Sure, Joe Schmoe might not know everything he needs to protect himself, but there are a couple of reasons he might have a hardware firewall anyway
Most WiFi boxes have firewalls in them (at least dumb NAT). It's often enabled by default and hard to turn *off*.
If Joe's got more than one computer, he'd probably got a NAT box so he can run all of them at once (unless he's doing that with Wifi, in which case see the previous item.)
NAT firewall boxes are as cheap as hubs these days, and have big scary advertising on the box about how they'll protect you from 3333V177777 h4X0rB0yz, so he bought that instead of a hub.
Most DSL and Cable Modem boxes _can_ do NAT, and maybe his ISP turns them on by default.
Maybe his wife told him to get a clue and use the firewall box because he's running Windows.
So how does OoBase compare to MySQL, Postgres, DB2, Oracle, etc.? Sure, it's as good as Access, but is it something that can support real database applications, or is it just a nice toy?
I see an article like this every year or so - usually it's blimps, some years it's solar-powered lightweight airplanes, but almost always it's a couple of years from completion. The big difference is that this project is from researchers expecting Somebody Else to deploy the system, as opposed to Entrepreneurs who expect to get Funding Real Soon Now.
But yes, it would really rock - and every couple of years delay means that the potential network speed tends to increase. Somebody else wrote that you could do TV broadcast from these things instead of satellite - that might be an actually viable business model, if you can get the spectrum allocation issues worked out.
Here in Collie-Foneya, it's been illegal for a year or two to have video screens in the front seat of a car, except for a few applications like navigation systems and rear-view-camera displays. You can have video in the back seat, to keep your kids entertained (probably a *positive* safety feature), but not the front. Using the iPod to feed a backseat display would probably be ok; using it to play music in the front is probably ok. But playing videos in the front seat is illegal.
If you've got a desktop machine as your office computer, it may or may not have a soundcard, depending on what you do, but a lot of cheap motherboards have the sound and video built in rather than using separate cards. That doesn't mean you'll have *speakers*, of course, but lots of people bring in headphones.
My main work computers for the last decade-plus have been laptops, and everything since the first or maybe second Pentium laptop generation has had a sound system and usually had tinny little speakers.
Way way too many websites have annoying sound, in addition to ads with sound. As soon as I got machines with soundcards built in, I found I needed to keep them turned off except when I'm actively trying to listen to sounds (typically music, but sometimes other things.) It's bad enough to have extra noise in my office from the computer fans, but at least that's whitish noise that isn't jumping up and down trying to get my attention. For the most part, background sounds on web sites are as annoying as leaf-blowers outside and almost as annoying as blinking ads. (There are obvious exceptions - if you're going to some band's website, it's not totally inappropriate for their music to be playing there, but that's the kind of situation where I'd turn on the sound to listen to their music clips anyway.)
Removing Extensions like Google that you dislike?
on
The Firemonger Project
·
· Score: 1
I don't want to install Google Toolbar on my browser. Is it easy to remove if you've installed this package? Extensions are typically harder to remove than themes or other simple stuff.
The current features for entering search terms in the address bar work just fine for me, and I'm concerned about the privacy implications of any deeper connection to Google or Yahoo or other search engines.
Scientology has Proven (tm) that the causes of your problems are related to the past lives of uptight dead space aliens called Thetans, and that you can cure your problems by getting your Inner Space Alien (ISA) into a better mood. But that's no contradiction to these results, because Scientology can make use of any other "real" science, and only objects to fake sciences like psychotherapy. You see, String Theory tells us that everything is made up of little tiny strings, and when your Thetans feel anxious and need to get your body to react anxiously so it's ready to fight or run away from whatever they're worried about, they start pulling on the little strings that trigger the neurotransmitters to let your body know what to do. The reason they do it this way is that Xenu can't read the chemicals remotely, though the little strings are triggered by mental energy that can be read more precisely with a very complex expensive device called an E-meter.
The big difference between Scientology's theories and Pastafarianism's is that anybody can make up stuff about the Flying Spaghetti Monster out of whole cloth, but only El-Ron and a few of his successors are Allowed to make up Official Scientology Stuff unless they want Bad Things to happen to them.
