They have a backup generator. I wouldn't colo anywhere without one. But even backup generators aren't foolproof. Many tier 1 colocators have had
problems from time to time.
Like I said, that's 2 boxes, so its completely comparable. The same cost for a single box would have been ~500-550 $US/mo. That said, I'm pretty damn skeptical you can get a provider in the UK with good bandwidth and peering for that low of a price, because the Internet content is still mostly in the US, which means unless the sites in question are local to the country, they'll be paying to haul the traffic over the atlantic, and that's not cheap. I'm not up on the situation enough to know if euronet is still trying to gouge people (I imagine not, with the explosion of first tiers in the US running bandwidth over there).
In any event, "tech support" isn't even the right term. Everyone expects the front line to be morons. It's certainly true when I've dealt with UK networking companies, as much as it is here. The question is: how hard is it to issue a trouble ticket which escalates to a real engineer who knows how to handle the situation? In their case, I was unimpressed, but that's a problem everywhere, and it won't go away even IF there are a lot of spare engineers on the market (which, recession or not, I'm not seeing), because you have to have managers and HR people who can discern good from bad, which many cannot. Some places, the engineers approve the engineers, but in those places, they already HAVE the good ones.
I've had a pair of boxes hosted at Maxim.net
for quite a while. The prices are low --
for 2 boxes and 1 dedicated meg (which we
can fill 100% all month long for the same price),
we pay around $850. These are for boxes we built,
so hardware is not included, but that's still
pretty impressive. Although I've never remote-rebooted (both boxes up 185 days running
linux since they had a power outage 6 mo ago),
they have telephone reboot, as well as some
services.
They just merged into a larger company, and they
finally got a trouble ticketing system, but
customer service is still pretty awful, so its
fortunate I rarely need it. They have a few very
clued network guys, if you can get them.
Honestly, I've gotten to the point where I
don't trust wires of any kind, let alone wireless.
It's hard enough to trust the endpoints and the encryption between in a secure exchange, never mind trusting your ethernet. Maybe if the
government wasn't all gung-ho about preventing nefarious criminals from getting encryption (as if the government opposing it would stop them), then the citizens would already have lightweight encryption capable of securing even a communications medium like this. But, hey, "law-abiding citizens don't need encryption", right?
The other thing you overlooked is that the GPL prevents you from restricting redistribution, which means the first person to get a copy of the source could then retransmit it for much less.
You can charge whatever you like in regards to the product you sell, if it is GPL. The requirement is that you distribute the source. If you only sell the product with the source, you're fine. Otherwise, you have to provide a method to someone to get the source. I'm under the impression if you sold a binary-only version of anything, you'd have to provide the source for cost-of-duplication only to those who requested it, but the availability of the source is the key proviso.
Rather than posting on/., I started with a letter to my representative, since I'm now a Texas resident. I objected to the bill on the following counts:
Provides no ability for consumers to "opt out" of the installation when buying a computer.
Provides no exemption for businesses
Filtering software has been shown to be highly inaccurate
Provides no provisions for non-Windows operating systems
The Bill's fiscal note attached estimated a financial impact of zero. If Dell is required to install a $50 censorware package that a business does not need, and other manufacturers like Compaq are not, by virtue of not being in Texas, that hurts Dell, and hurting Dell hurts Texas, especially in my area (I live in Round Rock, where Dell world hq is)
Not when you've got a half dozen staffers. They could have one person read the emails, send form letter responses, and tabulate the results, then forward on the 'unique' letters with special points on to the congressman for personal perusal.
Many people who ARE good at what they do could
be equally good at other things. One engineer
I work with was fed up with his law firm not
letting him work on technology cases like he wanted, so he became an engineer. With a bit
more school, I have no doubt I could do the
reverse. Or move into a different tech niche than
I'm in. If you know how to learn, and care enough to do it, you can move where the tides take you.
Actually, I'm a little glad the market is getting less frenzied. I was starting to get a bit annoyed that people who thought GRE used port 47 were making 6 figures.
So, I imagine they will be modifying the kernel in places. I don't see any mention of plans to publish. (Not that the lack of mention here implies they won't) But I was glad to see TiVo's code published, and it would be good to see mention of plans for that from them.
I think a lot of us has read some of the little half-funny half-scary short stories, mostly speculative bits about what the future could be like. We probably all remember the guy that let his poor girlfriend sneak a peak at his textbooks, even though it was likely to cost him jail time (but didn't). We recently got to read speculation here on the future as affected by the Napster decision.
