don't use interconnects.
If you're paying more than a dollar a metre for speaker cable, you're paying waaay too much. Home Despot is your friend.
"Halting State" by Charlie Stross is out in October. It starts with a bank-raid in an MMORPG by a bunch of orcs with a dragon for fire-support and then gets weird.
Right now, Google is building infrastructure -- distributed data centres, lighting up dark fibre trunks etc. That way they've got their own backbone capacity that can't be leveraged away from them with price hikes or lock-in deals between the existing backbone operators and the big telecomms/data providers.
Step two would be the biggie -- Google subcriber fibre. They enter the "last-mile" business and provide their own seriously high-capacity home and small business connections. The Big Guys can't touch them at this point since they have their own backbone operation already in place, not requiring essential ATM hookups that are a de facto momopoly. It's a Google-owned operation end-to-end.
The future is higher capacity, higher bandwidth for the individual user. BitTorrent, Joost and other streaming technologies are in their infancy but lots of distributed data centres are ideal for supporting that kind of messy ad-hoc multiple-connection data model. Google's planning seems to fit that vision of the future, but only if it gets down and dirty in the supply business rather than just standing off and expecting others to carry their content (if they feel like it) to the end-user who pays the bills.
Step 3? Profit! Expect the dinosaurs to scream like raped apes when something like this is sprung on them -- they believe they have a de-facto monopoly on last-mile (hence the Net Neutrality fan-dance), but it's something that could easily be overturned by the right iconoclastic company.
There were a number of factors that made the IBM team go for the Intel solution rather than the superior 68000 chip from Motorola. One factor was, as you said, the 8088 chip could use plentiful and well-understood 8-bit peripheral chips like the 8251, the 8259 etc. Another factor was that the 68000 had a long gestation -- Motorola had problems producing chips in quantity that could actually run at their specced speed of 8MHz. I played with an early dev board which had a working 68000 but it only ran at 4MHz. At that time the 8086/8088 were already available on the market, in quantity and that's what IBM needed.
The biggest factor though was probably the software. The x86 architecture was structurally similar to the venerable 8080 family's architecture (hence the memory segmentation, 8/16 bit registers etc.), and there was a lot of 8080-family code already out there, including developer toolsets like compilers and assemblers. Intel also released conversion tools for coders to convert existing 8080 code to run on their new 16-bit CPUs. The M68k was radically different structurally from Motorola's 6800 8-bit chips and its instruction set (although a dream to code new stuff for) did not make transitioning older 8-bit code easy.
Olden days, I was doing microprocessor coding for robotics. We used an assembler/linker package running on a couple of 16-bit minicomputers. One machine had enough RAM to do the assembly, the other had a hard disk that let us do the link of the object code to produce an image we could blow into an EPROM. Neither machine could do both parts.
Problem was the two machines were in adjacent rooms and networking was nonexistent. We'd run the assembly process, produce a paper tape of the object code, carry it to the other machine and then feed it into its paper-tape reader to do the link. This took a lot of time which delayed development.
I noticed the assembly process ran quite quickly but the linker machine was noticeably slower, and the machines weren't actually that far apart in reality. I got a couple of empty paper-tape spools and clamped them to the doorframes to act as guides, then I took an assembly tape being punched by the first machine out its room, around the doorframe, into the next room and fed it into the other machine's reader. A garbage bin acted as buffer storage for overflow tape, and it worked. I had to baby-sit it in case of jams or tangles but it cut the total assembly/link processing time by at least an hour.
COBOL... lessee, you're a bank shuffling billions of dollars around every day, reconciling accounts after business has closed. You've got a pile of 30-year-old continuously maintained COBOL doing the shuffling and adding up those BCD dollar and cents.
And then a whizzkid from down the corridor in the website support office suggests you abandon all that working, time-tested code and rewrite it all in some open-source incredibly cryptic code that looks like line-noise, that doesn't support native BCD, that...