Wide-Area effects, not just local
on
Sonic Torpedo Defense
·
· Score: 3, Informative
Sound travels much farther under water, especially if you need to crank up your speakers to 11 million to blow torpedoes out of the water. It's hard to get good information without destructive testing, but very-high-level-sound activities like some of the research the Navy is doing off Monterey Canyon appear to have very serious effects on whales and dolphins over a several hundred mile wide area - echolocation and inter-pack communications don't work very well if your eardrums are blown out. I don't know if this is quite as loud, but it wouldn't be surprising if it's a potentially serious problem for marine life.
Of course, as you say, the Navy would rather avoid having lots of sailors killed also, and sunken ships are a toxic mess, but the amount of sound it takes to trash a torpedo is a lot more than the amount you get from the torpedo's explosion.
When I was in college, I lived in a big house with a bunch of other people, and we had a lot of squirrels that would run across the roof. There was one girl who lived in the top corner room one summer who got annoyed by the squirrels making a racket every morning around 5am, especially because she usually went to bed around 3-4am. (She wasn't a hacker, but literature majors often keep similar hours.) So one night about 3am she turned her stereo speakers against the wall and cranked up the bass to wake up all the squirrels. After a couple days of this, they stopped hanging around for a while.
Do not use Happy Fun Electric Fish in Bathtub or Aquarium. Do not taunt Happy Fun Electric Fish. Keep Happy Fun Electric Fish in refrigerator when not in use. Do not let Happy Fun Electric Fish locate Sarah Conner..
There are two basic ways that networks connect to each other - peering and transit. In a transit arrangement, one network (typically the big one) agrees to deliver any traffic the other network hands it, in return for a bunch of money, and it typically either advertises a default route (telling a small customer that they can send it all their packets) or a bunch of detailed routes and a default (telling a dual-homed medium-large customer how good its connections are to lots of places, but that customer might use another carrier for destinations that are closer with that carrier.) If you're an end customer, or a small ISP buying service from a big ISP, that's usually what you buy.
Peering arrangements are different. Two networks that have a lot of traffic for each other will set up direct connections, split the direct costs of the connections, and not charge for accepting packets from the other carrier. But they'll only advertise the routes for their *own* customers. If two small ISPs peer with each other, typically they're each also buying transit service from big ISPs, but it's cheaper for them to dedicate a connection or put bits on a public peering point like MAE-West than to both pay their upstream ISPs.
The biggest ISPs in the US are called "Tier 1" ISPs, and they all peer with each other rather than buying transit, though they might buy transit for international connections, if they can't get the other side to buy transit from them. It seems flaky, but it makes business sense, or at least it did for a while. In some sense, being big enough that all the other Tier 1s will peer with you is what defines Tier 1, and aside from technical issues, it's a marketing thing - "See, we're one of the big players!" Peering and Transit don't mix very well - you either connect to a given carrier by peering, or by transit, or else you spend a long time hammering out custom arrangements about exactly which routes you'll accept and tweaking routing tables.
Cogent is a Wannabe-Tier-1. Their main business model is to put fiber into big multi-tenant office buildings and sell everybody 100-meg Ethernet for about the price other carriers charge for one or two T1s. If I were a customer, I wouldn't expect there to be enough upstream to really get that much bandwidth all the time, but I'd expect to get more than a T1 all the time, and a lot more than a T1 almost all the time. Level 3 has apparently decided they're not getting enough value out of the relationship (i.e. not sending Cogent enough packets to make it worth their while) to keep peering, and wants Cogent to either pay them for service or get transit from somebody else. They gave them about 50 days to make other arrangements, but Cogent decided to play chicken with them.
Right now banks, credit card companies, Paypal, eBay, etc. have two main problems with phishers:
Lack of incentive to fix the problem, because they're not at risk
Risk that tools they can use against phishers will backfire and cost them lots of money.
The combination of the two means that most of them aren't doing much to fix the problem. If your credit card gets phished, and some merchant accepts it, usually your risk is limited to $50, but the merchant gets dinged for the loss, not the credit card issuer, at least if the merchant does this very often. If your bank info gets phished, and the thief withdraws all your money, bummer for you, but the bank doesn't lose it. If the phisher tricks you into revealing all your ID details as "identification", in addition to your account number, then the phisher can make more money selling your identity than on the specific credit card.