I have to say -- this whole situation reads like that. If back in 1998, someone wrote one of these, and said, "...and so people were jailed for forwarding jokes without permission...", it would have seemed funny. Now, it is a mournful day in a world where legislators make laws seemingly without regard to the lives of the people they serve.
You're sitting in the theater. The lights dim. A hush falls over the crowd, as the screen lights up. 40 minutes of ads and previews later, film rolls on Blizzard's Diablo, the movie.
Not long after we settle on our hero, a hearty warrior from the barbarian highlands, the film begins skipping frames. The barbarian forges out of the rogue camp, beset upon by skeletons. Despite their being 20 feet away, they swing
their axes, and our hero staggers under the blows! Meanwhile, the movie keeps freezing up entirely for a second or two at a time.
Finally, no more than a few minutes into the action, the movie cuts to the familiar Diablo font:
I'm so glad. I enjoyed q3 for linux a great deal, at least up until the 1.27g patch -- which has horked up my video a fair bit, and email to Loki went unanswered. Also, with X 4.0.2+, getting it running ought to be a great deal easier -- I needed a special 3dFX written X server -- not
that shouldn't be necessary.
I wouldn't normally have played q3 at all --
I bought it only because of the penguin sticker
on the box at Frye's. Now, I'd probably buy
the windows release of their next game if there
was no linux (although I can wait for a slightly-delayed linux release). So id has gained a customer for many products by having one
on linux.
Has anyone else had experiences with Loki they can share? I'm going to contact them again as soon as I find the time, since nothing really changed on
my machine from release to release, but I'm hoping they're normally more responsive.
(Score: -1, Redundant)
Napster hopes to operate under a plan in which it will pay recording companies $200M a year. But recording companies are "cool to the plan, saying they are moving ahead with their own online plans".
All I want to know is: why the hell do the labels have lawyers that move at warp speed, and engineers that move like snails? If they'd gone ahead with their online plans a bit faster, they would have had results that convinced them it wasn't even necessary to sue Napster. I don't really want to go back to having to swap mp3s over ftp and irc.
They should spend more effort on progress, and less on suing people.
My experience with tests was similar; I've always been a strong test-taker, because I infer things automatically from the test itself. My math education in high school was extremely poor. I knew advanced algebra in 6th grade, but learned very little beyond that. I had to infer what f(x) was while taking the SAT (I went back and changed an answer;)). But what you may have experienced was simply that the SAT does not necessarily solely test knowledge; it tests aptitude, because it is tuned to find for college success. College success is NOT based, necessarily, on a strong knowledge of the material. As it turns out, the SAT, college success, and "intelligence", at least as measured by tests like the Stanford-Binet, all correlate very strongly, whereas grades correlate much more weakly.
I've seen many people, Gould included, who do very well in live, but take tests poorly -- although in his case, he specifically mentions being very apprehensive, etc. However, I have yet to see, "Well, I got a 1550 on the SAT, but I can't seem to get past Assistant Manager at Jack in the Box."
Part of the problem is that, while there is a subset of people that standardized testing under-measures, school administrators tend to be far less accurate still, so for the allocation of scarce educational resources, we should rely on the most accurate measurement first, until we can develop a better measurement.
Don't get me wrong -- narrative evaluations are valuable, but my experience with them as a student at UCSC was poor. Either they were glowing, or they were "Test A was excellent. Final was good. Overall was good to excellent.", just as you say. The move to have letter grades automatically in all classes (in addition) is an acknowledgement (in my mind), that no matter how much we'd like to all just be friendly, holistic, and generous in our appraisals, in a world with limited academic resources, we need a platform to compare students from diverse backgrounds. This, of course, is another advantage of the SAT, not mentioned in the article -- it is standard to all students. An "A" is not always an "A", as it were.
And anyone reading the Mismeasure of Man should realize that Stephen Jay Gould is, despite being generally witty and intelligent, an anti-testing radical produced by a system in which he tested poorly. This doesn't mean the SAT is inapplicable by and large: the SAT correlates strongly with collegiate academic success.
