There are a couple of obstacles to widespread adoption of IPv6; some are easier to overcome, and some are seriously hard.
Scalability is seriously hard - IPv6 was supposed to fix this, mostly by giving us a clean slate to reassign addresses in some hierarchical quasi-geographical aggregation basis, but that mostly turned out to be wildly optimistic handwaving that didn't reflect the underlying needs of business customers. It's not just that businesses want to have their own portable address space so they can change ISPs without renumbering; you can bully most of them out of that if you have to.
The big problem is that business customers need multiple homing, so their address space (whether it's provider-assigned or their own portable space) is advertised by multiple ISPs, so that if one ISP connection fails they're not dead in the water. DNS isn't enough to fix this - DNS caching means that you can't instantly activate your other IP address, and ongoing sessions that have already done their DNS lookup don't pay attention to it later. There's an ugly standards-process thing called shim6 that builds a sort of session-layer into the protocols, and some people think they'll be able to get end users to adopt this, but it breaks more things than just implementing IPv6 naturally does.
IPv6's cool features keep getting adapted to IPv4 - one of the biggest reasons for IPv6 was supposed to be better security (IPSEC does this for IPv4), and also things like multicast (works ok on IPv4, but ISPs don't have the economic models to support it widely, and IPv6 doesn't provide any special economic models), QoS (several versions work fine on IPv4 and on the MPLS used by many ISPs to haul IPv4 around, but ISPs are only starting to see economic usefulness, and it tends to burn router CPU if you're not careful). Most of the mobility solutions I've seen have looked like tunnels or NAT, so IPv4 can support them. The one feature that really can't be adapted is the bigger address spaces, and NAT, RFC1918, and CIDR have let us limp along.
Chicken-and-egg problems with content - until there are a lot of users with IPv6 support in their PCs, browser software, and ISP connections, there won't be a lot of cool web content that isn't also available on IPv4 (being "a cool IPv6 demo/implementation/developer site" doesn't count as cool; it's mass consumerism and widespread business-to-business that pays the bills.) And without cool IPv6-only or IPv6-faster content, consumers and b2b end users aren't going to bother. Bram Cohen might be able to fix this by making BitTorrent do cooler things on IPv6 than on IPv4, but basically it's a slow-growth problem. Cellphone companies in Japan and China that do the walled-garden content thing are the only other model I've seen that's got any major leverage, at least until the number of home-broadband users in Asia who aren't behind national-censorship firewalls becomes a lot larger.
Router performance - High-end routers are expensive, and ISP backbones and Colo center routers need really high-end routers to keep up. IPv6 addresses are bigger, so you can't just use your IPv4 router, which implements most of the high-speed packet dumb forwarding in ASICs and does more complex functions in CPUs - either you get a router with IPv6 ASIC support, or you do most of the work in the CPU. Also, routing tables (see previous scalability rant) get a lot bigger when you use IPv6 addresss, so the routing protocol calculations need 4x as much RAM and faster CPUs. This part can be partly solved by adding Moore's Law, shaking well, and waiting 2-3 years, but of course the amount of traffic on the Internet keeps growing. Some of this is helped by moving more of the backbone onto switched environments (whether Ethernet or MPLS flavored), which lets you move more of the routing work out to the edges, but you still need the horsepower.
Redesigning mail is hard, especially if you want it done in a way that lots of people actually adopt. DJB should know - he redesigned a mail server to work around many of the problems of Sendmail - and I appreciate his frustrations. Any time a spam discussion goes on long enough, people rant about how SMTP needs to be redesigned, but almost none of the ideas I've seen actually fix the spammer problem in ways that can be adopted incrementally by users who need compatibility with the existing email infrastructure, and since you're not going to flash-cut everybody in the world to your system, you need to support incremental growth.
The fundamental problem is that spammers and users are in an arms race, and the spammers keep thinking they can Make Money Fast, so they keep coming up with new attacks on our defenses. (I know that Rule #1 says that "Spammers are Stupid", but never underestimate the power of stupid people in large groups...). For instance, moving the cost of storage to the sender doesn't fix the problem, the way it might have back in 1999, not only because spammers can now use zombie armies to store the data, but because the data doesn't really change from recipient to recipient except for some obfuscation that can be generated on the fly; spammers really don't mind if they only have to "pay" to send their latest stock scam ad banner to the 10,000 people who clicked on the message instead of the 10,000,000 people who they currently ship it to, and it lets them collect the IP addresses of probable suckers in the process. And the only people who care about storage are mailbox service providers; legitimate senders don't care because disk space is nearly free, recipients don't care, for the same reason, and spammers don't care because it's the same banner so it doesn't even take up noticeable space on the zombies they're ripping off to host it.
The protocol redesigns that *have* been adopted widely are things like "use blogs instead of mailing lists" and "use IM instead of email" and "use really large webmail services to send email to the other users of those services". Spammers have adapted by using blogspam and IM spam and captcha-recognizers.
Oh, we absolutely ought to work on getting off the planet and out of the solar system before the sun blows up or a random asteroid hits us. But we've got some slack time before that happens, and if we're only going to spend a certain amount of money and (more importantly) a certain amount of bright scientists' and engineers' lifetimes working on space issues, it's much more efficient to use them first on space programs that do earth science (so we don't screw up the planet and die out before we can do the really hard problems of significant solar system or interstellar colonization, plus it may help keep the economy in better shape so there's more money to spend on space), and on programs that pay attention to asteroids and how to detect them and keep them from hitting us (again, so we don't get wiped out before we get out of here.)
