I could make the point that it's easier to keep a few atomic weapons then build new ones, and inspections wouldn't find them. But that doesn't leave room for debate:)
I can't say if atomic weapons can be build in secret. My gut feeling tells me they can. Think about China, or US, or even Japan. All are big enough so that a large well funded, distributed project could go undetected. Not every country is Irak.
Anyways, I wanted to say something else. Someone before us mentiones something about how conflicts grew in number of casualties all over our history, until the end of WW2, the likely reason beeing big powers avoided was for fear of a nuclear conflict. Now, imagine trying to enforce a no-nuclear pact against China. 1 bilion people, a quarter of a _BILION_ potential army. If they wanted to build anything, they could. Trying to stop them, succesful or not, would be the most bloody affair since WW2. Same with US, Russia, India, Japan, etc.
How about if Germany made a few nukes? US would try to invade, but who do you think the rest of Europe would side with? WW3?
This "disarmament with teeth" sounds ok on paper, but it's nor enforceable except against Irak and (we'll see) Iran.
The US gov't signed and Congress approved a treaty whose goal is to rid the world of nuclear weapons. They, and many others, obviously disagree. I don't know the text or the context, but i'll go on a limb and say they didn't mean to get rid of _all_ nukes. The big problem with the cold war was not the mere danger of using nuclear weapons: we used them before and with arguably good results. The real problem was that the number and the power of existing weapons could destroy all life on Earth a few times over. That is what all those treaties are trying to fix. Plus, it's too expensive to keep 20.000 nukes when you can keep 1000 and sleep just as well.
Nuclear inspections work. They worked to enforce the US and USSR's treaties, and they've worked to prevent other countries from building nukes (e.g. Iraq, Iran). If i were you I wouldn't mention Irak, considering it was invaded while the head of UN inspectors was saying on CNN they didn't find anything. This is a big example of how the inspections didn't work, not in the sense they didn't find what they were supposed, but that they weren't trusted enough to prevent war - which is ultimately their job.
And finaly: This seems to imply that the world is safer with nuclear weapons that without. Hell yes, i meant it. I'm pretty sure Eastern Europe where I live would have seen war the past 50 years if it weren't for the cold war. Which means my parent may have already died, and I wouldn't write this. So yes, I feel quite safe with nuclear weapons.
I wonder, where do you live? US maybe? That would explain a lot. I just thought a US citisen is really threatened only by a nuclear attak. Right now (and for the past 50 years) they were quite safe from anything else. The rest of us, on the other hand, know that war on home ground is a possibility. Also, without nuclear weapons there would be no balance - US would clearly be the best military power.
And how exactly would eliminating _all_ nuclear weapons help things? What really keeps most nations from building and using nuclear weapons is neither that UN disagree (hah) nor that it's too hard technically, It's the fact that most of their territory would turn to glass soon after. If now every nation would get rid of nuclear weapons do you know what would happen? They'd all stard building them right back and the first to finish would win. Only they'd probably have to use them to make the point.
Ok, ok, I wasn't talking about 2K or NT. From a user standpoint 2K is more or less the same with XP, they came out what? a year apart? and are way more similar then different. NT is The Grandpa. It deserves respect, but it was always a server OS first.
Thank you;) Truth is, I never got "under the hood" with windows. I've kept strictly portable for many years now. Usually I'm hapy about it, but since C# I'm started to envy you guys a bit.
Heh. I actually forgot about 2K:) Yes, the difference isn't so big, but then they're pretty much contemporary, and almost the same product. At least compared to '98, vista or millenium, there is little real difference between them. More like different packaging.
Don't understand the -1 mod. It's a valid post, with a legitimate (and argumented) point of view. And the sad part is probably the ones modding it down are doing it from an XP machine.