The SAT test questions are generally designed to be difficult to answer for people who attempt to study the test, rather than the material. Some people may actual gain knowledge of the material with SAT-specific testing, because certain important verbal or mathematical areas may be covered. Unlike the MCSE exam, perhaps, where there are a specific limited number of questions and they are repeated and can be memorized and smuggled out for test-prep courses, the SATs have the entire english language to use for analogies, for example.
That said, the effects of test-prep, even by the very best courses, are 120-150 points on the
high end, for the total test. Historically, about
75% or so of those points disappear within about 8
weeks time after the studying stops. The rest
is a "real gain". But that hardly invalidates the test. Notice there is no complaint by the president about the correlation between SAT scores and academic success in the article, only a complaints that, (1) the preoccupation with the test has led to time being spent on studying FOR the test, rather than just the subjects it examines and, (2) the test is not a 'holistic' look at a student's education.
To this absurdity, I offer some opinions: first, students of any type study for tests. When you sit even in a college class, when your professor chalks something up on a board, you write it down, and add it to your points to study especially. Are you studying for a test? Yes. Are you also learning? Quite possibly.
Second, the test need not be holistic. Nothing requires a school to specifically use an SAT score as a prime evaluation tool. In fact, it should be obvious, if the SAT correlated poorly with collegiate academic success, it would be only a small consideration, because colleges primarily want to seek students that will do well in their school.
No, the real problem here is that the UC system doesn't like to select students based on achievement or intelligence or their potential success, because that would leave out the people who can't hack it. Therefore, any test which places any sort of number on people's capabilities is bad. And, of course, since everyone in high school can get a 4.0 if standards are lowered enough, that poses no such problem. The need for everyone to pass, to feel ok, and to be labelled "educated" is what is TRULY responsible for the decline in academic performance. The subjects you teach and the rate you teach them at any level is a window. Through that window, you can capture an optimum learning environment for some subset of students -- the average, the low end, the high end, and you can extend the window with some things (like AP classes, say). Our education system up through high school has been slid down to the low end -- we want everyone to pass, to advance, regardless of their committment to learning, their parental participation, etc. Because of limited resources, the people at the high end sleep through high school and spend their time administering unix boxes remotely, making more than their parents. (Ok, not all of them, but one I know of, should be studying like a sophomore, but knows it all, and is making nearly $100k/yr, saving up for MIT).
UC Santa Cruz is a great example -- look how they've moved away from just narrative evaluations, because graduate schools want letter grades from students -- and they want a test, in every case, as well.
Yes, I meant US. If I can pick and choose any music I want at $2-3 a song, I'd pay to get the digital versions. I'd want an assurance I could re-download them, of course, if I lost them or something. Maybe have the songs I'd paid for tied to a login. But the convenience and ability to select song-by-song is worth it. It would be nice, but not necessary, to have a whole-cd discount available.
Since Napster, mp3.com, etc, came on the scene, I have to admit: I don't even want to buy CDs, because its too much trouble. I have bought a number of them, like the U2 cd, but they're just sitting around in shrinkwrap, unless I couldn't find high-kbps digital versions.
I'm happy to pay for the music. As a consumer, there's 2 things I want out of digital music (maybe 3):
I want it now, conveniently, and I want to get it again if I lose a hard drive without paying for it again
I want to be able to buy only what I want. One song from an artist shouldn't cost $14.
I want artists to get more money, so more artists are supported, and more people are encouraged to go for it
One does have to wonder, what are the record companies good for, in this scenario? But whatever. I want my music digital, I want to pick and choose songs, and if I get those things, I'll happily pay for them. (I'd gladly pay $2-3 a song)
Some people enjoy 'roleplaying' game in the sense of Diablo II -- action, in which you play a character. But Diablo is no more a roleplaying game than is Tomb Raider, where you 'play' Lara Croft. A true RPG, where people assume an alternate persona, requires a certain consistency, an effort, at building up a set of mannerisms, a coherent scheme of mores for an alternate personality.
If one person builds up a personality, and another just 'takes over', they will very likely be unable to play the character in the same manner. They know nothing of the characters history, their associations. They will compromise the integrity of the character, and therefore contribute to the compromise of the game.
Viewed in perhaps another way, if you played D&D with a group of friends, who accepted your character as 'real' for the purposes of their gameplay, would you expect them to welcome another person playing that same character, who had no real idea about them, because they paid you for the privelege? Undoubtedly not.
Having not played Everquest, I can't say whether this is a valid concern for them in particular, but having recognized the issue, and heard debate in general about whether people have a right to all their virtual property, I have considered the question. I don't believe they necessarily have the right, if it is made clear up front that it is not acceptable.