Building good space stations is another high priority - if we're going to get out of here, they're a lot more useful than putting humans on Mars, and any major colonization project will probably happen from space stations rather than launched from the ground.
Colonizing Mars or the moon, or probably more realistically asteroids, will probably happen eventually, but there's no sense in being a hurry to do it just to make politicians look good, when waiting a few decades or centuries means that Moore's Law gives us better robots (and realistically, most of the work in space will get done by robots) and we'll have time to get better materials (hopefully good enough to build space elevators.) And realistically, you're going to want cheap transport into orbit before making a trip to Mars, whether that's a space elevator or just a big rail-gun to ship raw materials into orbit to build space stations with.
The morning light is wasted on you because you're in bed asleep, and the evening light is useful to you because you've gotten out of the office before sunset. The way to take advantage of this is to drag yourself out of bed earlier, go to work earlier, and leave the office earlier.
The "Daylight Savings Time" way to do that is to reset your clock, so you can tell yourself in the morning that "it isn't really earlier, my clock says it's 8am instead of 7am, this isn't really all that hard" instead of "I hate getting up so early", and so you can tell your boss "Bye, boss, the clock said it was 9am when I got in and it's 6pm now, so I worked from 9-6" instead of "Bye, boss, I got in early and worked from 8-5, so I'm out of here".
And maybe the fact that the government told *everybody* to change their clocks means that your too-rigid-to-manage-flextime boss will let you get away with it, and your "I'm not a *&^(*&^(* morning person" attitude will let you feel better about dragging yourself out of bed earlier instead of sleeping in, but basically you've been tricked.
Farmers know this, of course - the cows are going to be up in the morning regardless of what the clock says, and since many small farmers have day jobs, DST doesn't mean they get more daylight, it just means they've got less time to do farm work before heading to their day job, unless it's a flex-time kind of job.
There are two ways to use ISDN for this. The standard way is to just have it be a very clean telephone connection carrying the vanilla telephone audio stream - G.711 8000 samples/sec 8-bit mu-law companded sampling of a 4kHz filtered audio, i.e. regular low-fi telephone audio, but no extra analog-flavored noise and hopefully a decent microphone. The other way is to use some kind of enhanced audio codec, such as one of the 7kHz 48kbps things, and use the ISDN to carry it as data; if you've got two B channels available you can get 128kbps which leaves you more room for a fancier codec, which is probably more common for music than for voice applications.
> DST made sense in the era before artificial lighting.
Huh? By "artificial lighting", I assume you mean "electricity" as opposed to "fire" (:-), though in fact there's a bit of overlap for oil lamps and gas lamps.
What makes sense is using daylight when you've got it, i.e. getting up around dawn and going to bed earlier. Farmers already do that... The reason for "Daylight Savings Time" is that having factories that all start at the same time means that workers are getting up for the day all at the same time, and Daylight Savings Time is a way for the government to tell everybody to get up earlier in the morning and go to bed earlier at night without having to be explicit about what they're doing. It makes much more sense for businesses that can use flex-time to do so, and for businesses that can't to open at some seasonally appropriate time of day, and now that commuting traffic probably burns more power than the electricity, it makes sense for businesses to stagger their hours a bit so that everybody isn't commuting at rush hour.
The Bush Administration announced a while back that NASA's priorities should be to get us back to the moon and come up with a way to get humans to Mars, and NASA's been complaining that it's interfering significantly with the budget for earth science projects - satellites and such. They only get so much money, and if they've got to put it into planning for human missions to places that should really be handled by robots at this point, then they don't have enough to do most of the other work.
This is Jenny. She and her family are having a picnic at the foot of a volcano. Oh no. The volcano has errupted. What do you do now Jenny? That's right duck and cover. What do you do Jimmy? Duck and cover. DUCK AND COVER!
Undersecretary Dudas says "and every nation is thinking how it can model [intellectual property governance] after the U.S"? That's happening not because it's a good idea but because the US Commerce Department is lobbying very aggressively to get everybody else to do what the US Intellectual Property industry wants.
He also asserts that "It's a proven system, over 200 years old.". But business model patents, which are probably the most abused and the most common violators of "obviousness", are only a few years old, especially the common "Do {some normal business practice} ON THE INTERNET" business models which cropped up after the Internet became popular in the late 1990s. Even software patents didn't exist before 1979, and for the early years they had to work by pretending to describe hardware.
Of course they want to put sentiments on the money about how "We" trust God - never mind that it's using God's name for profane purposes, it's there to give you a nice warm fuzzy feeling. The question at the time was whether you could trust the US Treasury to give you the gold or silver that the banknotes represented, or whether the coins were real gold or silver at full weight. During the War Between the States, the Confederacy printed banknotes at a vastly inflationary rate; I'm not sure how fast the US was doing the same. Later on, FDR confiscated Americans' gold, but soon after that the banknotes were telling you to trust God. When US banknotes were silver certificates, they said you should trust God, and could exchange the banknote for the silver it represented, but after a few years they stopped paying off in silver either. Now the nickel/copper/zinc coins still say it, but the Feds are complaining that it's costing them too much to make those either...