There's a very effective tool that most of these providers *could* use if they wanted to - creating fake account numbers that cause transactions by the phishers to get flagged. So you send them the phishing mail, they go to the phisher's website and enter the fake info, and when the phisher tries to spend it, you trace him. Of course, the bank needs to do this in a way that the *bank* isn't breaking laws against fraud etc. by doing it, and if the method becomes popular, some phishers will find ways to frame real people and flood the net with those phishes, so that the banks start losing lawsuits for busting the wrong people.
Yes, if you can create a paved-with-good-intentions worm that uses a given exploit, patches the hole, and propagates itself, there are some kinds of problems you can sometimes prevent, while risking destroying your network and infecting the people you do business with.
But anything that can do, a well-behaved cleanly-managed patch server can do much better and you don't have to
include worm propagation code in your patch system,
swamp your network with unpredictable traffic loads,
trash your users' machines at inconvenient times,
hope the worm reaches all the machines and not just most of them,
drop 200MB patches into dialup users' machines instead of waiting until they're in the office,
or do many other stupid dangerous things you haven't thought of yet.
I work for a Large Company which probably has 20,000 PCs managed by the IT department, running various versions of Windows. While the IT department are Clumsy and Evil, and any time they begin to resemble a competent organization their budget gets cut back again, they do run a number of patch server systems, most of which work much more reliably than they used to, and they run servers in most of the offices to handle printers and such. The anti-virus stuff gets queued from an internal server and Just Works, the monthly Microsoft Patch Tuesday stuff loads itself and runs, and if there are other problems that require us to install patches immediately that the central patch-tracking system can't forcefeed our machines, they'll send out an email telling everybody to run the install script.
If they didn't have a big honking network with many users working from home much of the time, they could cut their network load by downloading any software installs to local print servers, but it's usually not critical. A central server hits each user once across the WAN; a worm-based update has the possibility of sending just one copy to each office and only shoving lots of data to each machine once, but more realistic behaviour is that once something gets infected, it starts splattering all over the WAN and lots of machines in each office start splattering each other, so it's really not going to reduce WAN traffic significantly, and may crash LAN traffic.
And if you want to run a scanner-based system and don't want to hit everybody across the WAN, and don't have conveniently deployed servers everywhere, you can have a designated user in each building run the application, such as the Department Secretary or Local IT Grunt. It's much much cleaner than virusing everybody.
According to an article on NANOG, L3 gave Cogent 50 days advance warning to make other arrangements. Cogent didn't, preferring to play chicken and hope it made L3 look worse than Cogent so they'd back down. At this point, both drivers are barrelling down the road at each other, blindfolded, tossing spare steering wheels out the window, but unfortunately for Cogent, L3 is driving a bulldozer...
Peering is about two carriers deciding that they both benefit from providing each other with "free" service, splitting the direct costs of the interconnect and not charging for traffic; the alternative is that one carrier is a customer buying service from the other. (There are also some subtle technical differences, but it's basically about the money, and normally peering only routes packets between the customers of two carriers, and not between other carriers that either side also peers with.)
Cogent's business model is to sell large bandwidths for a low price, usually in multi-tenant office buildings. So they'd drop a fiber into the basement, and sell 100 Mbps ethernet connections to businesses in the building for about the price other carriers would charge for a T1 (that was back when a T1 was typically $1000 instead of $300; I haven't followed Cogent's prices in the last year or two.) Could you expect to get 100 Mbps consistently all the time? Not realistically, but you *could* expect to get lots more bandwidth than a T1 almost all the time, so it was a pretty competitive deal.
But at the end of the Interent boom, every carrier's finances looked pretty unstable, and a very aggressive business model that depends on getting free peering from big carriers while stealing their business customers looked extremely volatile:-) So does it make business sense for a Tier 1 provider to peer with Cogent as opposed to charging them money for Transit? Maybe, maybe not, and it looks like Level3 used to give them free peering but has changed their mind about it. Not the first time something like that has happened to Cogent - they've been back and forth on this with one or more carriers over the last few years. L3 seems to have decided that there's not enough reason to care about Cogent customers to give them free service.
