What is the main obstacle to building a craft that uses its engines to reduce its speed to well below orbital velocity while "hovering" outside the atmosphere in a non-sustainable orbital path until it's slow enough to reduce stress from air resistance and heat?
Fuel. It's technically possible to descend on rockets. Moon landings were done that way. But getting rid of orbital velocity costs as much as getting up to orbital velocity. Worse, you have to haul the fuel upstairs. That means stages, lots of stages. Look at Apollo - they launched something fifty stories tall and got back something the size of a mini-van.
Even if IBM settles with SCO, it's not over. There's other litigation. Red Hat is sueing SCO. There's an injunction against SCO in Germany. And sending out DMCA notices to Fortune 500 companies is sure to result in litigation.
There's going to be a break in the case this month, though. The judge gave SCO 30 days to state exactly what the supposed "infringements" are. Those 30 days run out on January 12, 2004.
The US Government bought a big share in Iridium, for which they basically get all the airtime they want.
When the Government bought in, Afghanistan and Iraq were still in the future. After the US bombed, invaded and occupied Afghanistan and Iraq, the people on the ground needed communications. Iridium is providing them. Without Iridium, the US probably would have spent more money frantically setting up communications systems than
Iridium cost.
Iridium handsets seem large by cell phone standards, but military radios with long range capability are still a backpack item or worse. There's more network capacity in the Iridium system than in military commo nets, and you can call any phone in the world.
Think of it as an instrument of empire, like the British East India Trading Company, not a business.
"Metrosexual" is a marketing term, intended to encourage shopping.
There's a delightful interview with Georgio Armani in the Economist a few weeks back that touches on this. He points out how hip-hop, unlike rock, induces young people to spend money on decorative objects and clothes.
I'd wondered why a played-out genre was so heavily promoted. Now I know.
This is incompatible with both the Microsoft business model and the design of Linux. But it's not impossible.
We could, in theory, have secure message-passing microkernels enforcing a mandatory security model running on secure machines with machine-checked proofs of correctness of both the code and the hardware at the VHDL level.
But every project to build such a thing has produced only a toy OS. All the verification projects are dead. C and C++ are hopeless for code verification. Java isn't really suitable for a low-level OS.
I worked on proof of correctness technologies in the 1980s. We didn't have enough CPU time for program proving back then; it took about 45 minutes on a VAX 11/780 (1 MIPS) to do the proofs for a 1000 line program. That would equate to about three seconds today. In a real proof environment, you're doing this about as often as you compile, so the proof process has to be reasonably fast. (You can cache quite a bit, though, and save time on reruns.)
But all this is stuff so far in advance of the crap we have today that it's not worth doing.
We don't even do obvious things like run browsers in jails, with a connection to a window and the net but no write access to anything else.
I would point out that the Junk Fax problem is largely a thing of the past.
Exactly. And I've received only one (1) junk phone call since Do Not Call went into effect.
Stopping spam can be done very effectively by legal means if the penalty applies to the person collecting the money. You can usually follow the money without much trouble. (If you can't, that's money-laundering, which is a bigger crime than spamming.)
Define adware, spyware, etc. as "computer intrusions" under the Computer Fraud and Abuse Act.
Require European-style "explicit, separate permission" for installs of such things as public policy, so they can't hide "permission" in some vague EULA.
Get Visa International to require that any site that accepts credit cards must have a digital certificate that gives the identity of the merchant, including their true name, DUNS number, and primary banking relationship. Any credit card site that doesn't do this gets their transactions reversed as soon as someone reports them to Visa.
That will make it easy to follow the money.
Enforce the new Federal spam law. It's weak, but it's something. A few high profile raids and arrests will do it.
Lobby the FTC on the details of the "do not spam" list. Insist that opt-out by domain be supported.
Insist that mail service providers that don't opt out their customers be required to disclose this in ads.
That's not enough.
Do they make a FAT file system image with long names during the boot process? And Microsoft can still break dual-booting on their side.
The basic question is this: will mainstream machines sold in retail channels still be able to run a non-Microsoft OS?