You've already spent money on PCs for everybody. At least for desktop users, LTSP gives you a way to keep using those machines for a long time, after you would have otherwise junked them. It wouldn't work in my environment - 98% of my organization uses laptops, because we work in the field and from home as well as from our offices - but if you're working from a desk, it's fine.
You'll still need servers, of course, and your servers will need upgrading, but it's a lot more concentrated and efficient.
If you break up the paragraphs where the Protestants and Jews break them up, there are four commandments dealing with Gods (Don't worship other gods, don't make statues of idols, don't use God's name in vain, take Saturdays off), and six dealing with your relationships with other people (honor your parents, don't kill, don't cheat on your spouse, don't steal, don't lie about your neighbors, don't be greedy about your neighbor's stuff.)
The Catholics like to make statues of various figures, so they split up the paragraphs so that the "graven images" bit doesn't get its own number, and coveting your neighbor's wife and you're neighbors other stuff get numbered separately. So for them it's three commandments about Gods and seven about dealing with other people.
By the way, if you look at how the Bush Administration's doing with the 10 commandments, they're not doing very well. Putting up statues of the 10 commandments at courthouses violates the graven-images bit, using religion as a justification for rabid right-wing politics violates #3, lying about weapons of mass destruction as an excuse for a war gets #6 and #9, having a war because you want to steal peoples' oil pretty much covers #8 and #10, and Republican politicians seem to get caught cheating on their wives about as much as Democrats do, even if they don't get impeached about it. But I guess they sometimes take Sunday off, except down at Guantanamo.
The real net neutrality arguments started happening when the telcos started to deploy high bandwidth to customers' homes to deploy television on them (especially when their execs started making boneheaded remarks, but there really are technical issues.)
The estimated bandwidth required for television is about 15 Mbps/house, to support a 9 Mbps High-def channel and a few low-def channel at the same time, and the various high-speed ADSL flavors mostly get about 20-50 Mbps depending on distance from your house to the green concentrator box, and there are similar bandwidth constraints to cable TV modem concentrators. The green box has fiber back to the telco office, and a typical telco office handles 10K-100K houses. Fiber-to-the-home systems have more bandwidth from the box to your house, but there's still typically around 25 Mbps per house between the box and the telco.
So if everybody's watching TV at 8pm, and they're all watching different channels, the telco office needs somewhere between 150 gigabits to 1.5 terabits per second. That's *way* more than it's getting today. After all, TV watching has much different statistics than either traditional Internet web+email content or even occasional Youtube watching - it's full bandwidth for a couple hours of primetime.
On the other hand, if the video signals are coming in as television-style content that's multicast, an OC48 2.4 Gbps feed could handle something like 200 high-def channels and 300 low-def channels. Internet-style multicast might or might not be able to handle it - as you start getting more people subscribing to content, it's going to hit the wall and choke at some point. On the other hand, if the telco or cable modem company manages it like a cable TV company selling channels, they can make sure everybody's got access to the "500 channels and nothing's on" vast wasteland of American television, and it'll work. It's not net neutrality, it's cable TV, but it works. There are hybrid models possible (e.g. the telco makes sure there's 100 channels of basic cable subscribed to the multicast feeds and the rest is first-come-first-served, with equipment enforcing the number of channels that get carried so it all fits in the telco office's available feed), but it's not clear that the telcos know how to sell that sort of thing. On the other hand, if they do too good a job of emulating the cable TV business, everybody's going to ignore them and use satellite dishes plus Youtube and Bittorrent.
The real trick with net neutrality is going to be getting the telcos to realize that they should sell you the non-TV part of the new bandwidth they're deploying as Internet bandwidth, with a pricing model different from "it's twice as big as your current bandwidth so we'll charge twice as much".
Oh, c'mon. The issue isn't whether they can take the cop to court and give him a ticket. Their objective was to catch the speeder and get him to stop. The fact that the perp was a cop, and reacted by using his position to threaten his accusers rather than to apologize, males it *really* hard to pass up a straight line like your article title:-)
Fortunately, because they were able to publicize his actions, they seem to be winning the PR war much more effectively than if he'd just done the right thing, and now he's been embarassed and presumably pressured by the town government or his bosses into backing down from his threat as well as having to drive slow for a while.
IPv6 was *supposed* to help with routing table complexity, because it was going to let us rebuild the address assignment and routing hierarchies and implement Step 3. In practice, nobody quite figured out Step 3, and the market needs for provider-independent addressing (so customers can change carriers) and customers homing to multiple ISPs (for reliability) have meant that nobody quite believes it any more. There have been some optimistic suggestions that never seem to quite cover the bases, and an ugly approach called shim6 that might help the multi-homing at the cost of real ugliness in many other places.
The internet is very much *not* hierarchical - current BGP tables are much closer to 183000 routes than 183 (it's a moving target, but somewhere in the range of 200,000 routes, which is possible to support now that you can get more than 128MB of RAM on a typical Cisco router.) In the US, there are about two dozen "Tier 1" ISPs that are mostly interconnected with each other, plus a bunch of medium, small, and specialized niche-market ISPs, plus all the hosting-center types of businesses, and the interconnections are pretty complex.
IPv6 was *supposed* to help with routing table complexity, because it was going to let us rebuild the address assignment and routing hierarchies and implement Step 3. In practice, nobody quite figured out Step 3, and the market needs for provider-independent addressing (so customers can change carriers) and customers homing to multiple ISPs (for reliability) have meant that nobody quite believes it any more. There have been some optimistic suggestions that never seem to quite cover the bases, and an ugly approach called shim6 that might help the multi-homing at the cost of real ugliness in many other places.
