Here's why no FreeS/WAN ipsec shipped in RedHat
on
Red Hat 7.2 Released
·
· Score: 2
The FreeS/WAN ipsec project (Mailing list here) has been developed entirely outside the US to prevent US export laws from restricting its distribution. The current version of the US export laws doesn't currently restrict it, and RedHat could probably ship it if they wanted to, but the laws have changed a number of times, and they only way to prevent the US government from changing them again to reimpose the previous restrictions is to continue not to accept US code. That was the policy before the terrorist attacks, and the FBI is now trying to grab every bit of control and access they can, so continuing to refuse to accept US contributions is an unfortunately wise policy.
That doesn't mean that FreeSWAN is easy to install - until the next version (Real Soon Now, probably within the month, which is supposed to do RPMs) you start off by doing a clean compile of your kernel, installing the FreeSWAN code, and compiling your kernel again, then configuring the actual config files. But the process is independent of the RedHat organization or anybody else developing code in the US, so they'll stay free. I've heard that Mandrake 8.1 comes with it installed, but I haven't tried it yet.
On the other hand, there's also PGPnet IPSEC for Windows, and IPSEC releases like Kame for some of the BSDs.
By the time you've paid for film developing, the cost of the disposable digital is about the same as the disposable film camera. Quality's nowhere near as good, but if they also email you the files, it can be ok.
Don't be such a snob, dude!:-) Not everybody's an imaging professional. If you want a really BAD camera, my second digital camera was 160x120 for $39 a couple years ago...
640x480 is really just fine for typical web pages - pictures of your cat or your cousin's kids, and most really cheap cameras are that resolution. Sure, it's not what you'd get with your thousand-dollar SLR with really great lenses - this is the digital equivalent of an Instamatic.
The interesting quality thing they did here is that they're not compressing the image much - 8MB for 24 shots means they're storing pictures as ~310KB instead of the more typical ~75KB JPEGs that other digital cameras I've seen use for 640x480 images. I don't know if this means they're doing JPEG, or if they're doing some low-CPU compression algorithm and saving money on CPU, or doing 8-bit-per-pixel uncompressed images instead of more useful color depth (unlikely but possible, and that really *would* make color suffer.)
IBM's patent resembles the way the JCL Job Control Language for IBM mainframes worked. Nobody ever actually programmed new JCL - they took some old thing somebody else had around, tweaked the variables and file names, and added the occasional new line, but there wasn't any *real* programming done*, just code stealing. And just *try* to patent "stealing other people's code" as a programming technique.
More seriously, I do like the example someone posted of Netscape Gold as prior art. Also, HTML is not really much more than the old troff -mm macros with angle brackets around them, and there were WYSIWYG editors for nroff/troff around in the 80s. And when did Slashdot start? Or web-based conferencing systems? Did any of them predate Bianca Troll's Graffiti Wall? The stuff just ain't new.
* Actually, there was a guy named Eli who worked in the basement who *did* write original JCL, but he was one of those rare wizards. Regular users never touched the stuff.
The old Sun/tmpfs filesystem design was good for storing files that you expected to use up rather than keep for long periods of time. Files might get paged out to swap space if the system needed memory and they hadn't been used recently, but very common behaviour was that they'd be created by one phase of a compiler run, read by another, and then deleted, before the system had to bother paging them to memory, and/tmpfs was smart enough to garbage-collect pages from deleted files so they wouldn't need to get paged out.
I primarily use RAMDISKS on Windows to accomplish the same thing - it's a convenient place to stash files I don't plan to keep, such as MSOffice attachments in incoming email messages, as well as to stash data I don't want on real disks, like decrypted emails (yes, I know they can get paged out, and there's memory persistence, but KGB/FBI/NSA attacks are really low on my threat models compared to generic theft.) It's an amazing performance win for many Windows applications, and Windows is happier running them from "disk" than running them from the temp files used by email while keeping the email messages open.