It took us a *long* time to get the Feds out of the way. MAE-East and MAE-West really were successful, but of course they couldn't keep up with the rapid growth of the Internet, and for the US Tier 1 providers, private peering has been a necessity for a long time. The carrier I work for has about 1 Gbps of public peering and 200 Gbps of private peering, and most of the other carriers we peer with are also too big to fit on the MAEs any more. On the other hand, Equinix has largely supplanted the MAEs, and mostly provides rack space, access, and cheap fiber cross-connects rather than active switching, so much of that "private peering" is still concentrated in half a dozen buildings, with the connections happening at Layer 1 instead of Layer 3.
Other markets have different evolutionary paths and different architectures - LINX, for instance, is huge (I lost track of its size when it passed 20 Gbps; it's probably 100 Gbps by now), and is extremely effective for the UK market and as a major interconnect for EU-wide carriers, and AMSIX is also pretty big. But London really *is* the geographically hub for UK business that uses the Internet, and is a good hop-off point for EU carriers that want to cross the Atlantic, while the US market was initially dominated by two locations 5000km apart, with some major users in between or within 500km on the coasts.
Most major ISPs use BGP for external routes, with filters to control what they do and don't accept. *Internal* routing is a different story - most ISPs do most of their internal routing with OSPF or ISIS (or perhaps EIGRP), but that doesn't mean there isn't also a lot of manual hackery to "improve" things. One of the advantages of MPLS or (back in the day) ATM infrastructure under the IP is that it gives you a lot more hooks for tuning your network, but if you're not running them, then static is one way to do it, if you don't mind the large amounts of manual effort and occasional brokenness. It's a tradeoff of fragile stability vs. much more resilient instability.
Most of the Tier 1 providers sell home internet connections and small-business internet connections and a wide variety of other services as well as selling to large enterprises and other ISPs. There may be a few who specialize or don't bother dealing with small pipes, or who only support small users indirectly through wholesale/channel arrangements, but most will sell you just about anything.
Most smaller ISPs focus on niche markets, including home and/or small businesses, but that doesn't mean Tier 1 ISPs don't also serve those markets.
Disclaimer: I work for a Tier 1 ISP, and we'll sell you just about anything if you don't mind our prices, but this is my own personal opinion about the market, not official corporate policy.
One of my friends bought a DivX player, but he was a gadget-freak and it was the Internet boom, so the only real constraint for him was shelf space near the TV, plus the problem of finding worthwhile content to rent and time to watch it. Everybody thought it was a pretty dumb idea, and if I remember correctly, the DRM system got cracked after it was mostly dead anyway, so the crack was strictly another nail in the coffin as opposed to the destruction of an industry.
Netflix, by contrast, was a low-tech approach (except that DVDs were still early-adopter back then) that absolutely rocked, because it matched what most customers generally wanted to do most of the time.
It's really important in environments like that to have a web server system, so you'd want Apache or THTTPD or some alternative, and you need SAMBA because somebody's going to have a Windows box you need to talk to. And obviously you're going to want some wireless tools to help you build wireless networks - at least Netstumbler.
A USB memory stick is really useful if you want to make servers run on unknown-condition hardware, and a USB hard drive can give you more space and power if you want to haul that around. Somebody else mentioned having various USB tools so you can download from digital cameras and other random devices.
If you've got a mapping program that you can fit onto your CDs, that can be really helpful also, in case people don't have the bandwidth to get to Google/Yahoo/Mapquest. I don't know of any that run on Linux (I've got some old ones that run on Windows, and on Linux machines I just use Google), but I assume there are some out there that can read Tiger data.
Many airports are carpeted, at least in some areas, and cleaning up moving walkways is probably not that easy either, especially the rubber-tread ones. Then there's the luggage rack on the parking shuttle busses... If you've got a super-sensitive machine, and somebody wanted to overload it, there are way too many opportunities.