The scarcity of items/characters in such a game is inevitable given larger demand than supply (almost a given, if people have something to try to do in a game), and a closed system. Diablo II is an excellent example. Too many players, not enough of certain items, closed system. But unlike an RPG, diablo 2 does not suffer a loss from the sale of items.
I wonder how the players and sellers would feel if sony/verant simply said: "ok, they are your property, feel free to sell. And, by the way, we're now offering characters and items, for an additional charge, for you to use in the game." And they could simply sell for less than whatever anyone could fetch on ebay, with a swift descent to a value of 0.
Actually, you do need servers, but you can have a big list, like a DNS cache file, and start from there. The advantages are a bit stronger than you think -- its not just 'the server', its 'all the servers'. And it shouldn't be too tough to have subdelegation, as well, with caching on the client side, to allow entire "chat" spaces to be moved off.
Disadvantages:
UDP is easier to spoof, because you don't need to predict sequence numbers. But random spoofing would require you KNOW when someone else was starting up a request -- ie, you'd have to be sniffing. If you were, it would be easy to hijack the tcp connection. This is why you don't hear complaints about spoofing of DNS very often. (The DNS exploits that involve cache poisoning are NOT examples of UDP spoofing)
limited bandwidth clients -- I'm asking about this. good question
Spoofing: not an issue, again. All you have to do is send a cookie back, and wait for a response before passing it on. Some sort of exchange is required anyhow, to confirm their join. The dead conversation thing isn't really an issue -- it would render YOU unable to hear, but anyone else talking would still be functional. Your final scenario can't be helped -- you can always DOS someone else, short of not knowing who they are -- but most people unleash DOS now trying to disrupt a channel, hack ops on it, etc. -- and that won't be effective in this scenario, so there's a diminished motivation.
Nothing popular yet, but at least one very talented software engineer I know of wants to create a DNS-based client-to-client chat service that would allow for a total distributed chat architecture, so that you could never DoS a server, only a single client. There would be
no such thing as "ops", and no need. Clientserver chat protocols had their day, and were good in their day. Time to change models.
Slashdot Mirror
They have a backup generator. I wouldn't colo anywhere without one. But even backup generators aren't foolproof. Many tier 1 colocators have had problems from time to time.
Like I said, that's 2 boxes, so its completely comparable. The same cost for a single box would have been ~500-550 $US/mo. That said, I'm pretty damn skeptical you can get a provider in the UK with good bandwidth and peering for that low of a price, because the Internet content is still mostly in the US, which means unless the sites in question are local to the country, they'll be paying to haul the traffic over the atlantic, and that's not cheap. I'm not up on the situation enough to know if euronet is still trying to gouge people (I imagine not, with the explosion of first tiers in the US running bandwidth over there).
In any event, "tech support" isn't even the right term. Everyone expects the front line to be morons. It's certainly true when I've dealt with UK networking companies, as much as it is here. The question is: how hard is it to issue a trouble ticket which escalates to a real engineer who knows how to handle the situation? In their case, I was unimpressed, but that's a problem everywhere, and it won't go away even IF there are a lot of spare engineers on the market (which, recession or not, I'm not seeing), because you have to have managers and HR people who can discern good from bad, which many cannot. Some places, the engineers approve the engineers, but in those places, they already HAVE the good ones.
I've had a pair of boxes hosted at Maxim.net for quite a while. The prices are low -- for 2 boxes and 1 dedicated meg (which we can fill 100% all month long for the same price), we pay around $850. These are for boxes we built, so hardware is not included, but that's still pretty impressive. Although I've never remote-rebooted (both boxes up 185 days running linux since they had a power outage 6 mo ago), they have telephone reboot, as well as some services.
They just merged into a larger company, and they finally got a trouble ticketing system, but customer service is still pretty awful, so its fortunate I rarely need it. They have a few very clued network guys, if you can get them.
Honestly, I've gotten to the point where I don't trust wires of any kind, let alone wireless. It's hard enough to trust the endpoints and the encryption between in a secure exchange, never mind trusting your ethernet. Maybe if the government wasn't all gung-ho about preventing nefarious criminals from getting encryption (as if the government opposing it would stop them), then the citizens would already have lightweight encryption capable of securing even a communications medium like this. But, hey, "law-abiding citizens don't need encryption", right?