There are several ways EFI could discourage Linux use:
The boot firmware might refuse to boot anything other than a signed operating system. The XBox is that way now. Microsoft would like that, but it's a bit too blatant and might attract unwanted antitrust attention, especially in the European Union.
Dual booting probably won't work for Microsoft operating systems. That's quite likely with Palladium-type DRM. Interposing a new boot loader breaks the chain of trust between the firmware and the OS. A Microsoft OS can tell that an unauthorized component has run before it loaded, and can refuse to run. Even if it runs, it won't let you access "protected content" (movies, music, etc.).
This will discourage casual Linux use.
You'll have to dedicate a machine to Linux.
The boot firmware might require some technology for which Microsoft has intellectual property rights. Microsoft's new push on enforcing their rights to long-name FAT file systems indicates a step in this direction. Does EFI require a long-name FAT file system? If it does, you won't be able to build a bootable Linux image without paying Microsoft. You'd probably have to build Linux images using a Microsoft machine. So the Linux-only user won't be able to build a bootable image from source.
There may be undocumented proprietary aspects to EFI that make it hard to write to. Watch for incorporation by reference of some Microsoft standard into the EFI standard. This is already a major problem with graphics cards.
It's a subtle strategy. It's not going to be impossible to boot Linux, but it looks like it's about to become more difficult.
It will still be possible to build machines that run Linux, and there will be companies that do so
and preload Linux.
But they'll make up their own distribution, like the Thiz Linux you find at Wal-Mart.
End user installation of Linux will decrease.
Red Hat's air supply will be cut off.
Once you see the whole strategy, you realize just how clever Microsoft is being about this. It's not so blatant as to provoke screams from the industry, but it's enough to put a big dent in Linux installs.
OnStar could be a great aid to terrorists. Plant bombs, and wait until the target drives by to detonate them. Load in a list of targets (politicians, cops, judges, journalists) and the next one to drive by gets it.
It's backwards client server
on
Skip The IP Address
·
· Score: 2, Insightful
That's an amusing approach.
The author must be a X-windows fanatic. He uses the terms "client" and "server" backwards. The end that sits there passively waiting for someone to connect is called the "client", and the end you run when you want to talk is called the "server".
Note that the "client" opens an Ethernet interface in promiscuous mode, so if you put this on a machine on a busy network, it's going to spend most of its time discarding packets.
Buy Deep Fritz 8 now for only $96.95! Get trounced.
Runs on 1 to 8 CPU IA-32 machines. It beat Gary Kasparov running on a 4-CPU 2.8GHz Xeon machine.
The machine used for that game was in rackmount packaging, but it fit on a tabletop.
You can buy workstations with an equivalent configuration.
This idea resurfaces every few years. It can be done, but the disadvantages outweigh the advantages. The unsprung weight problem restricts it to heavy, slow vehicles. It's been tried for tanks, rail transit, and now buses.
Rail transit has used motors in the wheel trucks for a century. Frank Sprague did the original motor design around 1890. But the motors are not actually in the wheels.
This is yet another attempt to breathe life into the boondoggle of massively parallel architectures.
Over the last few decades, there have been many exotic parallel architectures. Dataflow machines, connection machines, vector machines, hypercubes, associative memory machines (remember LINDA?), perfect shuffle machines, random-interconnect machines, networked memory machines, and partially-shared-memory machines have all come and gone. Some have come and gone more than once. None has been successful enough to sell commercially in quantity.
Very few of these machines have ever been purchased by any non-government entity.
There are two ends of the parallelism spectrum - the shared-memory symmetrical multiprocessor, where all memory is shared, and the networked cluster, where no memory is shared. Both are successful and widely used. Everything in between has been a flop.
Despite decades of failure, people keep coming up with new bad ways to hook CPUs together, and getting government agencies to fund them. It's more a pork program than a way to get real work done.
By the time one of these big wierdo machines is built, debugged, and programmed, it's outdated. A few years later, people are getting the same job done on desktops. Look at chess. In 1997, it took Deep Blue to beat Kasparov. Kasparov is now losing games to a desktop four-processor IA-32 machine.