So a friend of a friend is a firearms enthusiast, and had a duffle bag that he used to carry a bunch of guns and ammo when he went off shooting targets or Bambi's mother or whatever. Later he used that same duffle bag as carry-on luggage when he was flying somewhere, and he got lucky and got his bag swabbed by security. The machine was Not Happy about what it found, because in fact there _was_ explosive residue on the bag, and much hand inspection occurred. From what I remember, they did let him and his get on the plane; not sure if they let him hand-carry the bag or made him check it.
Dude, they were made before you were born. LISP 1.5 came out in about 1960, a few months after COBOL, and unfortunately COBOL caught on and LISP didn't - most languages developed after it were inferior, though some were designed for special environments that didn't have the resources to run or develop LISP.
Slashdot Mirror
So how does OoBase compare to MySQL, Postgres, DB2, Oracle, etc.? Sure, it's as good as Access, but is it something that can support real database applications, or is it just a nice toy?
But yes, it would really rock - and every couple of years delay means that the potential network speed tends to increase. Somebody else wrote that you could do TV broadcast from these things instead of satellite - that might be an actually viable business model, if you can get the spectrum allocation issues worked out.
Here in Collie-Foneya, it's been illegal for a year or two to have video screens in the front seat of a car, except for a few applications like navigation systems and rear-view-camera displays. You can have video in the back seat, to keep your kids entertained (probably a *positive* safety feature), but not the front. Using the iPod to feed a backseat display would probably be ok; using it to play music in the front is probably ok. But playing videos in the front seat is illegal.
My main work computers for the last decade-plus have been laptops, and everything since the first or maybe second Pentium laptop generation has had a sound system and usually had tinny little speakers.
Way way too many websites have annoying sound, in addition to ads with sound. As soon as I got machines with soundcards built in, I found I needed to keep them turned off except when I'm actively trying to listen to sounds (typically music, but sometimes other things.) It's bad enough to have extra noise in my office from the computer fans, but at least that's whitish noise that isn't jumping up and down trying to get my attention. For the most part, background sounds on web sites are as annoying as leaf-blowers outside and almost as annoying as blinking ads. (There are obvious exceptions - if you're going to some band's website, it's not totally inappropriate for their music to be playing there, but that's the kind of situation where I'd turn on the sound to listen to their music clips anyway.)
The current features for entering search terms in the address bar work just fine for me, and I'm concerned about the privacy implications of any deeper connection to Google or Yahoo or other search engines.
The big difference between Scientology's theories and Pastafarianism's is that anybody can make up stuff about the Flying Spaghetti Monster out of whole cloth, but only El-Ron and a few of his successors are Allowed to make up Official Scientology Stuff unless they want Bad Things to happen to them.
Of course, as you say, the Navy would rather avoid having lots of sailors killed also, and sunken ships are a toxic mess, but the amount of sound it takes to trash a torpedo is a lot more than the amount you get from the torpedo's explosion.
When I was in college, I lived in a big house with a bunch of other people, and we had a lot of squirrels that would run across the roof. There was one girl who lived in the top corner room one summer who got annoyed by the squirrels making a racket every morning around 5am, especially because she usually went to bed around 3-4am. (She wasn't a hacker, but literature majors often keep similar hours.) So one night about 3am she turned her stereo speakers against the wall and cranked up the bass to wake up all the squirrels. After a couple days of this, they stopped hanging around for a while.
Do not use Happy Fun Electric Fish in Bathtub or Aquarium.
Do not taunt Happy Fun Electric Fish.
Keep Happy Fun Electric Fish in refrigerator when not in use.
Do not let Happy Fun Electric Fish locate Sarah Conner..
Peering arrangements are different. Two networks that have a lot of traffic for each other will set up direct connections, split the direct costs of the connections, and not charge for accepting packets from the other carrier. But they'll only advertise the routes for their *own* customers. If two small ISPs peer with each other, typically they're each also buying transit service from big ISPs, but it's cheaper for them to dedicate a connection or put bits on a public peering point like MAE-West than to both pay their upstream ISPs.
The biggest ISPs in the US are called "Tier 1" ISPs, and they all peer with each other rather than buying transit, though they might buy transit for international connections, if they can't get the other side to buy transit from them. It seems flaky, but it makes business sense, or at least it did for a while. In some sense, being big enough that all the other Tier 1s will peer with you is what defines Tier 1, and aside from technical issues, it's a marketing thing - "See, we're one of the big players!" Peering and Transit don't mix very well - you either connect to a given carrier by peering, or by transit, or else you spend a long time hammering out custom arrangements about exactly which routes you'll accept and tweaking routing tables.