Both of the Turok games were pretty fun, too. (not the 3rd, which was BS deathmatch)
I think the only real question is: did Nintendo maybe bury a Sega Genesis in your backyard, or something?
The other thing you overlooked is that the GPL prevents you from restricting redistribution, which means the first person to get a copy of the source could then retransmit it for much less.
You can charge whatever you like in regards to the product you sell, if it is GPL. The requirement is that you distribute the source. If you only sell the product with the source, you're fine. Otherwise, you have to provide a method to someone to get the source. I'm under the impression if you sold a binary-only version of anything, you'd have to provide the source for cost-of-duplication only to those who requested it, but the availability of the source is the key proviso.
Provides no ability for consumers to "opt out" of the installation when buying a computer.
Provides no exemption for businesses
Filtering software has been shown to be highly inaccurate
Provides no provisions for non-Windows operating systems
The Bill's fiscal note attached estimated a financial impact of zero. If Dell is required to install a $50 censorware package that a business does not need, and other manufacturers like Compaq are not, by virtue of not being in Texas, that hurts Dell, and hurting Dell hurts Texas, especially in my area (I live in Round Rock, where Dell world hq is)
I urge all Texas residents to do the same, today.
Not when you've got a half dozen staffers. They could have one person read the emails, send form letter responses, and tabulate the results, then forward on the 'unique' letters with special points on to the congressman for personal perusal.
Many people who ARE good at what they do could be equally good at other things. One engineer I work with was fed up with his law firm not letting him work on technology cases like he wanted, so he became an engineer. With a bit more school, I have no doubt I could do the reverse. Or move into a different tech niche than I'm in. If you know how to learn, and care enough to do it, you can move where the tides take you.
Actually, I'm a little glad the market is getting less frenzied. I was starting to get a bit annoyed that people who thought GRE used port 47 were making 6 figures.
So, I imagine they will be modifying the kernel in places. I don't see any mention of plans to publish. (Not that the lack of mention here implies they won't) But I was glad to see TiVo's code published, and it would be good to see mention of plans for that from them.
I think a lot of us has read some of the little half-funny half-scary short stories, mostly speculative bits about what the future could be like. We probably all remember the guy that let his poor girlfriend sneak a peak at his textbooks, even though it was likely to cost him jail time (but didn't). We recently got to read speculation here on the future as affected by the Napster decision.
I have to say -- this whole situation reads like that. If back in 1998, someone wrote one of these, and said, "...and so people were jailed for forwarding jokes without permission...", it would have seemed funny. Now, it is a mournful day in a world where legislators make laws seemingly without regard to the lives of the people they serve.
You're sitting in the theater. The lights dim. A hush falls over the crowd, as the screen lights up. 40 minutes of ads and previews later, film rolls on Blizzard's Diablo, the movie.
Not long after we settle on our hero, a hearty warrior from the barbarian highlands, the film begins skipping frames. The barbarian forges out of the rogue camp, beset upon by skeletons. Despite their being 20 feet away, they swing their axes, and our hero staggers under the blows! Meanwhile, the movie keeps freezing up entirely for a second or two at a time.
Finally, no more than a few minutes into the action, the movie cuts to the familiar Diablo font:
Your Connection Has Been Interrupted
Roll credits.
I'm so glad. I enjoyed q3 for linux a great deal, at least up until the 1.27g patch -- which has horked up my video a fair bit, and email to Loki went unanswered. Also, with X 4.0.2+, getting it running ought to be a great deal easier -- I needed a special 3dFX written X server -- not that shouldn't be necessary.
I wouldn't normally have played q3 at all -- I bought it only because of the penguin sticker on the box at Frye's. Now, I'd probably buy the windows release of their next game if there was no linux (although I can wait for a slightly-delayed linux release). So id has gained a customer for many products by having one on linux.
Has anyone else had experiences with Loki they can share? I'm going to contact them again as soon as I find the time, since nothing really changed on my machine from release to release, but I'm hoping they're normally more responsive.
(Score: -1, Redundant)
Napster hopes to operate under a plan in which it will pay recording companies $200M a year. But recording companies are "cool to the plan, saying they are moving ahead with their own online plans".
All I want to know is: why the hell do the labels have lawyers that move at warp speed, and engineers that move like snails? If they'd gone ahead with their online plans a bit faster, they would have had results that convinced them it wasn't even necessary to sue Napster. I don't really want to go back to having to swap mp3s over ftp and irc.