Both of those applications assume they're using main system RAM, and that they're relatively disposable - if the system crashes in the middle of a compile, you'll re-run the compile anyway. Some of the other approaches to RAM-disk boxes provide separate battery backup, which gives you persistence.
There are a number of applications, including journaling databases and file systems, which require data to be written to non-volatile storage before they continue, and providing microsecond-latency writes to battery-backed ramdisk instead of few-millisecond latency for spinning disks makes a hugedifference when you're trying to do thousands of transactions per second.
One of the classic accelerators was the old Legato Prestoserve - it had a MB or so of battery-backed RAM, which was enough to provide a non-volatile buffer for the time it took for disk drives to write. Machines have gotten a lot faster since then, and disk drives now usually come with buffers of a few MB, but that kind of approach can still be a win.
There's a lot of cultural difference between what happens on the net and what happens on TV; TV is highly centralized information presentation, while the Net is highly decentralized - person-to-person email, plus web sites that range from individual rants to formal broadcasts by large news organizations, plus search tools that let you find things you're looking for without some editing service compiling them into a package for you. Esther Dyson has comented since at least Release 2.1 about the asymmetry between Net-based and centralized information sources, most famously commenting that the Net may be good for conspiracy and rumors, but TV is better at propaganda.
Look at the information you're seeing, and if you were old enough to be media-literate during the Gulf War, think about how the messages were managed then, including coverage on TV, news wire services, editorials, interviews with government sources. It was done better during the Gulf War because Bush Sr. could take his time, while Bush Jr. had this thrown at him, plus the press has a strong talent for going for the emotional, intense stories around the WTC scene, which creates an energy that Bush can use but can't control as easily.
Email was more useful than the web for the beginnings of the story - I first heard by phone call from a friend who'd been watching early-morning TV, and then started getting emails. CNN.com was slashdotted, and did extremely well getting anything at all up and running with that demand load - just because the web lets everybody publish information to everybody else doesn't mean you don't turn to a few centralized sites for breaking news:-) Email also had the advantage that it's much lower bit volume and scales better than the web because of the large peer-to-peer connectivity, and it has different failure modes than wired and wireless telephony so people near the affected sites could get messages out more reliably.
The net being what it is, I googled for Esther Dyson conspiracy propaganda and found a bunch of references including this interview with Esther Dyson:
AD: You mentioned in your book the different characteristics of the Internet and television, the former being an instrument of conspiracy and the latter an instrument of propaganda. Could you explain that a bit more?
ED: It is a bit of a simplification, but what I want to say is that propaganda is a centrally propagated formal truth, which can be good or bad. Conspiracy is an undermining, decentralized force. The point is depending whether the centralizing authority is good or bad, the conspiracy is good or bad. The role of the Internet is to be an undermining force. If I had to choose between the Internet and television, I would choose the Internet and its conspiracy over television and its propaganda.
Brin's point is that Moore's Law means we *will* have ubiquitous networked cameras whether we like it or not. Optics, electronics, networking, and storage keep getting cheaper rapidly - We *are* falling off the cliff - the issue is whether we're aiming on the way down. If you read Brin's book, he agrees there are lots of ways that this technology can be abused, but considers it to be much safer if we're watching the government, because they *will* be watching us.
Those radio-based X10 cameras are an example, as are $29 internet-cams, 802.11 wireless, and cheap bandwidth - if it weren't for anti-server policies on most cable modem companies, there'd be relatively common "Neighborhood Watch" cameras run by lots of random people. Cu-SeeMe quality fits in modem bandwidths - even 802.11 can handle dozens of broadcasts. The only place we can't easily watch is police offices.