It's difficult to screw up COBOL code by putting a period in the wrong place, or missing out an ampersand. It won't compile, it won't run, it won't drop a billion dollars down a rathole. There's a reason it was designed like that and that is the reason it is still widely used.
Why do some programming languages (C, Perl) have to be cryptic? It doesn't save much time in the compile, it produces executables that are the same size as non-cryptic languages like Pascal or COBOL. About the only reason I can see is it saves the programmer some effort at the keyboard. Is that all?
The SR-71 required multiple tanker operations for any sort of operational flight. The profile was usually to tanker up in international airspace, penetrate "enemy" airspace at high speed and high altitude burning fuel like crazy, snap pictures over the target (assuming it was not obscured by clouds or smoke) and then dash out and then refuel again. It took days or even weeks to prepare any given SR-71 recon mission as refueling mistakes could cost them the plane (as it did in one instance when the wrong sort of fuel was loaded onto a tanker).
The SR-71's fuel wasn't stock Jet-A, it was a specially concocted mix that I've heard cost about as much as single malt whisky, litre for litre.
"The evidence that Microsoft does this and uses that information... Is?"
Clippy. For the 98% of people out there who aren't computer-literate uberhackers, Clippy is help and succour personified when the damn computer doesn't do what they want. Once they're up to speed they can dispense with Clippy's assistance, or switch it back on when they try something new. It's like training wheels -- once you're a Kool Kid you look down on the dorks still cycling around in circles on four wheels and carefully forget you were ever the same as them.
MS made a careful and calculated decision to put Clippy and co. into their premier software package and they made it after talking to a lot of potential customers, studying them carefully and figuring out what they needed and how to make it available in a way they would want to use. Blind guesses and stabs in the dark are how GPL UIs are created (I was going to say "designed" but there's little evidence of rational thought behind most OSS UI design decisions) which is why nearly all of it is a pain to use, even if the functional code underneath is competent.
If you want ten million people to use a piece of software you need to test its UI on thousands of subjects, young and old, techies and computer illiterates. You need to run it on laptops, on CRTs, on LCDs, in rooms with bad lighting and in direct sunlight. You need to find people with palsy who can't control a mouse, people with colour blindness and generally bad eyesight, people with dyslexia, people whose first language isn't English, and you can't use them more than once for any given piece of software testing.
To analyse those results you need teams of psych grad and industrial T&M researchers to go over the videotapes of each subject with stopwatches measuring each eye movement, each hesitant gesture, each mistake, each fumble to produce big databases of results and dig deep for the golden mean that gives the average user the best result in the UI. You then feed that info to your software designers who go away and produce the next iteration of code that has to go through the same horribly expensive (in time AND money) testing mangle again. And again. And again.
Aha! What about beta testing, you cry? Sorry, all that UI development is done a long time before anyone outside the company ever sees a beta release version. By then they're hunting bugs and fine-checking that the techies who will run and install this stuff for users can do their part in making ten million copies work; the UI is pretty much set in stone by then and a hundred million dollars has already been spent.
Slashdot Mirror
AFAIR the actual Moon suits never returned to Earth. They were left in the ascent stage of the Lunar Module after the crew transferred back into the tiny Command Module in Moon orbit. There just wasn't space (so to speak) in the Command Module for them and it wasn't like they were going to be used again.
She has a face-to-face meeting with the Prime Minister, the senior elected official of Her Majesty's Government once a week to talk about Britain and he Commonwealth, its current state and its place in the world. She receives a box of State papers every day to work through -- as Head of State she signs off on all treaties and international agreements negotiated between the UK and other countries.
She's been doing this for more than fifty years now, day in day out, week in week out. She's heard everything, the good news and the bad. Apart from the anodyne Christmas address to the nation she keeps her mouth shut about it, as it is her duty to do so.