I happen to agree with the parent, any OS before Windows XP was way too inadequate. Actually, from a business point of view, the biggest flaw of XP is it's way too good. It took Microsoft 6 years to come with a successor (Vista), and it's gonna be a very tough sell - convincing people they don't need their XP anymore isn't a piece of cake. I for one before XP was using mostly linux for desktop. Not because I was a big fan (although I was) but because it was a better choice compared to '98 & company. Not so much now, because the need isn't the same.
Wikipedia's Neutral Point of View ideology may help here. Sounds like wishful thinking but they managed to make it work. Helps a lot they translated a vague concept (NPV) into a clear and short set of guidelines and made people respect it.
Expelling a student for cheating when statistics (and common sense) tell 50% of students do the same, yes, it is unethical. Basicaly you're expelling him because his number came up. What i'm talking about is a realistic policy to prevent cheating. You can't expell half the students (don't even go there...), it's well, kinda not nice to expell by lottery just to make examples (which doesn't even work). So my 2 cents would be to find alternative punishments, something that can be more often and easely apply and which would "touch" the student without ruining his career. Something like dressing in yellow for a month.
Now if you want a more in-depth analysis (yawn... it's late!) I'll say things aren't likely to change very soon. Why? Because cheating is too useful for both sides. Its best use is to cover the cracks in the educational system. Classes are boring? The teacher is unqualified? The subject is outdated? No worry, some will cram, some will barely get by and all will cheat more or less. And they'll pass. Imagine how embarrasing would be for a teacher to have 90% of his class fail. Not really looking good on his resume, right? Did you notice how teachers are more lenient with cheating when they're doing a bad work with the class? And the strictest teachers usually do their job better then average? For the students it's most useful in dodging the less interesting subjects, or in getting by when that's all you want, or in avoiding a bad teacher etc. For the teachers it's also another power tool. When so many cheat, all you have to do is point your finger and someone is expelled. And make no mistake, a teacher half awake knows when students cheat.
For cheating to go away for good the system would have to work for real. And it's not gonna happen any time soon.
I never get why people always go to extremes in these cases. As if they're taboo and people are ashamed of it so they either pretend not to notice or overreact. To expel a student over cheating when statistics say 50% do it is... well, not very ethical. Not to do anything is just as wrong, of course. But I've yet to see a university having a tough but reasonable policy on cheating and implementing it.
Just use your imagination (and natural sadism if you have it) and make up 5-6 punishments for cheating that don't involve the student getting expelled. Let's see... - failing the exam (always), and beeing examined at the end of the year by a comision - paying a fine. Yes, money. - having to take extra/advanced classes related to said subject - take ethics classes (if the university has them) - lots and lots of extra papers to make the next semester - (my favorite) assist the offended teacher for the next semester - "volunteer" work for the university (washing the dishes?)
Seven already? I was just started to think of what you can do with a free workforce around the campus.
It's a punishment, so it doesn't have to pretend it isn't. And given the alternative is expulsion, the students will definitely prefer it. Also, it's a much better deterrent. When somebody is expelled, nobody sees him anymore, so he'll be soon forgotten. But if you see the guy each morning washing the floors... you tend to remember why he's doing it.
It's a matter of relationship. Last time (a long ago) when i did pair programming it was lots of fun and very useful. The one writing code took care of the micromanagement and had in mind at most the function he was working on, and the other was two steps ahead. If the side man keeps checking syntax and formating then the whole thing is useless.
It's a bit of a cultural thing too. For you, cams at every street corner are the pinnacle of Big Brother. In other countries they're just means to an end. And it's not just a matter of opinion, it may be that this system wouldn't fit well in America. But the growing feeling in the internet nation is that it's unnatural for a piece of civilisation not to be watched. Either by satelite, tourist's cameras or CCTVs. Now since that's obviously where we're heading (and I don't even disagree) we'd better start getting ready.
One very important point is who gets to watch these cameras. And who gets to watch _all_ these cameras. Two distinct questions, mind you. You can either go with very few people, or almost everybody. For the first question, I tend to go with the second approach, because on the long run it's impossible to restrict access and I wouldn't want to find out the mob can watch but i can't. As for who can search the data feed from all those cameras... that would better be restricted. Can't imagine a good thing average joe would get from this, but i wouldn't let the local bag guys look up my daily routine.