ZKS's business model was always a gamble, but it was the critical thing they had to offer. There are a number of people in the cypherpunks community who've developed remailers, but the critical problem was how to keep enough of them running to provide security, diversity, and reliability. (The typical problem remailers have is keeping their ISPs from getting upset by complaints from recipients of unwanted anonymous mail, and running a faster-response-time system like ZKS requires comsuming more upstream bandwidth as well.) What their business model had to offer, besides a good enough friendly user interface to make it easy for the general public to use, was a financial incentive for ISPs to want to run remailers, because they're paying customers rather than problem users. We were never sure whether they'd succeed, but it would be a great thing if they did. Oh, well....
Safeweb is one of several anonymizing services, of which the first well-known one was www.anonymizer.com. There are a couple of serious problems with it, one technical, one trust-related. On the other hand, Triangle Boy is really cool.
The technical problem is that their service uses Javascript, and doesn't work if you're not running Javascript. That means that any time you're using the system, you're vulnerable to any other JS problems on any other web page your browser encounters, until you turn JS back off. IIRC, Safeweb does attempt to clean up JS and other dangerous stuff from pages it displays to you, but it's still a risk. Also, I'm not that impressed with their Javascript, though I'm not an expert on the stuff - my problem was that under Mozilla ~0.91, they pop up windows to do the secure browsing in, and they're not really quite the shape of my screen, though that could have been Mozilla's fault. I sent email to the Safeweb folks about the fundamental "You're using Javascript" problem, and got a really prompt reply from their technical management, which was good, but they fundamentally didn't get it, which bothered me.
The other problem is trust - in general, you always need to be concerned about whether a service like this is trustable, both because of the intent of the people running it (are they ratting you out to somebody) and the security of their systems (if their server is 0wned by CrackerZ, you're not secure.) As I mentioned, Triangle Boy is really cool - it's a sort of distributed set of volunteer-run anonymizing servers, which keep moving around to prevent blocking services from blocking them, and Safeweb announced that they were going to be using this to provide censorship-free web access for people in China, the Middle East, and other places with censorship problems. The catch - they've got funding from In-Q-Tel, the CIA venture fund. It's probably entirely legit, and certainly good enough for most purposes - but how paranoid you need to be depends on who's really out to get you. ZeroKnowledge was very upfront about what their trustability levels were (plus I knew the folks there, and they were well-connected to the cypherpunks community.)
While The Bunker is a technically cool site, and provides lots of entertainment for its PR folks, one of the really cool things about this announcement is that it's an internet hosting/colo site that's actually doing well enough to need more real estate. I don't know how much is overall growth in the UK and European markets for hosting/colo, and how much is individual competence out-running the other players. But it's a major contrast to the US market, which is hopelessly overbuilt and glutted - companies who run them are either responding to the wonderful simultaneity of the dot-com crash and construction boom by going Chapter 11 or by building even more space to run their competitors farther into the ground:-) Many of them, of course, are doing what they can to create market differentiation, such as running consulting services, and there are major differences in the business plans of independent hosting centers, like Exodus, and the ISP-related hosting centers, like Genuity, AT&T, Level 3, etc. that are partly there to sell internet bandwidth.
Also, there have been whole ecologies of businesses around them, like companies providing dedicated managed computers in colo spaces, and companies providing shared hosting on the managed computers, and companies paying shared-hosting companies to market more shared hosting (either through legitimate business or pyramid scams), and ASPs running applications on the dedicated managed computers used by the shared-hosting customers as well as providing services to non-web-based business out in the real world, and spam hunting businesses trying to protect you from the pyramid-spammers selling shared hosting, and content provider businesses using the dedicated hosting to serve content to shared-hosting customers or other dedicated hosting customers, and billing companies providing billing services for those Internet businesses that actually can bill somebody, and advertising services trying to get the various hosting users to carry their ads.
Disclaimer: some of the folks who run The Bunker and also Havenco on Sealand are friends of mine, and my employer's also in that business, but I'm not speaking for any of them.