Replacing BGP wouldn't really help - any replacement would have to solve most of the same problems, though perhaps you could change the size of the calculations by a factor of log N if you got lucky. One of the real problems had been that a certain large router vendor used to charge way too much for RAM, about 10x the market price for painting their logo on the chips (ok, and for giving you a warrantee, but commodity memory really works fine.) But it's no longer a problem to get more than 128MB of RAM, which had been the common maximum a few years ago when the table sizes were approaching 100K entries. Of course, IPv6 makes all the tables at least 4x as large, because the addresses are bigger, but two years of Moore's Law has helped with that.
Also, your mention of transit and survivability is confused. In the US, the roughly two dozen big Tier 1 backbone ISPs are all heavily interconnected with each other, mostly in about 6-7 geographically obvious big cities, and the Tier 2, local, and niche-player ISPs almpst all get service from two or more bigger ISPs. Free peering vs. paid transit is mostly a commercial question these days - the price of transit has been in free-fall for a decade, to the extent that a number of ISPs no longer mind having to pay instead of doing free peering. The old MAEs have largely been supplanted by carrier hotels such as Equinix. In Europe, almost everybody seems to interconnect at LINX or AMSIX or both, so the commercial models are a bit different, but they've got adequately distributed hardware and buildings as well. Asia still has issues because some countries have telecomm quasi-monopolies that interfere with good architecture, but the other countries do just fine (especially Japan, Singapore, Taiwan, Hong Kong.) The Taiwan earthquake's effect was a geographical problem - just about all the north-south connectivity really does need to go by there, unless it's going to go as far east as Hawaii or else go to North America. Schools don't really affect the peering market, except for special cases like the Internet2 research networks.
After my father "retired", i.e. told his company he was retiring and wanted to start taking his pension, he worked there as a consultant full-time for a year or two before cutting back to half-time, and it took him a couple of years to _actually_ retire. But he was a research chemist, and research is the kind of thing you can do part-time.
Professional programming usually isn't part-time work, at least if you're working for a company that's producing a product to sell as opposed to doing in-house projects to support other activities. It's typically feast-or-famine schedule, with the usual deadline crunches. Now that the 90s boom is over, there may be less of the 80-hour-week-deathmarch kind of thing going on, and programmers may be more likely to have lives rather than being 25-year-olds with an infinite tolerance for caffeine, but that still tends to be the environment.
So if you want to work part-time, you'll need to look a bit longer for a gig than if you want to be full-time. On the other hand, if you want to work occasional full-time gigs, then contract/temp work does fine for that. Or if you want to do sysadmin work, that's often flexible about schedule.
Spamd's basically a lightweight MTA that you use as a proxy front-end for your real MTA.
It's still running SMTP, but instead of running a full-scale turing-machine-complete bells-and-whistles mail forwarding and delivery extravaganza like Sendmail, or the somewhat lighter postfix / qmail / etc., all of which have to manage complex delivery rules, spam filters, and mail relay capabilities, spamd is basically configured to do some simple spam-repeller rules, relay mail that's potentially not spam to your real MTA, and optionally harass anyone that looks spammy. This makes 50-90% of the email go away, letting your memory-hogging CPU intensive spam filters and mail delivery engine handle the more clever spam and the occasional real message.
It doesn't take that much resource to do basic SMTP and greylisting. After all, Sendmail originally ran on a PDP-11, as did UUCP, and it's mostly a simple state machine that does a couple of handshake plus keeping a simple database lookup - and the database doesn't need to be an SQL engine, just a simple hash table or Berkeley DB or equivalent, keeping track of an IP address and a timestamp, plus another list or two of valid email usernames, known evil sites, etc.
Spamd has another big advantage, which is that cleverness of spam content and cleverness of spam delivery method aren't closely correlated. The spam that puts lots of work into embedding images in the message body or being the most uniquely plaintive Nigerian widow may be harder for your CPU-burning Bayesian filters to crunch, but it may be delivered from known zombies or high-performance delivery engines that don't slow down to respond to return codes, so you can save a lot of transferred bits and CPU just by greylisting or by running slowly for the first few seconds. And some of the really clever spam/phishing writers are using stolen address space, which is a really clever delivery system that doesn't survive greylisting either.
Spamming makes money for some people, because "there's a sucker born every minute, and two to take him", and spammers are happy to sell to the sucker, the two that want to take him, or both. It doesn't matter that some spammers are wanabees who lose money; if they fall off the map, there are more replacements on the way. If spamming didn't work, there wouldn't be so much of it, and it wouldn't be increasing in volume so fast.
We don't care why spammers think they can find suckers - we only care how, because that offers some hints for how to detect them and either kill them or distract them into areas where they spam each other and leave us alone, optionally while we sell them bandwidth or imaginary hosting space in Nigeria or whatever. If we can make it technically infeasible or economically non-profitable to spam *us*, they'll stop spamming us, but otherwise they'll keep it up.
Occam's Razor is fundamentally an aesthetic preference, not a physical law.