In my side of the world most (90%) of the grading is done by exams. Very little is graded based on papers and such, at all levels of education. And let me tell you, it sucks. We watch your system with more then a little envy. Why? Because exams don't measure _at all_ creativity or initiative. Only how well you learned _the subject_, and only the subject, and how smart you are. And btw, you can cheat too. This may be a cultural thing, but cheating in exams is quite tolerated by the faculty. Which I used to hate because allowing students to cheat prevents noticing and solving lots of problems with the education in general. "Look, they're passing, so we must be doing our job right".
Funny that kids today are pretty good at video games that are longer and more complicated than the ones we played, eh?
Are you kidding?! Last time i watched someone play a contemporary game for half hour i was "WTF?! It's like click, point, click, point, click, point, click, point, click, point, click, point, click, point". Boring as hell. I remember about a week after discovering Dune 2 I was eating with my parents and suddenly i let the fork fall into the plate (splashing stew around) and exclaimed: "U! If i shape it [the defence] like an U, they'll all come right in!". My parents were a bit put of:) Warcraft 1 I couldn't finish because of the "puzzles", but once i got the rules for 2 (defend and harvest i think) i won easely.
I agree btw, when you say: "It would have been hard *not* to program". Now programming is designed to be done by the few, and it feels. I don't think any amount of work can make that go away. A "programmingforbeginners.org" would help a lot, though...
See? You're reading it from a developer point of view (90% of atomic functions offered are identical), and i meant it from a user point of view (does pretty much the same thing). That's exactly the problem, the incapacity to see things with a purely pragmatic eye: the user's. I suffered a lot because of it, mostly by implementic lots of cool features to my software only to find the clients extatic at the most basic functions. Now i'm asking twice, and i've learned that most times i can actualy get away with not doing the ugliest parts, simply because "hard" doesn't mean "useful"; and i can push other features, simpler for me and better for them.
These are not stupid or uneducated people. They use Firefox at home. If they hit Slashdot from work, they're likely to be doing it via IE.
Come on people, it's a browser. We computer people tend to lose a great deal by getting stuck on minor issues like what browser people use. There are many very intelligent people who use internet explorer. It's a fact. And they are't even exceptions. The truth is, 90% of functionality is the same. The difference doesn't justify what we make of it.
I'm not hitting on your comment btw. It's just the habit of not seeing the forest because of the trees that i have a problem with.
Anyways, what I really wanted to say: a side effect of this comment system is that it'll favor a lot more comments with a score of 3 or lower. They are the comments usualy hidden by default, and most people who'd read them didn't bother reload. I predict from now on comments will have more meaningful subjects from now on, and a lot less "Re:". Even if a comment is low-rated, if it has an interesting subject it may incline people to click on it.
Not really. Costs pretty much mean salaries, various services, wiring and boxes. Assuming the wiring can take the upgraded speed, and in all newer infrastructures it can, most of the new investment goes into "boxes": routers, software, modems at the client's side etc. All the other don't feel any increase in costs.
Any ISP of 100 people has like 30 guys who install wires, 30 who sell, 30 who make coffe, wipe the floors or do client service and 10 (at most) who have anything to do with the real technology. Of which 2 are newbies and 3 are backup:) Most of the infrastructure is the same. Salaries, rent, electricity bill, taxes, gas, all don't have anything to do with the speed of the service.
We're assuming the ISP's always work at the maxumum speed possible, and any upgrade, even by 10%, means complete replacement of their technology. That is not true. They, like any other business, provide the speed which offers them the best profit. Which is definitely not the best possible speed. Infrastructure upgrades are something that's done every 5-10 years, and ensures the _possibility_ of a much better service. Like going from twisted cable to ethernet, or from ethernet to fibers. Each allows for a 100x increase potential, which allows small upgrades for the next 5-10 years. Any other way would be wasteful.