"Oh Dad, Poor Dad" - Wow, Blast from the past! I was in that play in high school summer theater, a few decades ago, playing "Dad", the corpse:-)
The big US backbone carriers don't peer at the NAPs - not enough bandwidth, not enough control over performance. They mostly do private peering with each other, and with any middle-tier folks big enough to do peering. International peering is a different story - the bandwidths have typically been smaller, and there are more complex arrangements between European networks and US networks (and the rest of the world gets even stranger.) The usual goal is to carry 90-95% of your other-tier-1 traffic by private peering, partly for performance and partly for economics. For instance, AT&T has a bit over 40Gbits/sec of private peering connections in the US, in addition to their NAP connections (though something like 5Gbits of that is the @Home network) - while the NAPs and MAEs are no longer 100Mbps shared fast ethernets, they're still not that big. On the other hand, most of the peering between big carriers is in the same cities as the NAPs, especially when there are carriers with POPs in the same telco buildings or carrier hotels, which is fairly common. So nuking a dozen big cities, or having Dr. Evil arrange simultaneous earthquakes in them or hire a large number of guys named "Bubba" with backhoes, will still cause major disruption, because there's still a lot of concentration there. (The Internet has liberated us from geography, making it possible for anybody to work from anywhere in the world, which is why everybody moved to the San Francisco Bay Area....)
If you're more realistically paranoid than that, look at the number of root domain servers. There's been recent discussion about what they're doing for security and reliability, mainly worrying about crackers disrupting the databases. (Beyond, of course, the bigger problems are the relationships between ICANN, NSI, the UDRP, disgruntled postmasters, etc.:-)
Vernor Vinge's book "A Deepness In The Sky" has a lot of discussion about what you could do with locator smart-dust; it's obviously speculative fiction, but it does a great job of looking at the potential for technology. Think about the effects of small (fictionally nanotech, but really small is probably enough) devices that communicate with their neighbors, have some computing power, and can do relative location detection. What could you do with that?
I can't find the obligatory Beowulf post:-) It's actually applicable here - one of the more interesting things to do with Smart Dust and similar locator technologies is for them to talk to each other about where they are and to detect changes in their relative positions. It's not just a server thing. Vernor Vinge's book "A Deepness In The Sky" has a lot of discussion about what you could do with locator smart-dust; it's obviously speculative fiction, but it does a great job of looking at the potential for technology.
I took one course in college that did team programming projects. (This was ~25 years ago, in a computer simulation course.) We had groups of 4-6 people to work on projects. The prof set up one ground rule, called Assassination, which says that any group of N-1 people can assign any grade they want to the Nth person. The primary purpose was to boot slackers, but it could also be used to reward people who deserved it.
Working in teams is one of the most critical skills that computer programmers need to learn - I hope it's taught more today than it was back when I was in school, though the institutional demand for gradability makes it hard to do. It's more common in graduate school, where research projects often need to be done as a team, but the current market has been for undergrad-producing factories to feed the dot-com boom - perhaps that'll change a bit.
Some of the important skills include how to coordinate work on shared code, how to divide up work when it is more separable, how to tell the difference, how to cooperate on design, how to find out the real requirements (since customers seldom hand you those on a plate, much less do so correctly...), how to tell when there are personality conflicts or ego problems, how to back off when you're the one with the ego problem (:-), how to match up different individuals' skills with the problems to be solved,.... Some of the new "extreme programming" approaches especially have to deal with these problems, but everyday business and amateur programming environments do as well.
Oh, and if by any chance Karl, Setsuko, Renee, or Temporary Emergency Acting Professor Eric are reading this, hi!
Quantum computing can be used for cryptanalysis, letting you solve problems, such as factoring, that are the core of cryptosystems like RSA and Diffie-Hellman. Quantum Cryptography is entirely different - it's a technique for sending bits securely down a fiber, using quantum techniques to tell whether someone's tried to eavesdrop on it. This is really useful if you've got a spare fiber connecting you to your recipient and you're worried about KGB eavesdroppers, but isn't too useful in the real world.
The early QC algorithms also had a significant chance of finding a wrong answer, with no way to control what you got. On the other hand, the interesting problems that they solve are NP-hard problems like factoring, for which you can quickly verify whether an answer is correct or not.