As we learn more and more physics and biology, the world is getting to be a weirder and weirder place. Atoms and other particles we can interact with seem to be made out of quarks, some of which you can only see by bashing large particles together, quarks might or might not be made out of little strings of 10-or-11-dimensional nothingness, quantum effects mean that all the stuff isn't necessarily where or what you think it is, vaccuum is no longer empty space but now has pairs of particles appearing and disappearing for no apparent reason, large parts of the universe are probably "dark matter" we haven't found yet, optionally pushed around by "dark energy", and there might or might not be infinite numbers of universes reflecting those quantum effects, and/or there might be lots of other universes that have different physical laws, some of which make them vanish faster or slower than ours after bigger or smaller bangs, and which mostly don't support the kind of physics needed for us to exist. Meanwhile your brain may look like a piece of meat running a closed-source operating system in a wash of chemicals, what you perceive as "yourself" is looking like a pretty complex illusion, and if you look at all the complexity of evolution and the weird stuff it's come up with, it's surprising you're around to perceive it at all.
Basically, it looks much simpler to decide that a god made the place, whether it's one of the traditional ones or the FSM, though clearly any god must have had enough sense of humor to make all those particle-wave things, humans, cats to laugh at the humans, and inordinate types of beetles. And Maxwell's daemons are looking just as likely as some of the ways that quantum mechanics play dice with the universe. On the other hand, that doesn't give you much in the way of tools to make predictions, and as you say, making predictions is one of the big things that science is for.
You're already going to have some constant DNS load from spammers who aren't using their own ISP's DNS server; this won't increase it much, because you're still using static configs. Most ISPs have caching DNS servers, and most zombies and other virus-driven spammers are going to be using their ISP's DNS servers, not targeting any special ones.
Are there legitimate mail servers that consistently fail greylisting? Or is it a problem with how they're configured (e.g. some MCSE sysadmin doesn't understand all the options to Exchange 2007?)
False positives are more annoying in a corporate environment, but usually if a mailserver is at least halfway competently misconfigured then the user will get to see your well-written message saying "Sorry if my spam-blocker confused your mail server, call my postmaster at 1-415-555-1212 or www.example.com/postmaster" and you can whitelist them when they call to complain. (Because yes, it's realistically much more likely that you'll want to whitelist a potential customer than bother getting them to fix their email server.) And you're going to want to whitelist frequent contacts anyway, just to avoid slowing down their mail.
False negatives are ok. If greylisting only cuts your spam load by 50% instead of 90% or 99%, it's still cutting the CPU load on your better spam filters in half, and letting you be more careful about filtering the real email out of the flood of better-implemented spam.
Also, you can get fancy about greylisting selectively if you want - take all those hyper-aggressive take-no-prisoners admit-no-mistakes RBLs, and the Linux-user-hating dynamic-address DUL blocklists, and greylist that stuff even if you're not going to greylist the rest of the internet. You'll be blocking most of Zombieland, and very little corporate email that way, and it'll still cut your spam load. If you know there are countries where you don't do business, e.g. Korea, China, and Nigeria, you can put them on your greylist targets as well.
Slashdot Mirror
Her Stanford talk on how she learned to do VLSI design herself was cool too.
The big problem is that business customers need multiple homing, so their address space (whether it's provider-assigned or their own portable space) is advertised by multiple ISPs, so that if one ISP connection fails they're not dead in the water. DNS isn't enough to fix this - DNS caching means that you can't instantly activate your other IP address, and ongoing sessions that have already done their DNS lookup don't pay attention to it later. There's an ugly standards-process thing called shim6 that builds a sort of session-layer into the protocols, and some people think they'll be able to get end users to adopt this, but it breaks more things than just implementing IPv6 naturally does.
The fundamental problem is that spammers and users are in an arms race, and the spammers keep thinking they can Make Money Fast, so they keep coming up with new attacks on our defenses. (I know that Rule #1 says that "Spammers are Stupid", but never underestimate the power of stupid people in large groups...). For instance, moving the cost of storage to the sender doesn't fix the problem, the way it might have back in 1999, not only because spammers can now use zombie armies to store the data, but because the data doesn't really change from recipient to recipient except for some obfuscation that can be generated on the fly; spammers really don't mind if they only have to "pay" to send their latest stock scam ad banner to the 10,000 people who clicked on the message instead of the 10,000,000 people who they currently ship it to, and it lets them collect the IP addresses of probable suckers in the process. And the only people who care about storage are mailbox service providers; legitimate senders don't care because disk space is nearly free, recipients don't care, for the same reason, and spammers don't care because it's the same banner so it doesn't even take up noticeable space on the zombies they're ripping off to host it.
The protocol redesigns that *have* been adopted widely are things like "use blogs instead of mailing lists" and "use IM instead of email" and "use really large webmail services to send email to the other users of those services". Spammers have adapted by using blogspam and IM spam and captcha-recognizers.
Building good space stations is another high priority - if we're going to get out of here, they're a lot more useful than putting humans on Mars, and any major colonization project will probably happen from space stations rather than launched from the ground.
Colonizing Mars or the moon, or probably more realistically asteroids, will probably happen eventually, but there's no sense in being a hurry to do it just to make politicians look good, when waiting a few decades or centuries means that Moore's Law gives us better robots (and realistically, most of the work in space will get done by robots) and we'll have time to get better materials (hopefully good enough to build space elevators.) And realistically, you're going to want cheap transport into orbit before making a trip to Mars, whether that's a space elevator or just a big rail-gun to ship raw materials into orbit to build space stations with.