Back to the problem at hand: when faced with increased bittorrent trafic, they make a more or less in depth marketing analisys: what do they gain in cutting the trafic, and what do they loose in clients who give up their service? I don't know your market, but in my city broadband isp's are pretty zonal, so you can't easely switch.
On the bright side, here in Bucharest all city is covered in fiber, and the last mile is ethernet. So we have real metro speeds of 1-10 MB/s (ie 10-100 Mb, as much as ethernet allows), and since the fiber makes the internet market between isp very open, the price of external internet has dropped too (I pay around 20 euros for half a meg - a lot more then I ever use, since most heavy trafic is internal). But we do have a lot of real (read fierce) competition between isp too.
Very nice is the distinction between internal/external trafic. First internal meant the local isp, with speeds over 1 MB. Then most of the city. And in the last year, without _any_ change in my subscription, it's expanding to several larger cities. Actually, any subscription at my isp has the same internal speeds. I like it:)
To put this another way, if everyone saturated their pipe, they would have to charge upwards of 10x for your cable or DSL connection as they currently do.
And i'd have belived that, if i hadn't seen yahoo mail jump from 4 megs to 250 megs only days after it became apparent gmail was here to stay.
I know a bit about isp business, and most (read 95%) percent of costs go to things without any connection to actual speed. To double the speed costs less then 10% of what you invested so far. They _can_ provide better service, it's just cheaper and more convenient to throttle existing trafic first. They're not evil, just practical.
Best I can think of, so far, is to give your computer to charity and buy a new one. Of course, before giving it away you'll have to wipe the hard-drive clean.
Deep Space One was it. Faster, stronger and a lot smarter. Wasn't built to go outside the solar system, but in engineering terms is just as successful - it's done a lot more then it was supposed to.
Slashdot Mirror
I could make the point that it's easier to keep a few atomic weapons then build new ones, and inspections wouldn't find them. But that doesn't leave room for debate
I can't say if atomic weapons can be build in secret. My gut feeling tells me they can. Think about China, or US, or even Japan. All are big enough so that a large well funded, distributed project could go undetected. Not every country is Irak.
Anyways, I wanted to say something else. Someone before us mentiones something about how conflicts grew in number of casualties all over our history, until the end of WW2, the likely reason beeing big powers avoided was for fear of a nuclear conflict. Now, imagine trying to enforce a no-nuclear pact against China. 1 bilion people, a quarter of a _BILION_ potential army. If they wanted to build anything, they could. Trying to stop them, succesful or not, would be the most bloody affair since WW2. Same with US, Russia, India, Japan, etc.
How about if Germany made a few nukes? US would try to invade, but who do you think the rest of Europe would side with? WW3?
This "disarmament with teeth" sounds ok on paper, but it's nor enforceable except against Irak and (we'll see) Iran.
The US gov't signed and Congress approved a treaty whose goal is to rid the world of nuclear weapons. They, and many others, obviously disagree.
I don't know the text or the context, but i'll go on a limb and say they didn't mean to get rid of _all_ nukes. The big problem with the cold war was not the mere danger of using nuclear weapons: we used them before and with arguably good results. The real problem was that the number and the power of existing weapons could destroy all life on Earth a few times over. That is what all those treaties are trying to fix.
Plus, it's too expensive to keep 20.000 nukes when you can keep 1000 and sleep just as well.
Nuclear inspections work. They worked to enforce the US and USSR's treaties, and they've worked to prevent other countries from building nukes (e.g. Iraq, Iran).
If i were you I wouldn't mention Irak, considering it was invaded while the head of UN inspectors was saying on CNN they didn't find anything. This is a big example of how the inspections didn't work, not in the sense they didn't find what they were supposed, but that they weren't trusted enough to prevent war - which is ultimately their job.
And finaly:
This seems to imply that the world is safer with nuclear weapons that without.