Most car stereos have Auxiliary jacks on them, so you can plug in other sound sources. Sometimes they're installed properly; sometimes the jack isn't reachable but it's still back there if you want to look for it. Portable MP3 players range from $100 El Cheapo sets to $260 Archos jukeboxes with 6GB laptop drives in them. Plug in , Turn on, Rock out. Depending on the voltage your MP3 player uses, you might want to get a cigarette-lighter adapter to power it, or hotwire from the back of the lighter, or especially for one of the lower-capacity units, just use rechargeable batteries (or builtins, if they have them.)
I'm in the ATM business (Asynchronous Transfer Mode, not Cash Machines, though we do them too), where Telemedicine has been a staple of hype vendors for decades. At least in the US, telemedicine is not realistic except for a few contrived situations, mainly because medical licensing is done on a state-by-state basis, at least for dramatic silly things like this. Usually anywhere that can afford robot doctors are high-tech enough big-city big hostpitals that have real doctors. There is a realistic case for remote support, which is a (human) doctor in one city where the patient is talking to a specialist in another city and sharing pictures back and forth. There may be emergency medicine situations where a paramedic needs to consult a specialist, but that's usually a wireless situation. If it weren't for the structure of the medical insurance systems dominating US medicine, there might also be applications for a nurse with a camera at a small office working with a doctor who's telecommuting from some other location, but the main situations where that makes sense are rural areas that don't have the right kind of doctor within an hour or two drive.
(In a non-insurance-dominated free market, people could pay for what they wanted, which would probably include cost-effective non-bureaucratically-oriented structures like that. And in a socialized-medicine market, you'd probably have either lots of doctors, if you believe its proponents, or not enough money for experimental technology, if you believe its opponents, or less restriction on what the medical service can do as long as it saves the service money.)
> Source Route Bridging anyone? try that with ethernet.
Well, that's just the point, isn't it? Source Route bridging was designed by people who Just Didn't Get It - it's doing almost all the work of a router without gaining any of the benefits. It doesn't deserve as much bashing as Microsoft, who also Never Did Get It about networking, because the folks who did SNA and its predecessors were trying to design protocols that would run well on systems with slightly more horsepower than digital watches, back when RAM actually cost money. It's possible to use Transparent Bridging on many kinds of IBM systems, and these days any of the mainframes will run TCP/IP. And almost all the applications can either be handled using tn3270 (since there aren't a lot of Genuine Original Green-Screen 3270 terminals left - almost everyone uses PCs with various boards in them) or if necessary encapsulated in IP using DLSW or other tunnelware.
Token ring was often more reliable than Real Ethernet Thick-wire with Vampire Taps - the mechanical connectors were better than just chomping your way into a coax cable with the possible risk of trashing the code enough to get reflections. I'm not sure if it was better than connectorized thinwire or not, and it certainly wasn't better than Cat5 10baseT, at least if you used genuine Cat5 and didn't cheat with Cat3 (or didn't use cable-TV coax instead of proper thinwire:-) Back when I used to deal with this awful stuff, I had one customer for whom the original Ugly Shielded Twisted Pair token-ring really did make sense - they were a city's Department of Water and Power, and they had lots of Really Big Electric Motors which were a really bad electrically noisy environment, and their network needed all the help if could get. These days I hope they're using fiber.
Slashdot Mirror
That doesn't mean that FreeSWAN is easy to install - until the next version (Real Soon Now, probably within the month, which is supposed to do RPMs) you start off by doing a clean compile of your kernel, installing the FreeSWAN code, and compiling your kernel again, then configuring the actual config files. But the process is independent of the RedHat organization or anybody else developing code in the US, so they'll stay free. I've heard that Mandrake 8.1 comes with it installed, but I haven't tried it yet.
On the other hand, there's also PGPnet IPSEC for Windows, and IPSEC releases like Kame for some of the BSDs.