Cogent is a Wannabe-Tier-1. Their main business model is to put fiber into big multi-tenant office buildings and sell everybody 100-meg Ethernet for about the price other carriers charge for one or two T1s. If I were a customer, I wouldn't expect there to be enough upstream to really get that much bandwidth all the time, but I'd expect to get more than a T1 all the time, and a lot more than a T1 almost all the time. Level 3 has apparently decided they're not getting enough value out of the relationship (i.e. not sending Cogent enough packets to make it worth their while) to keep peering, and wants Cogent to either pay them for service or get transit from somebody else. They gave them about 50 days to make other arrangements, but Cogent decided to play chicken with them.
- Lack of incentive to fix the problem, because they're not at risk
- Risk that tools they can use against phishers will backfire and cost them lots of money.
The combination of the two means that most of them aren't doing much to fix the problem. If your credit card gets phished, and some merchant accepts it, usually your risk is limited to $50, but the merchant gets dinged for the loss, not the credit card issuer, at least if the merchant does this very often. If your bank info gets phished, and the thief withdraws all your money, bummer for you, but the bank doesn't lose it. If the phisher tricks you into revealing all your ID details as "identification", in addition to your account number, then the phisher can make more money selling your identity than on the specific credit card.There's a very effective tool that most of these providers *could* use if they wanted to - creating fake account numbers that cause transactions by the phishers to get flagged. So you send them the phishing mail, they go to the phisher's website and enter the fake info, and when the phisher tries to spend it, you trace him. Of course, the bank needs to do this in a way that the *bank* isn't breaking laws against fraud etc. by doing it, and if the method becomes popular, some phishers will find ways to frame real people and flood the net with those phishes, so that the banks start losing lawsuits for busting the wrong people.
But anything that can do, a well-behaved cleanly-managed patch server can do much better and you don't have to
I work for a Large Company which probably has 20,000 PCs managed by the IT department, running various versions of Windows. While the IT department are Clumsy and Evil, and any time they begin to resemble a competent organization their budget gets cut back again, they do run a number of patch server systems, most of which work much more reliably than they used to, and they run servers in most of the offices to handle printers and such. The anti-virus stuff gets queued from an internal server and Just Works, the monthly Microsoft Patch Tuesday stuff loads itself and runs, and if there are other problems that require us to install patches immediately that the central patch-tracking system can't forcefeed our machines, they'll send out an email telling everybody to run the install script.
If they didn't have a big honking network with many users working from home much of the time, they could cut their network load by downloading any software installs to local print servers, but it's usually not critical. A central server hits each user once across the WAN; a worm-based update has the possibility of sending just one copy to each office and only shoving lots of data to each machine once, but more realistic behaviour is that once something gets infected, it starts splattering all over the WAN and lots of machines in each office start splattering each other, so it's really not going to reduce WAN traffic significantly, and may crash LAN traffic.
And if you want to run a scanner-based system and don't want to hit everybody across the WAN, and don't have conveniently deployed servers everywhere, you can have a designated user in each building run the application, such as the Department Secretary or Local IT Grunt. It's much much cleaner than virusing everybody.
According to an article on NANOG, L3 gave Cogent 50 days advance warning to make other arrangements. Cogent didn't, preferring to play chicken and hope it made L3 look worse than Cogent so they'd back down. At this point, both drivers are barrelling down the road at each other, blindfolded, tossing spare steering wheels out the window, but unfortunately for Cogent, L3 is driving a bulldozer...
On the other hand, the Chinese don't go in for Giant Steam-Powered Mecha Robots, so this could be, like, cool...
Cogent's business model is to sell large bandwidths for a low price, usually in multi-tenant office buildings. So they'd drop a fiber into the basement, and sell 100 Mbps ethernet connections to businesses in the building for about the price other carriers would charge for a T1 (that was back when a T1 was typically $1000 instead of $300; I haven't followed Cogent's prices in the last year or two.) Could you expect to get 100 Mbps consistently all the time? Not realistically, but you *could* expect to get lots more bandwidth than a T1 almost all the time, so it was a pretty competitive deal.