The final choice of sites for the ITER project came down to Japan and France -- the US was about the only major ITER partner which supported Japan. Most of the other partners are European and supported the French option, Cadarache. It was rumoured that the US opposed the choice of Cadarache in part because the French government didn't support the Iraq invasion in 2003. How true this is we don't know but in that period there was a lot of anti-French hysteria in the higher ranks of the US government which may have contributed.
Hospital use -- how do you disinfect it? Will it survive being wiped down with astringent cleaning solutions several times a day?
Warehousing -- drop it several times on a concrete floor from a metre up, landing on all four corners. Does it survive that experience and continue to function OK?
There are portable data-logging devices that will survive that sort of treatment (and worse) and they're available today, but they don't come cheap. Any Apple netbook-type device is not going to find much of a market in those sorts of areas.
The UK has about 70 tonnes of Pu-239 in storage. Using small-ball physics packages, that's enough Pu to make about 7,000 nuclear weapons. Right now we've got about 200 nukes capable of being used in a shooting match (about the same number as Israel is suspected of owning) and we really don't want any more. We don't need more plutonium.
We also don't need more tritium -- there's a couple of warehouses filled with tritium-filled Trimphone dials which are being left to decay through half-life because it's more trouble than it's worth to recover or dispose of the tritium. Some enterprising fellows even made novelty keyrings from tritium capsules -- I have a couple of them hanging on a desklamp next to me even as I type this. We might need lots of tritium in the future to fuel fusion reactors but that's not a certainty, and lithium-breeding in the fusion reactors will probably produce enough for our needs.
The reason to build new PWRs in the UK is that they demonstrably work and they have a long service history behind them; there should be no surprises and gotchas. We also have fuel production and reprocessing lines set up to manufacture and recycle PWR fuel elements. Using less-well-known reactor designs could mean problems down the road, and we've been through that with the Advanced Gas-cooled Reactors (AGRs) we built in the 1980s. They're great reactors, even more efficient than PWRs but they cost a lot more to build and operate, and their unique fuel rods require special and expensive handling.
You (I presume you are American) already have one. It's called... ITER. The US whined and moaned over the possibility of ITER being built in France, delaying its start by a couple of years. Eventually the US was over-ruled by nearly all the other countries who actually wanted to get on with developing fusion as a possible power source. Cadarache in France was finally chosen as the site and the project is now up and running.
There are a number of other fusion research projects going on around the world, but ITER's job is to figure out how to build a magnetic-containment fusion reactor that will deliver electrical power into a grid, and it's the biggest and best-equipped (and best-funded) of the current fusion projects.
EDIT: I may have spoken too soon, as I'm not sure the US has actually agreed to contribute to funding ITER, at least in this budget cycle.
Philips made radio tubes for the Wehrmacht, Kriegmarine and Luftwaffe during WWII while Holland was occupied by the Nazis. Remarkably the tubes suffered a high failure rate, but only after several hours of flawless operation, enough to get them past inspection and initial fitment but not much longer. Odd that.
The pistol they used in the test at the Royal Armoury was not particularly modern -- it was a GI-standard Colt 1911A1 firing milspec .45ACP ball ammo.
From yesterday's local news on the BBC:
http://news.bbc.co.uk/1/hi/scotland/edinburgh_and_east/7864982.stm
The police want to talk to his donkey. Afterwards, they want the donkey to shut the fuck up.
"So the government gets to choose who will watch over the government?"
All the parties represented in the Commons put forward candidates for a Peerage in proportion to the number of seats they hold. They are often party politicians but not always -- for example Mary Warnock was made a Baroness and Life Peer in 1984 after her work chairing the Committee of Enquiry into Human Fertilisation and Embryology. As a member of the House of Lords she has been instrumental in shaping the legislation that has helped make the UK a world leader in cloning and embryology research.
"A cop can just take you from the street and lock you up for 42/28 days without anyone asking why(like a judge ...)"