Figuring out more effective ways to use clusters is far more cost effective than putting a National Supercomputer Center in some Congressman's district in Outer Nowhere. There's a whole chain of these tax-funded "National Supercomputer Centers".
The "Alabama Supercomputer Center" has ended up as an ISP for the public school system, hosting E-mail accounts and such. It's all pork.
Really bad idea: Kazaa/Brilliant Digital unloading that computational task onto all their zombies, the Kazaa clients? They can do it; read the Kazaa customer agreement. They 0wn your cycles.
Negligence by the retailer in following the supplier's instructions may make the retailer liable to the supplier, but that doesn't relieve the supplier of any obligations.
More popular in the 1980s.
on
Source Code Escrow
·
· Score: 4, Informative
Source code escrow was quite common in the early days of microcomputer software. Back then, the software companies were little and their customers were big, and they had to keep the big companies happy. Now, it's the other way round.
Some of the early source code escrow companies themselves went bust. You need a software escrow agent that's likely to be around for decades.
There are still companies offering software escrow services, but it's a minor business.
For historical reasons, revolving around the "process table" design in PDP-11 UNIX, the schedulers in UNIX-like OSs were originally very dumb. The one in V6 UNIX was so bad that if three processes were compute-bound, the scheduler would stall.
Mainframe schedulers have been O(1) for a long time. The one for UNIVAC 1108 EXEC 8 was O(1) in 1967. (And you can still buy a Unisys ClearPath server with that code in it.)
The traditional implementation is a set of FIFO queues, one for each priority. For systems with only a few priorities, that's just fine. As the number of different priorities increases, you spend too much time checking the empty queues, so there are tree-like structures to get around that problem.
Scheduling today is more complex because of cache issues. If you context-switch too much, you thrash in the CPU caches. On the other hand, everybody has so much memory today that paging is mostly unnecessary.
From an antitrust perspective, this is right. The labels blew it so badly in online music distribution that they failed to achieve significant market share. They tried, but failed through sheer incompetence.
If you read the ReplayTV revision history, it looks like each new release removes functionality. "Commercial advance" was dropped when the 5500 series came out.
Because the purchasing public continually searches for the best buy, and regards the offer of ``Free'' merchandise or service to be a special bargain, all such offers must be made with extreme care so as to avoid any possibility that consumers will be misled or deceived....
when the purchaser is told that an article is ``Free'' to him if another article is purchased, the word ``Free'' indicates that he is paying nothing for that article and no more than the regular price for the other. Thus, a purchaser has a right to believe that the merchant will not directly and immediately recover, in whole or in part, the cost of the free merchandise or service by marking up the price of the article which must be purchased, by the substitution of inferior merchandise or service, or otherwise.
...
When making ``Free'' or similar offers all the terms, conditions and obligations upon which receipt and retention of the ``Free'' item are contingent should be set forth clearly and conspicuously at the outset of the offer so as to leave no reasonable probability that the terms of the offer might be misunderstood. Stated differently, all of the terms, conditions and obligations should appear in close conjunction with the offer of ``Free'' merchandise or service.
That's clear enough. It's binding on the supplier as well as the retailer; the supplier can't pass the buck here.
Any advertisement, including any advertisement over the
Internet, soliciting the purchase or lease of a product or service,
or any combination thereof, that requires, as a condition of sale,
the purchase or lease of a different product or service, or any
combination thereof, shall conspicuously disclose in the
advertisement the price of all those products or services.
The Bunker, in Britain, offers bombproof web hosting, in their underground data center inside an underground military base. Starting from 125 pounds per month for a dedicated 1U server. Linux hosting available.
Some wall transformers use switching power supplies, with a small high-frequency transformer after the chopper, and some are linear, with a front-end transformer running at 50/60Hz.
Worth mentioning is the forged UL safety label problem. If a power supply doesn't have a UL or CSA label, don't buy it. It will probably catch fire if shorted or even heavily loaded. A UL label must have the UL logo and a certification number. UL certifications can be looked up here. This is worth doing for desktop computer power supplies, for which forged certifications are a big problem.
The ones that catch fire are almost invariably uncertified. UL requires that no single component failure can cause a fire.