But at the end of the Interent boom, every carrier's finances looked pretty unstable, and a very aggressive business model that depends on getting free peering from big carriers while stealing their business customers looked extremely volatile :-) So does it make business sense for a Tier 1 provider to peer with Cogent as opposed to charging them money for Transit? Maybe, maybe not, and it looks like Level3 used to give them free peering but has changed their mind about it. Not the first time something like that has happened to Cogent - they've been back and forth on this with one or more carriers over the last few years. L3 seems to have decided that there's not enough reason to care about Cogent customers to give them free service.
It took us a *long* time to get the Feds out of the way. MAE-East and MAE-West really were successful, but of course they couldn't keep up with the rapid growth of the Internet, and for the US Tier 1 providers, private peering has been a necessity for a long time. The carrier I work for has about 1 Gbps of public peering and 200 Gbps of private peering, and most of the other carriers we peer with are also too big to fit on the MAEs any more. On the other hand, Equinix has largely supplanted the MAEs, and mostly provides rack space, access, and cheap fiber cross-connects rather than active switching, so much of that "private peering" is still concentrated in half a dozen buildings, with the connections happening at Layer 1 instead of Layer 3.
Other markets have different evolutionary paths and different architectures - LINX, for instance, is huge (I lost track of its size when it passed 20 Gbps; it's probably 100 Gbps by now), and is extremely effective for the UK market and as a major interconnect for EU-wide carriers, and AMSIX is also pretty big. But London really *is* the geographically hub for UK business that uses the Internet, and is a good hop-off point for EU carriers that want to cross the Atlantic, while the US market was initially dominated by two locations 5000km apart, with some major users in between or within 500km on the coasts.
Most major ISPs use BGP for external routes, with filters to control what they do and don't accept. *Internal* routing is a different story - most ISPs do most of their internal routing with OSPF or ISIS (or perhaps EIGRP), but that doesn't mean there isn't also a lot of manual hackery to "improve" things. One of the advantages of MPLS or (back in the day) ATM infrastructure under the IP is that it gives you a lot more hooks for tuning your network, but if you're not running them, then static is one way to do it, if you don't mind the large amounts of manual effort and occasional brokenness. It's a tradeoff of fragile stability vs. much more resilient instability.
Most smaller ISPs focus on niche markets, including home and/or small businesses, but that doesn't mean Tier 1 ISPs don't also serve those markets.
Disclaimer: I work for a Tier 1 ISP, and we'll sell you just about anything if you don't mind our prices, but this is my own personal opinion about the market, not official corporate policy.
Netflix, by contrast, was a low-tech approach (except that DVDs were still early-adopter back then) that absolutely rocked, because it matched what most customers generally wanted to do most of the time.
A USB memory stick is really useful if you want to make servers run on unknown-condition hardware, and a USB hard drive can give you more space and power if you want to haul that around. Somebody else mentioned having various USB tools so you can download from digital cameras and other random devices.
If you've got a mapping program that you can fit onto your CDs, that can be really helpful also, in case people don't have the bandwidth to get to Google/Yahoo/Mapquest. I don't know of any that run on Linux (I've got some old ones that run on Windows, and on Linux machines I just use Google), but I assume there are some out there that can read Tiger data.
Many airports are carpeted, at least in some areas, and cleaning up moving walkways is probably not that easy either, especially the rubber-tread ones. Then there's the luggage rack on the parking shuttle busses... If you've got a super-sensitive machine, and somebody wanted to overload it, there are way too many opportunities.
So a friend of a friend is a firearms enthusiast, and had a duffle bag that he used to carry a bunch of guns and ammo when he went off shooting targets or Bambi's mother or whatever. Later he used that same duffle bag as carry-on luggage when he was flying somewhere, and he got lucky and got his bag swabbed by security. The machine was Not Happy about what it found, because in fact there _was_ explosive residue on the bag, and much hand inspection occurred. From what I remember, they did let him and his get on the plane; not sure if they let him hand-carry the bag or made him check it.
LISP Machine Wikipedia Page