The "Daylight Savings Time" way to do that is to reset your clock, so you can tell yourself in the morning that "it isn't really earlier, my clock says it's 8am instead of 7am, this isn't really all that hard" instead of "I hate getting up so early", and so you can tell your boss "Bye, boss, the clock said it was 9am when I got in and it's 6pm now, so I worked from 9-6" instead of "Bye, boss, I got in early and worked from 8-5, so I'm out of here".
And maybe the fact that the government told *everybody* to change their clocks means that your too-rigid-to-manage-flextime boss will let you get away with it, and your "I'm not a *&^(*&^(* morning person" attitude will let you feel better about dragging yourself out of bed earlier instead of sleeping in, but basically you've been tricked.
Farmers know this, of course - the cows are going to be up in the morning regardless of what the clock says, and since many small farmers have day jobs, DST doesn't mean they get more daylight, it just means they've got less time to do farm work before heading to their day job, unless it's a flex-time kind of job.
There are two ways to use ISDN for this. The standard way is to just have it be a very clean telephone connection carrying the vanilla telephone audio stream - G.711 8000 samples/sec 8-bit mu-law companded sampling of a 4kHz filtered audio, i.e. regular low-fi telephone audio, but no extra analog-flavored noise and hopefully a decent microphone. The other way is to use some kind of enhanced audio codec, such as one of the 7kHz 48kbps things, and use the ISDN to carry it as data; if you've got two B channels available you can get 128kbps which leaves you more room for a fancier codec, which is probably more common for music than for voice applications.
Huh? By "artificial lighting", I assume you mean "electricity" as opposed to "fire" (:-), though in fact there's a bit of overlap for oil lamps and gas lamps.
What makes sense is using daylight when you've got it, i.e. getting up around dawn and going to bed earlier. Farmers already do that... The reason for "Daylight Savings Time" is that having factories that all start at the same time means that workers are getting up for the day all at the same time, and Daylight Savings Time is a way for the government to tell everybody to get up earlier in the morning and go to bed earlier at night without having to be explicit about what they're doing. It makes much more sense for businesses that can use flex-time to do so, and for businesses that can't to open at some seasonally appropriate time of day, and now that commuting traffic probably burns more power than the electricity, it makes sense for businesses to stagger their hours a bit so that everybody isn't commuting at rush hour.
The Bush Administration announced a while back that NASA's priorities should be to get us back to the moon and come up with a way to get humans to Mars, and NASA's been complaining that it's interfering significantly with the budget for earth science projects - satellites and such. They only get so much money, and if they've got to put it into planning for human missions to places that should really be handled by robots at this point, then they don't have enough to do most of the other work.
This is Jenny. She and her family are having a picnic at the foot of a volcano. Oh no. The volcano has errupted. What do you do now Jenny? That's right duck and cover. What do you do Jimmy? Duck and cover. DUCK AND COVER!
He also asserts that "It's a proven system, over 200 years old.". But business model patents, which are probably the most abused and the most common violators of "obviousness", are only a few years old, especially the common "Do {some normal business practice} ON THE INTERNET" business models which cropped up after the Internet became popular in the late 1990s. Even software patents didn't exist before 1979, and for the early years they had to work by pretending to describe hardware.
Of course they want to put sentiments on the money about how "We" trust God - never mind that it's using God's name for profane purposes, it's there to give you a nice warm fuzzy feeling. The question at the time was whether you could trust the US Treasury to give you the gold or silver that the banknotes represented, or whether the coins were real gold or silver at full weight. During the War Between the States, the Confederacy printed banknotes at a vastly inflationary rate; I'm not sure how fast the US was doing the same. Later on, FDR confiscated Americans' gold, but soon after that the banknotes were telling you to trust God. When US banknotes were silver certificates, they said you should trust God, and could exchange the banknote for the silver it represented, but after a few years they stopped paying off in silver either. Now the nickel/copper/zinc coins still say it, but the Feds are complaining that it's costing them too much to make those either...
You'll still need servers, of course, and your servers will need upgrading, but it's a lot more concentrated and efficient.
The Catholics like to make statues of various figures, so they split up the paragraphs so that the "graven images" bit doesn't get its own number, and coveting your neighbor's wife and you're neighbors other stuff get numbered separately. So for them it's three commandments about Gods and seven about dealing with other people.
By the way, if you look at how the Bush Administration's doing with the 10 commandments, they're not doing very well. Putting up statues of the 10 commandments at courthouses violates the graven-images bit, using religion as a justification for rabid right-wing politics violates #3, lying about weapons of mass destruction as an excuse for a war gets #6 and #9, having a war because you want to steal peoples' oil pretty much covers #8 and #10, and Republican politicians seem to get caught cheating on their wives about as much as Democrats do, even if they don't get impeached about it. But I guess they sometimes take Sunday off, except down at Guantanamo.
The estimated bandwidth required for television is about 15 Mbps/house, to support a 9 Mbps High-def channel and a few low-def channel at the same time, and the various high-speed ADSL flavors mostly get about 20-50 Mbps depending on distance from your house to the green concentrator box, and there are similar bandwidth constraints to cable TV modem concentrators. The green box has fiber back to the telco office, and a typical telco office handles 10K-100K houses. Fiber-to-the-home systems have more bandwidth from the box to your house, but there's still typically around 25 Mbps per house between the box and the telco.