Hell yes, i meant it. I'm pretty sure Eastern Europe where I live would have seen war the past 50 years if it weren't for the cold war. Which means my parent may have already died, and I wouldn't write this. So yes, I feel quite safe with nuclear weapons.
I wonder, where do you live? US maybe? That would explain a lot. I just thought a US citisen is really threatened only by a nuclear attak. Right now (and for the past 50 years) they were quite safe from anything else. The rest of us, on the other hand, know that war on home ground is a possibility.
Also, without nuclear weapons there would be no balance - US would clearly be the best military power.
And how exactly would eliminating _all_ nuclear weapons help things?
What really keeps most nations from building and using nuclear weapons is neither that UN disagree (hah) nor that it's too hard technically, It's the fact that most of their territory would turn to glass soon after.
If now every nation would get rid of nuclear weapons do you know what would happen? They'd all stard building them right back and the first to finish would win. Only they'd probably have to use them to make the point.
Don't you agree?
Yup, found out the first CS course at MIT is Scheme. Impressed.
You mean it's not there?! Damn, that's 30 minutes of my life I'm not gonna get back.
Ups... I might have done something stupid. He's 4 Insightful now. Live and learn...
Ok, ok, I wasn't talking about 2K or NT. From a user standpoint 2K is more or less the same with XP, they came out what? a year apart? and are way more similar then different. NT is The Grandpa. It deserves respect, but it was always a server OS first.
Thank you ;) Truth is, I never got "under the hood" with windows. I've kept strictly portable for many years now. Usually I'm hapy about it, but since C# I'm started to envy you guys a bit.
Heh. I actually forgot about 2K
Don't understand the -1 mod. It's a valid post, with a legitimate (and argumented) point of view. And the sad part is probably the ones modding it down are doing it from an XP machine.
I happen to agree with the parent, any OS before Windows XP was way too inadequate. Actually, from a business point of view, the biggest flaw of XP is it's way too good. It took Microsoft 6 years to come with a successor (Vista), and it's gonna be a very tough sell - convincing people they don't need their XP anymore isn't a piece of cake.
I for one before XP was using mostly linux for desktop. Not because I was a big fan (although I was) but because it was a better choice compared to '98 & company. Not so much now, because the need isn't the same.
Not his site. This time(?) it's actually quite legit, not a blog and no advertising.
Wikipedia's Neutral Point of View ideology may help here. Sounds like wishful thinking but they managed to make it work. Helps a lot they translated a vague concept (NPV) into a clear and short set of guidelines and made people respect it.
Expelling a student for cheating when statistics (and common sense) tell 50% of students do the same, yes, it is unethical. Basicaly you're expelling him because his number came up.
What i'm talking about is a realistic policy to prevent cheating. You can't expell half the students (don't even go there...), it's well, kinda not nice to expell by lottery just to make examples (which doesn't even work). So my 2 cents would be to find alternative punishments, something that can be more often and easely apply and which would "touch" the student without ruining his career. Something like dressing in yellow for a month.
Now if you want a more in-depth analysis (yawn... it's late!) I'll say things aren't likely to change very soon. Why? Because cheating is too useful for both sides. Its best use is to cover the cracks in the educational system. Classes are boring? The teacher is unqualified? The subject is outdated? No worry, some will cram, some will barely get by and all will cheat more or less. And they'll pass. Imagine how embarrasing would be for a teacher to have 90% of his class fail. Not really looking good on his resume, right? Did you notice how teachers are more lenient with cheating when they're doing a bad work with the class? And the strictest teachers usually do their job better then average?
For the students it's most useful in dodging the less interesting subjects, or in getting by when that's all you want, or in avoiding a bad teacher etc.
For the teachers it's also another power tool. When so many cheat, all you have to do is point your finger and someone is expelled. And make no mistake, a teacher half awake knows when students cheat.
For cheating to go away for good the system would have to work for real. And it's not gonna happen any time soon.