Slashdot Mirror
Fuel. It's technically possible to descend on rockets. Moon landings were done that way. But getting rid of orbital velocity costs as much as getting up to orbital velocity. Worse, you have to haul the fuel upstairs. That means stages, lots of stages. Look at Apollo - they launched something fifty stories tall and got back something the size of a mini-van.
There's going to be a break in the case this month, though. The judge gave SCO 30 days to state exactly what the supposed "infringements" are. Those 30 days run out on January 12, 2004.
Iridium handsets seem large by cell phone standards, but military radios with long range capability are still a backpack item or worse. There's more network capacity in the Iridium system than in military commo nets, and you can call any phone in the world.
Think of it as an instrument of empire, like the British East India Trading Company, not a business.
There's a delightful interview with Georgio Armani in the Economist a few weeks back that touches on this. He points out how hip-hop, unlike rock, induces young people to spend money on decorative objects and clothes.
I'd wondered why a played-out genre was so heavily promoted. Now I know.
Type "program verification" into Google, and you get a "Work at Google" paid ad.
We could, in theory, have secure message-passing microkernels enforcing a mandatory security model running on secure machines with machine-checked proofs of correctness of both the code and the hardware at the VHDL level.
But every project to build such a thing has produced only a toy OS. All the verification projects are dead. C and C++ are hopeless for code verification. Java isn't really suitable for a low-level OS.
I worked on proof of correctness technologies in the 1980s. We didn't have enough CPU time for program proving back then; it took about 45 minutes on a VAX 11/780 (1 MIPS) to do the proofs for a 1000 line program. That would equate to about three seconds today. In a real proof environment, you're doing this about as often as you compile, so the proof process has to be reasonably fast. (You can cache quite a bit, though, and save time on reruns.)
But all this is stuff so far in advance of the crap we have today that it's not worth doing. We don't even do obvious things like run browsers in jails, with a connection to a window and the net but no write access to anything else.
That's why I got out of security years ago.
Exactly. And I've received only one (1) junk phone call since Do Not Call went into effect.
Stopping spam can be done very effectively by legal means if the penalty applies to the person collecting the money. You can usually follow the money without much trouble. (If you can't, that's money-laundering, which is a bigger crime than spamming.)
That's not enough. Do they make a FAT file system image with long names during the boot process? And Microsoft can still break dual-booting on their side.
There are several ways EFI could discourage Linux use:
It's a subtle strategy. It's not going to be impossible to boot Linux, but it looks like it's about to become more difficult.
It will still be possible to build machines that run Linux, and there will be companies that do so and preload Linux. But they'll make up their own distribution, like the Thiz Linux you find at Wal-Mart. End user installation of Linux will decrease. Red Hat's air supply will be cut off.
Once you see the whole strategy, you realize just how clever Microsoft is being about this. It's not so blatant as to provoke screams from the industry, but it's enough to put a big dent in Linux installs.
Too risque for American TV, the Trojan Games ads have run in the UK.
OnStar could be a great aid to terrorists. Plant bombs, and wait until the target drives by to detonate them. Load in a list of targets (politicians, cops, judges, journalists) and the next one to drive by gets it.
The author must be a X-windows fanatic. He uses the terms "client" and "server" backwards. The end that sits there passively waiting for someone to connect is called the "client", and the end you run when you want to talk is called the "server".
Note that the "client" opens an Ethernet interface in promiscuous mode, so if you put this on a machine on a busy network, it's going to spend most of its time discarding packets.
Send this guy a roll of duct tape.
Buy Deep Fritz 8 now for only $96.95! Get trounced. Runs on 1 to 8 CPU IA-32 machines. It beat Gary Kasparov running on a 4-CPU 2.8GHz Xeon machine. The machine used for that game was in rackmount packaging, but it fit on a tabletop. You can buy workstations with an equivalent configuration.
Rail transit has used motors in the wheel trucks for a century. Frank Sprague did the original motor design around 1890. But the motors are not actually in the wheels.