By the time you've paid for film developing, the cost of the disposable digital is about the same as the disposable film camera. Quality's nowhere near as good, but if they also email you the files, it can be ok.
640x480 is really just fine for typical web pages - pictures of your cat or your cousin's kids, and most really cheap cameras are that resolution. Sure, it's not what you'd get with your thousand-dollar SLR with really great lenses - this is the digital equivalent of an Instamatic.
The interesting quality thing they did here is that they're not compressing the image much - 8MB for 24 shots means they're storing pictures as ~310KB instead of the more typical ~75KB JPEGs that other digital cameras I've seen use for 640x480 images. I don't know if this means they're doing JPEG, or if they're doing some low-CPU compression algorithm and saving money on CPU, or doing 8-bit-per-pixel uncompressed images instead of more useful color depth (unlikely but possible, and that really *would* make color suffer.)
The Necoro-nomicon ?
More seriously, I do like the example someone posted of Netscape Gold as prior art. Also, HTML is not really much more than the old troff -mm macros with angle brackets around them, and there were WYSIWYG editors for nroff/troff around in the 80s. And when did Slashdot start? Or web-based conferencing systems? Did any of them predate Bianca Troll's Graffiti Wall? The stuff just ain't new.
* Actually, there was a guy named Eli who worked in the basement who *did* write original JCL, but he was one of those rare wizards. Regular users never touched the stuff.
I primarily use RAMDISKS on Windows to accomplish the same thing - it's a convenient place to stash files I don't plan to keep, such as MSOffice attachments in incoming email messages, as well as to stash data I don't want on real disks, like decrypted emails (yes, I know they can get paged out, and there's memory persistence, but KGB/FBI/NSA attacks are really low on my threat models compared to generic theft.) It's an amazing performance win for many Windows applications, and Windows is happier running them from "disk" than running them from the temp files used by email while keeping the email messages open.
Both of those applications assume they're using main system RAM, and that they're relatively disposable - if the system crashes in the middle of a compile, you'll re-run the compile anyway. Some of the other approaches to RAM-disk boxes provide separate battery backup, which gives you persistence.
One of the classic accelerators was the old Legato Prestoserve - it had a MB or so of battery-backed RAM, which was enough to provide a non-volatile buffer for the time it took for disk drives to write. Machines have gotten a lot faster since then, and disk drives now usually come with buffers of a few MB, but that kind of approach can still be a win.
Look at the information you're seeing, and if you were old enough to be media-literate during the Gulf War, think about how the messages were managed then, including coverage on TV, news wire services, editorials, interviews with government sources. It was done better during the Gulf War because Bush Sr. could take his time, while Bush Jr. had this thrown at him, plus the press has a strong talent for going for the emotional, intense stories around the WTC scene, which creates an energy that Bush can use but can't control as easily.
Email was more useful than the web for the beginnings of the story - I first heard by phone call from a friend who'd been watching early-morning TV, and then started getting emails. CNN.com was slashdotted, and did extremely well getting anything at all up and running with that demand load - just because the web lets everybody publish information to everybody else doesn't mean you don't turn to a few centralized sites for breaking news :-) Email also had the advantage that it's much lower bit volume and scales better than the web because of the large peer-to-peer connectivity, and it has different failure modes than wired and wireless telephony so people near the affected sites could get messages out more reliably.
The net being what it is, I googled for Esther Dyson conspiracy propaganda and found a bunch of references including this interview with Esther Dyson:
Those radio-based X10 cameras are an example, as are $29 internet-cams, 802.11 wireless, and cheap bandwidth - if it weren't for anti-server policies on most cable modem companies, there'd be relatively common "Neighborhood Watch" cameras run by lots of random people. Cu-SeeMe quality fits in modem bandwidths - even 802.11 can handle dozens of broadcasts. The only place we can't easily watch is police offices.