I never get why people always go to extremes in these cases. As if they're taboo and people are ashamed of it so they either pretend not to notice or overreact. To expel a student over cheating when statistics say 50% do it is... well, not very ethical. Not to do anything is just as wrong, of course. But I've yet to see a university having a tough but reasonable policy on cheating and implementing it.
Just use your imagination (and natural sadism if you have it) and make up 5-6 punishments for cheating that don't involve the student getting expelled. Let's see...
- failing the exam (always), and beeing examined at the end of the year by a comision
- paying a fine. Yes, money.
- having to take extra/advanced classes related to said subject
- take ethics classes (if the university has them)
- lots and lots of extra papers to make the next semester
- (my favorite) assist the offended teacher for the next semester
- "volunteer" work for the university (washing the dishes?)
Seven already? I was just started to think of what you can do with a free workforce around the campus.
It's a punishment, so it doesn't have to pretend it isn't. And given the alternative is expulsion, the students will definitely prefer it. Also, it's a much better deterrent. When somebody is expelled, nobody sees him anymore, so he'll be soon forgotten. But if you see the guy each morning washing the floors... you tend to remember why he's doing it.
3. Getting your interface specs from a dieing man after being helicoptered across town
Omg! This one is scary!
It's a matter of relationship. Last time (a long ago) when i did pair programming it was lots of fun and very useful. The one writing code took care of the micromanagement and had in mind at most the function he was working on, and the other was two steps ahead. If the side man keeps checking syntax and formating then the whole thing is useless.
It's a bit of a cultural thing too. For you, cams at every street corner are the pinnacle of Big Brother. In other countries they're just means to an end. And it's not just a matter of opinion, it may be that this system wouldn't fit well in America. But the growing feeling in the internet nation is that it's unnatural for a piece of civilisation not to be watched. Either by satelite, tourist's cameras or CCTVs. Now since that's obviously where we're heading (and I don't even disagree) we'd better start getting ready.
One very important point is who gets to watch these cameras. And who gets to watch _all_ these cameras. Two distinct questions, mind you. You can either go with very few people, or almost everybody. For the first question, I tend to go with the second approach, because on the long run it's impossible to restrict access and I wouldn't want to find out the mob can watch but i can't. As for who can search the data feed from all those cameras... that would better be restricted. Can't imagine a good thing average joe would get from this, but i wouldn't let the local bag guys look up my daily routine.
In my side of the world most (90%) of the grading is done by exams. Very little is graded based on papers and such, at all levels of education. And let me tell you, it sucks. We watch your system with more then a little envy. Why? Because exams don't measure _at all_ creativity or initiative. Only how well you learned _the subject_, and only the subject, and how smart you are. And btw, you can cheat too. This may be a cultural thing, but cheating in exams is quite tolerated by the faculty. Which I used to hate because allowing students to cheat prevents noticing and solving lots of problems with the education in general. "Look, they're passing, so we must be doing our job right".
Funny that kids today are pretty good at video games that are longer and more complicated than the ones we played, eh?
:) Warcraft 1 I couldn't finish because of the "puzzles", but once i got the rules for 2 (defend and harvest i think) i won easely.
Are you kidding?! Last time i watched someone play a contemporary game for half hour i was "WTF?! It's like click, point, click, point, click, point, click, point, click, point, click, point, click, point". Boring as hell. I remember about a week after discovering Dune 2 I was eating with my parents and suddenly i let the fork fall into the plate (splashing stew around) and exclaimed: "U! If i shape it [the defence] like an U, they'll all come right in!". My parents were a bit put of
I agree btw, when you say: "It would have been hard *not* to program". Now programming is designed to be done by the few, and it feels. I don't think any amount of work can make that go away. A "programmingforbeginners.org" would help a lot, though...
I've always wandered... did anybody try to teach lisp to people as a first language?
See? You're reading it from a developer point of view (90% of atomic functions offered are identical), and i meant it from a user point of view (does pretty much the same thing). That's exactly the problem, the incapacity to see things with a purely pragmatic eye: the user's. I suffered a lot because of it, mostly by implementic lots of cool features to my software only to find the clients extatic at the most basic functions. Now i'm asking twice, and i've learned that most times i can actualy get away with not doing the ugliest parts, simply because "hard" doesn't mean "useful"; and i can push other features, simpler for me and better for them.