Over the last few decades, there have been many exotic parallel architectures. Dataflow machines, connection machines, vector machines, hypercubes, associative memory machines (remember LINDA?), perfect shuffle machines, random-interconnect machines, networked memory machines, and partially-shared-memory machines have all come and gone. Some have come and gone more than once. None has been successful enough to sell commercially in quantity. Very few of these machines have ever been purchased by any non-government entity.
There are two ends of the parallelism spectrum - the shared-memory symmetrical multiprocessor, where all memory is shared, and the networked cluster, where no memory is shared. Both are successful and widely used. Everything in between has been a flop.
Despite decades of failure, people keep coming up with new bad ways to hook CPUs together, and getting government agencies to fund them. It's more a pork program than a way to get real work done.
By the time one of these big wierdo machines is built, debugged, and programmed, it's outdated. A few years later, people are getting the same job done on desktops. Look at chess. In 1997, it took Deep Blue to beat Kasparov. Kasparov is now losing games to a desktop four-processor IA-32 machine.
Figuring out more effective ways to use clusters is far more cost effective than putting a National Supercomputer Center in some Congressman's district in Outer Nowhere. There's a whole chain of these tax-funded "National Supercomputer Centers". The "Alabama Supercomputer Center" has ended up as an ISP for the public school system, hosting E-mail accounts and such. It's all pork.
Really bad idea: Kazaa/Brilliant Digital unloading that computational task onto all their zombies, the Kazaa clients? They can do it; read the Kazaa customer agreement. They 0wn your cycles.
Bad idea. It will just result in spammers taking over even more zombies to spam.
Negligence by the retailer in following the supplier's instructions may make the retailer liable to the supplier, but that doesn't relieve the supplier of any obligations.
Some of the early source code escrow companies themselves went bust. You need a software escrow agent that's likely to be around for decades. There are still companies offering software escrow services, but it's a minor business.
Iron Mountain has a software escrow business, and they offer some stories of software released from escrow. The most common situation is bankruptcy of a supplier.
Mainframe schedulers have been O(1) for a long time. The one for UNIVAC 1108 EXEC 8 was O(1) in 1967. (And you can still buy a Unisys ClearPath server with that code in it.)
The traditional implementation is a set of FIFO queues, one for each priority. For systems with only a few priorities, that's just fine. As the number of different priorities increases, you spend too much time checking the empty queues, so there are tree-like structures to get around that problem.
Scheduling today is more complex because of cache issues. If you context-switch too much, you thrash in the CPU caches. On the other hand, everybody has so much memory today that paging is mostly unnecessary.
From an antitrust perspective, this is right. The labels blew it so badly in online music distribution that they failed to achieve significant market share. They tried, but failed through sheer incompetence.
As for what Replay is doing with their "free" offer, it appears to violate the Federal Trade Commission Guidelines for use of the word "Free". These are quite specific.
when the purchaser is told that an article is ``Free'' to him if another article is purchased, the word ``Free'' indicates that he is paying nothing for that article and no more than the regular price for the other. Thus, a purchaser has a right to believe that the merchant will not directly and immediately recover, in whole or in part, the cost of the free merchandise or service by marking up the price of the article which must be purchased, by the substitution of inferior merchandise or service, or otherwise.
That's clear enough. It's binding on the supplier as well as the retailer; the supplier can't pass the buck here.
California also requires this: (Business and Professions Code 17509).
-
Any advertisement, including any advertisement over the
Internet, soliciting the purchase or lease of a product or service,
or any combination thereof, that requires, as a condition of sale,
the purchase or lease of a different product or service, or any
combination thereof, shall conspicuously disclose in the
advertisement the price of all those products or services.
That's clear enough.The Bunker, in Britain, offers bombproof web hosting, in their underground data center inside an underground military base. Starting from 125 pounds per month for a dedicated 1U server. Linux hosting available.
Worth mentioning is the forged UL safety label problem. If a power supply doesn't have a UL or CSA label, don't buy it. It will probably catch fire if shorted or even heavily loaded. A UL label must have the UL logo and a certification number. UL certifications can be looked up here. This is worth doing for desktop computer power supplies, for which forged certifications are a big problem. The ones that catch fire are almost invariably uncertified. UL requires that no single component failure can cause a fire.