So if everybody's watching TV at 8pm, and they're all watching different channels, the telco office needs somewhere between 150 gigabits to 1.5 terabits per second. That's *way* more than it's getting today. After all, TV watching has much different statistics than either traditional Internet web+email content or even occasional Youtube watching - it's full bandwidth for a couple hours of primetime.
On the other hand, if the video signals are coming in as television-style content that's multicast, an OC48 2.4 Gbps feed could handle something like 200 high-def channels and 300 low-def channels. Internet-style multicast might or might not be able to handle it - as you start getting more people subscribing to content, it's going to hit the wall and choke at some point. On the other hand, if the telco or cable modem company manages it like a cable TV company selling channels, they can make sure everybody's got access to the "500 channels and nothing's on" vast wasteland of American television, and it'll work. It's not net neutrality, it's cable TV, but it works. There are hybrid models possible (e.g. the telco makes sure there's 100 channels of basic cable subscribed to the multicast feeds and the rest is first-come-first-served, with equipment enforcing the number of channels that get carried so it all fits in the telco office's available feed), but it's not clear that the telcos know how to sell that sort of thing. On the other hand, if they do too good a job of emulating the cable TV business, everybody's going to ignore them and use satellite dishes plus Youtube and Bittorrent.
The real trick with net neutrality is going to be getting the telcos to realize that they should sell you the non-TV part of the new bandwidth they're deploying as Internet bandwidth, with a pricing model different from "it's twice as big as your current bandwidth so we'll charge twice as much".
Fortunately, because they were able to publicize his actions, they seem to be winning the PR war much more effectively than if he'd just done the right thing, and now he's been embarassed and presumably pressured by the town government or his bosses into backing down from his threat as well as having to drive slow for a while.
IPv6 was *supposed* to help with routing table complexity, because it was going to let us rebuild the address assignment and routing hierarchies and implement Step 3. In practice, nobody quite figured out Step 3, and the market needs for provider-independent addressing (so customers can change carriers) and customers homing to multiple ISPs (for reliability) have meant that nobody quite believes it any more. There have been some optimistic suggestions that never seem to quite cover the bases, and an ugly approach called shim6 that might help the multi-homing at the cost of real ugliness in many other places.
The internet is very much *not* hierarchical - current BGP tables are much closer to 183000 routes than 183 (it's a moving target, but somewhere in the range of 200,000 routes, which is possible to support now that you can get more than 128MB of RAM on a typical Cisco router.) In the US, there are about two dozen "Tier 1" ISPs that are mostly interconnected with each other, plus a bunch of medium, small, and specialized niche-market ISPs, plus all the hosting-center types of businesses, and the interconnections are pretty complex.
IPv6 was *supposed* to help with routing table complexity, because it was going to let us rebuild the address assignment and routing hierarchies and implement Step 3. In practice, nobody quite figured out Step 3, and the market needs for provider-independent addressing (so customers can change carriers) and customers homing to multiple ISPs (for reliability) have meant that nobody quite believes it any more. There have been some optimistic suggestions that never seem to quite cover the bases, and an ugly approach called shim6 that might help the multi-homing at the cost of real ugliness in many other places.
Replacing BGP wouldn't really help - any replacement would have to solve most of the same problems, though perhaps you could change the size of the calculations by a factor of log N if you got lucky. One of the real problems had been that a certain large router vendor used to charge way too much for RAM, about 10x the market price for painting their logo on the chips (ok, and for giving you a warrantee, but commodity memory really works fine.) But it's no longer a problem to get more than 128MB of RAM, which had been the common maximum a few years ago when the table sizes were approaching 100K entries. Of course, IPv6 makes all the tables at least 4x as large, because the addresses are bigger, but two years of Moore's Law has helped with that.
Also, your mention of transit and survivability is confused. In the US, the roughly two dozen big Tier 1 backbone ISPs are all heavily interconnected with each other, mostly in about 6-7 geographically obvious big cities, and the Tier 2, local, and niche-player ISPs almpst all get service from two or more bigger ISPs. Free peering vs. paid transit is mostly a commercial question these days - the price of transit has been in free-fall for a decade, to the extent that a number of ISPs no longer mind having to pay instead of doing free peering. The old MAEs have largely been supplanted by carrier hotels such as Equinix. In Europe, almost everybody seems to interconnect at LINX or AMSIX or both, so the commercial models are a bit different, but they've got adequately distributed hardware and buildings as well. Asia still has issues because some countries have telecomm quasi-monopolies that interfere with good architecture, but the other countries do just fine (especially Japan, Singapore, Taiwan, Hong Kong.) The Taiwan earthquake's effect was a geographical problem - just about all the north-south connectivity really does need to go by there, unless it's going to go as far east as Hawaii or else go to North America. Schools don't really affect the peering market, except for special cases like the Internet2 research networks.
Professional programming usually isn't part-time work, at least if you're working for a company that's producing a product to sell as opposed to doing in-house projects to support other activities. It's typically feast-or-famine schedule, with the usual deadline crunches. Now that the 90s boom is over, there may be less of the 80-hour-week-deathmarch kind of thing going on, and programmers may be more likely to have lives rather than being 25-year-olds with an infinite tolerance for caffeine, but that still tends to be the environment.