These are not stupid or uneducated people. They use Firefox at home. If they hit Slashdot from work, they're likely to be doing it via IE.
Come on people, it's a browser. We computer people tend to lose a great deal by getting stuck on minor issues like what browser people use. There are many very intelligent people who use internet explorer. It's a fact. And they are't even exceptions. The truth is, 90% of functionality is the same. The difference doesn't justify what we make of it.
I'm not hitting on your comment btw. It's just the habit of not seeing the forest because of the trees that i have a problem with.
Anyways, what I really wanted to say: a side effect of this comment system is that it'll favor a lot more comments with a score of 3 or lower. They are the comments usualy hidden by default, and most people who'd read them didn't bother reload. I predict from now on comments will have more meaningful subjects from now on, and a lot less "Re:". Even if a comment is low-rated, if it has an interesting subject it may incline people to click on it.
Not really. Costs pretty much mean salaries, various services, wiring and boxes. Assuming the wiring can take the upgraded speed, and in all newer infrastructures it can, most of the new investment goes into "boxes": routers, software, modems at the client's side etc. All the other don't feel any increase in costs.
:) Most of the infrastructure is the same. Salaries, rent, electricity bill, taxes, gas, all don't have anything to do with the speed of the service.
:)
Any ISP of 100 people has like 30 guys who install wires, 30 who sell, 30 who make coffe, wipe the floors or do client service and 10 (at most) who have anything to do with the real technology. Of which 2 are newbies and 3 are backup
We're assuming the ISP's always work at the maxumum speed possible, and any upgrade, even by 10%, means complete replacement of their technology. That is not true. They, like any other business, provide the speed which offers them the best profit. Which is definitely not the best possible speed. Infrastructure upgrades are something that's done every 5-10 years, and ensures the _possibility_ of a much better service. Like going from twisted cable to ethernet, or from ethernet to fibers. Each allows for a 100x increase potential, which allows small upgrades for the next 5-10 years. Any other way would be wasteful.
Back to the problem at hand: when faced with increased bittorrent trafic, they make a more or less in depth marketing analisys: what do they gain in cutting the trafic, and what do they loose in clients who give up their service? I don't know your market, but in my city broadband isp's are pretty zonal, so you can't easely switch.
On the bright side, here in Bucharest all city is covered in fiber, and the last mile is ethernet. So we have real metro speeds of 1-10 MB/s (ie 10-100 Mb, as much as ethernet allows), and since the fiber makes the internet market between isp very open, the price of external internet has dropped too (I pay around 20 euros for half a meg - a lot more then I ever use, since most heavy trafic is internal). But we do have a lot of real (read fierce) competition between isp too.
Very nice is the distinction between internal/external trafic. First internal meant the local isp, with speeds over 1 MB. Then most of the city. And in the last year, without _any_ change in my subscription, it's expanding to several larger cities. Actually, any subscription at my isp has the same internal speeds. I like it
To put this another way, if everyone saturated their pipe, they would have to charge upwards of 10x for your cable or DSL connection as they currently do.
And i'd have belived that, if i hadn't seen yahoo mail jump from 4 megs to 250 megs only days after it became apparent gmail was here to stay.
I know a bit about isp business, and most (read 95%) percent of costs go to things without any connection to actual speed. To double the speed costs less then 10% of what you invested so far. They _can_ provide better service, it's just cheaper and more convenient to throttle existing trafic first. They're not evil, just practical.
Best I can think of, so far, is to give your computer to charity and buy a new one. Of course, before giving it away you'll have to wipe the hard-drive clean.
Deep Space One was it. Faster, stronger and a lot smarter. Wasn't built to go outside the solar system, but in engineering terms is just as successful - it's done a lot more then it was supposed to.