Ummm, no. Longest detention without judicial oversight is 24 hours in the UK. After that a suspect's case is brought before a judge who can order an extension of detention without charge once the police have laid out the reasons they want to hold the suspect in this manner. After that the detention order is reviewed regularly by a senior judge who will usually give the police a few days at a time before they have to come back and explain themselves. If it stretches out the judge will want to hear really really good reasons why the detention order should be extended at all.
In the very rare case of someone being held for the existing 28-day period without charge the case for detention will be reviewed by the judiciary at least five or six times. It's almost unknown for someone to be held for an extended period and then released without charge. The usual reason for the police to do this is because there are more serious charges being investigated -- e.g. the suspect has been found in possession of explosives but they're trying to track down who supplied them, if they are implicated in other crimes etc. They *will* be charged with something in that circumstance.
They're available in various x86 processor and RAM capacities, including Transmeta CPUs; the older models (Celeron 266, PII 300) are quite cheap today. Fat battery packs are available (Sony branded ones cost serious bucks but 3rd party units are a lot cheaper) that will run to 12 hours or more uptime. Replacing the HDD with a SSD will save you more battery power. Linux is readily ported onto most of the C1 variants and they all have PCMCIA or CardBus slots to support WiFi.
The accessory I regret not getting for my old PB was a ballistic-nylon shoulder holster for carrying it around.
They need to be very reliable -- they're going to be running 24/7/365, handling several kilowatts of power, operating as efficiently as they can so they don't waste energy unnecessarily. If I had to design something like that I'd overrate everything by a factor of 5 or more to ensure reliability and uptime and doing that will add to the pricetag.
There's also other factors governing the cost. These inverters aren't mass-manufactured as such, and the makers need to provide a long-term maintenance opreration with technicians and spares on tap, not just for warranty servicing. There's also liability and type testing -- having a few of these inverters bursting into flames due to a design defect would be bad news, financially speaking.
The reactor at Torness is an Advanced Gas-cooled Reactor (AGR), a British design built in the 1980s and optimised for efficiency. As it turned out, fuel costs for nukes haven't risen particularly over the last couple of decades so fuel efficiency isn't as important as the designers thought it would be. That's why the UK built cheaper GE-designed Pressurised Water Reactors (PWR) after the last of the AGRs was commissioned. They're not as efficient in turning fuel into power as the AGRs but they're cheaper to build and operate and they were available off a production line instead of being hand-built one-offs.
Here's the tale of a visit to the Torness nuclear power station by an old denizen of Slashdot.
I believe the higher capacity drives will force a rethink on how data is stored and accessed on standalone machines like laptops and desktops. I've only got a couple of terabytes of data on this machine and doing a file search over the five (I think, I can't actually remember how many drives I've got fitted in this thing) disks is already pretty time-consuming. The solution will be to add intelligence to the disk interface so that data indexing is done pre-emptively and the results cached on the fly.
The first generation of hybrid drives are already here but they're only at the beginning of their development cycle. HDD recording densities will increase as will flash RAM densities and that will improve access times but only for the most commonly accessed data.
Imagine a 10Tb HDD built in the classic 3.5" wide form factor, with 256Gb of 1024-bit-wide 150MWord/sec flash memory or MRAM on the controller board acting as cache. The spinning disk becomes a backing store for the flash where data is kept "fresh" by a smart algorithm. The drive spins down intelligently when not needed, saving power and reducing heat dissipation.
Higher recording densities are only one part of the future of disk drive technology.
OFC wire has a use in audio, but not for speakers. Oxygen-free copper has almost no copper oxide crystals in its makeup hence there's no piezoelectric noise generated when the wires flex. It makes great cable for microphones and guitars and other instruments that move during a performance and which carry low-amplitude signals. The difference in resistance between regular copper and OFC is almost immeasurable.
Speaker wires shouldn't move when they've been laid into place. Even if they did they carry high voltages and significant current which dwarfs any induced piezo noise from oxide inclusions.