So if you want to work part-time, you'll need to look a bit longer for a gig than if you want to be full-time. On the other hand, if you want to work occasional full-time gigs, then contract/temp work does fine for that. Or if you want to do sysadmin work, that's often flexible about schedule.
Those weren't the droids you were looking for, you can move along.
It's still running SMTP, but instead of running a full-scale turing-machine-complete bells-and-whistles mail forwarding and delivery extravaganza like Sendmail, or the somewhat lighter postfix / qmail / etc., all of which have to manage complex delivery rules, spam filters, and mail relay capabilities, spamd is basically configured to do some simple spam-repeller rules, relay mail that's potentially not spam to your real MTA, and optionally harass anyone that looks spammy. This makes 50-90% of the email go away, letting your memory-hogging CPU intensive spam filters and mail delivery engine handle the more clever spam and the occasional real message.
It doesn't take that much resource to do basic SMTP and greylisting. After all, Sendmail originally ran on a PDP-11, as did UUCP, and it's mostly a simple state machine that does a couple of handshake plus keeping a simple database lookup - and the database doesn't need to be an SQL engine, just a simple hash table or Berkeley DB or equivalent, keeping track of an IP address and a timestamp, plus another list or two of valid email usernames, known evil sites, etc.
Spamd has another big advantage, which is that cleverness of spam content and cleverness of spam delivery method aren't closely correlated. The spam that puts lots of work into embedding images in the message body or being the most uniquely plaintive Nigerian widow may be harder for your CPU-burning Bayesian filters to crunch, but it may be delivered from known zombies or high-performance delivery engines that don't slow down to respond to return codes, so you can save a lot of transferred bits and CPU just by greylisting or by running slowly for the first few seconds. And some of the really clever spam/phishing writers are using stolen address space, which is a really clever delivery system that doesn't survive greylisting either.
Spamming makes money for some people, because "there's a sucker born every minute, and two to take him", and spammers are happy to sell to the sucker, the two that want to take him, or both. It doesn't matter that some spammers are wanabees who lose money; if they fall off the map, there are more replacements on the way. If spamming didn't work, there wouldn't be so much of it, and it wouldn't be increasing in volume so fast.
We don't care why spammers think they can find suckers - we only care how, because that offers some hints for how to detect them and either kill them or distract them into areas where they spam each other and leave us alone, optionally while we sell them bandwidth or imaginary hosting space in Nigeria or whatever. If we can make it technically infeasible or economically non-profitable to spam *us*, they'll stop spamming us, but otherwise they'll keep it up.
As we learn more and more physics and biology, the world is getting to be a weirder and weirder place. Atoms and other particles we can interact with seem to be made out of quarks, some of which you can only see by bashing large particles together, quarks might or might not be made out of little strings of 10-or-11-dimensional nothingness, quantum effects mean that all the stuff isn't necessarily where or what you think it is, vaccuum is no longer empty space but now has pairs of particles appearing and disappearing for no apparent reason, large parts of the universe are probably "dark matter" we haven't found yet, optionally pushed around by "dark energy", and there might or might not be infinite numbers of universes reflecting those quantum effects, and/or there might be lots of other universes that have different physical laws, some of which make them vanish faster or slower than ours after bigger or smaller bangs, and which mostly don't support the kind of physics needed for us to exist. Meanwhile your brain may look like a piece of meat running a closed-source operating system in a wash of chemicals, what you perceive as "yourself" is looking like a pretty complex illusion, and if you look at all the complexity of evolution and the weird stuff it's come up with, it's surprising you're around to perceive it at all.
Basically, it looks much simpler to decide that a god made the place, whether it's one of the traditional ones or the FSM, though clearly any god must have had enough sense of humor to make all those particle-wave things, humans, cats to laugh at the humans, and inordinate types of beetles. And Maxwell's daemons are looking just as likely as some of the ways that quantum mechanics play dice with the universe. On the other hand, that doesn't give you much in the way of tools to make predictions, and as you say, making predictions is one of the big things that science is for.
You're already going to have some constant DNS load from spammers who aren't using their own ISP's DNS server; this won't increase it much, because you're still using static configs. Most ISPs have caching DNS servers, and most zombies and other virus-driven spammers are going to be using their ISP's DNS servers, not targeting any special ones.
False positives are more annoying in a corporate environment, but usually if a mailserver is at least halfway competently misconfigured then the user will get to see your well-written message saying "Sorry if my spam-blocker confused your mail server, call my postmaster at 1-415-555-1212 or www.example.com/postmaster" and you can whitelist them when they call to complain. (Because yes, it's realistically much more likely that you'll want to whitelist a potential customer than bother getting them to fix their email server.) And you're going to want to whitelist frequent contacts anyway, just to avoid slowing down their mail.
False negatives are ok. If greylisting only cuts your spam load by 50% instead of 90% or 99%, it's still cutting the CPU load on your better spam filters in half, and letting you be more careful about filtering the real email out of the flood of better-implemented spam.
Also, you can get fancy about greylisting selectively if you want - take all those hyper-aggressive take-no-prisoners admit-no-mistakes RBLs, and the Linux-user-hating dynamic-address DUL blocklists, and greylist that stuff even if you're not going to greylist the rest of the internet. You'll be blocking most of Zombieland, and very little corporate email that way, and it'll still cut your spam load. If you know there are countries where you don't do business, e.g. Korea, China, and Nigeria, you can put them on your greylist targets as well.
Dude, at least get the joke right :-)