ZKS's business model was always a gamble, but it was the critical thing they had to offer. There are a number of people in the cypherpunks community who've developed remailers, but the critical problem was how to keep enough of them running to provide security, diversity, and reliability. (The typical problem remailers have is keeping their ISPs from getting upset by complaints from recipients of unwanted anonymous mail, and running a faster-response-time system like ZKS requires comsuming more upstream bandwidth as well.) What their business model had to offer, besides a good enough friendly user interface to make it easy for the general public to use, was a financial incentive for ISPs to want to run remailers, because they're paying customers rather than problem users. We were never sure whether they'd succeed, but it would be a great thing if they did. Oh, well....
http://slashdot.org/comments.pl?sid=22261&cid=2388 370 is another thread of this same discussion - see comments there.
The technical problem is that their service uses Javascript, and doesn't work if you're not running Javascript. That means that any time you're using the system, you're vulnerable to any other JS problems on any other web page your browser encounters, until you turn JS back off. IIRC, Safeweb does attempt to clean up JS and other dangerous stuff from pages it displays to you, but it's still a risk. Also, I'm not that impressed with their Javascript, though I'm not an expert on the stuff - my problem was that under Mozilla ~0.91, they pop up windows to do the secure browsing in, and they're not really quite the shape of my screen, though that could have been Mozilla's fault. I sent email to the Safeweb folks about the fundamental "You're using Javascript" problem, and got a really prompt reply from their technical management, which was good, but they fundamentally didn't get it, which bothered me.
The other problem is trust - in general, you always need to be concerned about whether a service like this is trustable, both because of the intent of the people running it (are they ratting you out to somebody) and the security of their systems (if their server is 0wned by CrackerZ, you're not secure.) As I mentioned, Triangle Boy is really cool - it's a sort of distributed set of volunteer-run anonymizing servers, which keep moving around to prevent blocking services from blocking them, and Safeweb announced that they were going to be using this to provide censorship-free web access for people in China, the Middle East, and other places with censorship problems. The catch - they've got funding from In-Q-Tel, the CIA venture fund. It's probably entirely legit, and certainly good enough for most purposes - but how paranoid you need to be depends on who's really out to get you. ZeroKnowledge was very upfront about what their trustability levels were (plus I knew the folks there, and they were well-connected to the cypherpunks community.)
Also, there have been whole ecologies of businesses around them, like companies providing dedicated managed computers in colo spaces, and companies providing shared hosting on the managed computers, and companies paying shared-hosting companies to market more shared hosting (either through legitimate business or pyramid scams), and ASPs running applications on the dedicated managed computers used by the shared-hosting customers as well as providing services to non-web-based business out in the real world, and spam hunting businesses trying to protect you from the pyramid-spammers selling shared hosting, and content provider businesses using the dedicated hosting to serve content to shared-hosting customers or other dedicated hosting customers, and billing companies providing billing services for those Internet businesses that actually can bill somebody, and advertising services trying to get the various hosting users to carry their ads.
Disclaimer: some of the folks who run The Bunker and also Havenco on Sealand are friends of mine, and my employer's also in that business, but I'm not speaking for any of them.
"Oh Dad, Poor Dad" - Wow, Blast from the past! I was in that play in high school summer theater, a few decades ago, playing "Dad", the corpse
If you're more realistically paranoid than that, look at the number of root domain servers. There's been recent discussion about what they're doing for security and reliability, mainly worrying about crackers disrupting the databases. (Beyond, of course, the bigger problems are the relationships between ICANN, NSI, the UDRP, disgruntled postmasters, etc.
Vernor Vinge's book "A Deepness In The Sky" has a lot of discussion about what you could do with locator smart-dust; it's obviously speculative fiction, but it does a great job of looking at the potential for technology. Think about the effects of small (fictionally nanotech, but really small is probably enough) devices that communicate with their neighbors, have some computing power, and can do relative location detection. What could you do with that?