As for the skin effect also mentioned by somebody above, it only comes into play in the multimegahertz frequency range. Radio amateurs buy silver-plated coax for this reason but it only makes a noticable difference in attenuation at VHF frequencies and above. At audio frequencies there's no skin effect worth mentioning. Multicore wire doesn't enhance the skin effect as it's an electromagnetic effect and the skin effect would tend to concenetrate the current flow in the outside of the entire wire bundle, not on each individual core.
If you want really good speaker wires get the heaviest solid twin-core mains wiring you can get your hands on and use the shortest single runs you can lay down between the amplifier and the speakers. Screening the cable from mains pickup may do some good but is usually pretty irrelevant given the currents running in the wires. Solder the speaker wire directly onto the speaker connections or the crossover unit in the speaker cabinet, don't use interconnects.
If you're paying more than a dollar a metre for speaker cable, you're paying waaay too much. Home Despot is your friend.
"Halting State" by Charlie Stross is out in October. It starts with a bank-raid in an MMORPG by a bunch of orcs with a dragon for fire-support and then gets weird.
This is blue-sky thinking but...
Right now, Google is building infrastructure -- distributed data centres, lighting up dark fibre trunks etc. That way they've got their own backbone capacity that can't be leveraged away from them with price hikes or lock-in deals between the existing backbone operators and the big telecomms/data providers.
Step two would be the biggie -- Google subcriber fibre. They enter the "last-mile" business and provide their own seriously high-capacity home and small business connections. The Big Guys can't touch them at this point since they have their own backbone operation already in place, not requiring essential ATM hookups that are a de facto momopoly. It's a Google-owned operation end-to-end.
The future is higher capacity, higher bandwidth for the individual user. BitTorrent, Joost and other streaming technologies are in their infancy but lots of distributed data centres are ideal for supporting that kind of messy ad-hoc multiple-connection data model. Google's planning seems to fit that vision of the future, but only if it gets down and dirty in the supply business rather than just standing off and expecting others to carry their content (if they feel like it) to the end-user who pays the bills.
Step 3? Profit! Expect the dinosaurs to scream like raped apes when something like this is sprung on them -- they believe they have a de-facto monopoly on last-mile (hence the Net Neutrality fan-dance), but it's something that could easily be overturned by the right iconoclastic company.
There were a number of factors that made the IBM team go for the Intel solution rather than the superior 68000 chip from Motorola. One factor was, as you said, the 8088 chip could use plentiful and well-understood 8-bit peripheral chips like the 8251, the 8259 etc. Another factor was that the 68000 had a long gestation -- Motorola had problems producing chips in quantity that could actually run at their specced speed of 8MHz. I played with an early dev board which had a working 68000 but it only ran at 4MHz. At that time the 8086/8088 were already available on the market, in quantity and that's what IBM needed.
The biggest factor though was probably the software. The x86 architecture was structurally similar to the venerable 8080 family's architecture (hence the memory segmentation, 8/16 bit registers etc.), and there was a lot of 8080-family code already out there, including developer toolsets like compilers and assemblers. Intel also released conversion tools for coders to convert existing 8080 code to run on their new 16-bit CPUs. The M68k was radically different structurally from Motorola's 6800 8-bit chips and its instruction set (although a dream to code new stuff for) did not make transitioning older 8-bit code easy.
Olden days, I was doing microprocessor coding for robotics. We used an assembler/linker package running on a couple of 16-bit minicomputers. One machine had enough RAM to do the assembly, the other had a hard disk that let us do the link of the object code to produce an image we could blow into an EPROM. Neither machine could do both parts.
Problem was the two machines were in adjacent rooms and networking was nonexistent. We'd run the assembly process, produce a paper tape of the object code, carry it to the other machine and then feed it into its paper-tape reader to do the link. This took a lot of time which delayed development.