I can't find the obligatory Beowulf post :-) It's actually applicable here - one of the more interesting things to do with Smart Dust and similar locator technologies is for them to talk to each other about where they are and to detect changes in their relative positions. It's not just a server thing. Vernor Vinge's book "A Deepness In The Sky" has a lot of discussion about what you could do with locator smart-dust; it's obviously speculative fiction, but it does a great job of looking at the potential for technology.
Working in teams is one of the most critical skills that computer programmers need to learn - I hope it's taught more today than it was back when I was in school, though the institutional demand for gradability makes it hard to do. It's more common in graduate school, where research projects often need to be done as a team, but the current market has been for undergrad-producing factories to feed the dot-com boom - perhaps that'll change a bit.
Some of the important skills include how to coordinate work on shared code, how to divide up work when it is more separable, how to tell the difference, how to cooperate on design, how to find out the real requirements (since customers seldom hand you those on a plate, much less do so correctly...), how to tell when there are personality conflicts or ego problems, how to back off when you're the one with the ego problem (:-), how to match up different individuals' skills with the problems to be solved, .... Some of the new "extreme programming" approaches especially have to deal with these problems, but everyday business and amateur programming environments do as well.
Oh, and if by any chance Karl, Setsuko, Renee, or Temporary Emergency Acting Professor Eric are reading this, hi!
A few fnords to distract the form-error trap;...
93 million miles and an ozone layer seems about right.
Good reference - Brassard's Bibliography
The early QC algorithms also had a significant chance of finding a wrong answer, with no way to control what you got. On the other hand, the interesting problems that they solve are NP-hard problems like factoring, for which you can quickly verify whether an answer is correct or not.
Most car stereos have Auxiliary jacks on them, so you can plug in other sound sources. Sometimes they're installed properly; sometimes the jack isn't reachable but it's still back there if you want to look for it. Portable MP3 players range from $100 El Cheapo sets to $260 Archos jukeboxes with 6GB laptop drives in them. Plug in , Turn on, Rock out. Depending on the voltage your MP3 player uses, you might want to get a cigarette-lighter adapter to power it, or hotwire from the back of the lighter, or especially for one of the lower-capacity units, just use rechargeable batteries (or builtins, if they have them.)
(In a non-insurance-dominated free market, people could pay for what they wanted, which would probably include cost-effective non-bureaucratically-oriented structures like that. And in a socialized-medicine market, you'd probably have either lots of doctors, if you believe its proponents, or not enough money for experimental technology, if you believe its opponents, or less restriction on what the medical service can do as long as it saves the service money.)
Well, that's just the point, isn't it? Source Route bridging was designed by people who Just Didn't Get It - it's doing almost all the work of a router without gaining any of the benefits. It doesn't deserve as much bashing as Microsoft, who also Never Did Get It about networking, because the folks who did SNA and its predecessors were trying to design protocols that would run well on systems with slightly more horsepower than digital watches, back when RAM actually cost money. It's possible to use Transparent Bridging on many kinds of IBM systems, and these days any of the mainframes will run TCP/IP. And almost all the applications can either be handled using tn3270 (since there aren't a lot of Genuine Original Green-Screen 3270 terminals left - almost everyone uses PCs with various boards in them) or if necessary encapsulated in IP using DLSW or other tunnelware.
Token ring was often more reliable than Real Ethernet Thick-wire with Vampire Taps - the mechanical connectors were better than just chomping your way into a coax cable with the possible risk of trashing the code enough to get reflections. I'm not sure if it was better than connectorized thinwire or not, and it certainly wasn't better than Cat5 10baseT, at least if you used genuine Cat5 and didn't cheat with Cat3 (or didn't use cable-TV coax instead of proper thinwire :-) Back when I used to deal with this awful stuff, I had one customer for whom the original Ugly Shielded Twisted Pair token-ring really did make sense - they were a city's Department of Water and Power, and they had lots of Really Big Electric Motors which were a really bad electrically noisy environment, and their network needed all the help if could get. These days I hope they're using fiber.