I noticed the assembly process ran quite quickly but the linker machine was noticeably slower, and the machines weren't actually that far apart in reality. I got a couple of empty paper-tape spools and clamped them to the doorframes to act as guides, then I took an assembly tape being punched by the first machine out its room, around the doorframe, into the next room and fed it into the other machine's reader. A garbage bin acted as buffer storage for overflow tape, and it worked. I had to baby-sit it in case of jams or tangles but it cut the total assembly/link processing time by at least an hour.
The ultimate Bond Villain accessory toy.
http://www.segatoys.co.jp/yumeneko/index.html
Whatever you do don't pull its tail.
COBOL... lessee, you're a bank shuffling billions of dollars around every day, reconciling accounts after business has closed. You've got a pile of 30-year-old continuously maintained COBOL doing the shuffling and adding up those BCD dollar and cents.
And then a whizzkid from down the corridor in the website support office suggests you abandon all that working, time-tested code and rewrite it all in some open-source incredibly cryptic code that looks like line-noise, that doesn't support native BCD, that...
It's difficult to screw up COBOL code by putting a period in the wrong place, or missing out an ampersand. It won't compile, it won't run, it won't drop a billion dollars down a rathole. There's a reason it was designed like that and that is the reason it is still widely used.
Why do some programming languages (C, Perl) have to be cryptic? It doesn't save much time in the compile, it produces executables that are the same size as non-cryptic languages like Pascal or COBOL. About the only reason I can see is it saves the programmer some effort at the keyboard. Is that all?
The SR-71 required multiple tanker operations for any sort of operational flight. The profile was usually to tanker up in international airspace, penetrate "enemy" airspace at high speed and high altitude burning fuel like crazy, snap pictures over the target (assuming it was not obscured by clouds or smoke) and then dash out and then refuel again. It took days or even weeks to prepare any given SR-71 recon mission as refueling mistakes could cost them the plane (as it did in one instance when the wrong sort of fuel was loaded onto a tanker).
The SR-71's fuel wasn't stock Jet-A, it was a specially concocted mix that I've heard cost about as much as single malt whisky, litre for litre.
"The evidence that Microsoft does this and uses that information... Is?"
Clippy. For the 98% of people out there who aren't computer-literate uberhackers, Clippy is help and succour personified when the damn computer doesn't do what they want. Once they're up to speed they can dispense with Clippy's assistance, or switch it back on when they try something new. It's like training wheels -- once you're a Kool Kid you look down on the dorks still cycling around in circles on four wheels and carefully forget you were ever the same as them.
MS made a careful and calculated decision to put Clippy and co. into their premier software package and they made it after talking to a lot of potential customers, studying them carefully and figuring out what they needed and how to make it available in a way they would want to use. Blind guesses and stabs in the dark are how GPL UIs are created (I was going to say "designed" but there's little evidence of rational thought behind most OSS UI design decisions) which is why nearly all of it is a pain to use, even if the functional code underneath is competent.
If you want ten million people to use a piece of software you need to test its UI on thousands of subjects, young and old, techies and computer illiterates. You need to run it on laptops, on CRTs, on LCDs, in rooms with bad lighting and in direct sunlight. You need to find people with palsy who can't control a mouse, people with colour blindness and generally bad eyesight, people with dyslexia, people whose first language isn't English, and you can't use them more than once for any given piece of software testing.
To analyse those results you need teams of psych grad and industrial T&M researchers to go over the videotapes of each subject with stopwatches measuring each eye movement, each hesitant gesture, each mistake, each fumble to produce big databases of results and dig deep for the golden mean that gives the average user the best result in the UI. You then feed that info to your software designers who go away and produce the next iteration of code that has to go through the same horribly expensive (in time AND money) testing mangle again. And again. And again.
Aha! What about beta testing, you cry? Sorry, all that UI development is done a long time before anyone outside the company ever sees a beta release version. By then they're hunting bugs and fine-checking that the techies who will run and install this stuff for users can do their part in making ten million copies work; the UI is pretty much set in stone by then and a hundred million dollars has already been spent.