They should spend more effort on progress, and less on suing people.
My experience with tests was similar; I've always been a strong test-taker, because I infer things automatically from the test itself. My math education in high school was extremely poor. I knew advanced algebra in 6th grade, but learned very little beyond that. I had to infer what f(x) was while taking the SAT (I went back and changed an answer ;)). But what you may have experienced was simply that the SAT does not necessarily solely test knowledge; it tests aptitude, because it is tuned to find for college success. College success is NOT based, necessarily, on a strong knowledge of the material. As it turns out, the SAT, college success, and "intelligence", at least as measured by tests like the Stanford-Binet, all correlate very strongly, whereas grades correlate much more weakly.
I've seen many people, Gould included, who do very well in live, but take tests poorly -- although in his case, he specifically mentions being very apprehensive, etc. However, I have yet to see, "Well, I got a 1550 on the SAT, but I can't seem to get past Assistant Manager at Jack in the Box."
Part of the problem is that, while there is a subset of people that standardized testing under-measures, school administrators tend to be far less accurate still, so for the allocation of scarce educational resources, we should rely on the most accurate measurement first, until we can develop a better measurement.
Don't get me wrong -- narrative evaluations are valuable, but my experience with them as a student at UCSC was poor. Either they were glowing, or they were "Test A was excellent. Final was good. Overall was good to excellent.", just as you say. The move to have letter grades automatically in all classes (in addition) is an acknowledgement (in my mind), that no matter how much we'd like to all just be friendly, holistic, and generous in our appraisals, in a world with limited academic resources, we need a platform to compare students from diverse backgrounds. This, of course, is another advantage of the SAT, not mentioned in the article -- it is standard to all students. An "A" is not always an "A", as it were.
And anyone reading the Mismeasure of Man should realize that Stephen Jay Gould is, despite being generally witty and intelligent, an anti-testing radical produced by a system in which he tested poorly. This doesn't mean the SAT is inapplicable by and large: the SAT correlates strongly with collegiate academic success.
The SAT test questions are generally designed to be difficult to answer for people who attempt to study the test, rather than the material. Some people may actual gain knowledge of the material with SAT-specific testing, because certain important verbal or mathematical areas may be covered. Unlike the MCSE exam, perhaps, where there are a specific limited number of questions and they are repeated and can be memorized and smuggled out for test-prep courses, the SATs have the entire english language to use for analogies, for example.
That said, the effects of test-prep, even by the very best courses, are 120-150 points on the high end, for the total test. Historically, about 75% or so of those points disappear within about 8 weeks time after the studying stops. The rest is a "real gain". But that hardly invalidates the test. Notice there is no complaint by the president about the correlation between SAT scores and academic success in the article, only a complaints that, (1) the preoccupation with the test has led to time being spent on studying FOR the test, rather than just the subjects it examines and, (2) the test is not a 'holistic' look at a student's education.
To this absurdity, I offer some opinions: first, students of any type study for tests. When you sit even in a college class, when your professor chalks something up on a board, you write it down, and add it to your points to study especially. Are you studying for a test? Yes. Are you also learning? Quite possibly.
Second, the test need not be holistic. Nothing requires a school to specifically use an SAT score as a prime evaluation tool. In fact, it should be obvious, if the SAT correlated poorly with collegiate academic success, it would be only a small consideration, because colleges primarily want to seek students that will do well in their school.
No, the real problem here is that the UC system doesn't like to select students based on achievement or intelligence or their potential success, because that would leave out the people who can't hack it. Therefore, any test which places any sort of number on people's capabilities is bad. And, of course, since everyone in high school can get a 4.0 if standards are lowered enough, that poses no such problem. The need for everyone to pass, to feel ok, and to be labelled "educated" is what is TRULY responsible for the decline in academic performance. The subjects you teach and the rate you teach them at any level is a window. Through that window, you can capture an optimum learning environment for some subset of students -- the average, the low end, the high end, and you can extend the window with some things (like AP classes, say). Our education system up through high school has been slid down to the low end -- we want everyone to pass, to advance, regardless of their committment to learning, their parental participation, etc. Because of limited resources, the people at the high end sleep through high school and spend their time administering unix boxes remotely, making more than their parents. (Ok, not all of them, but one I know of, should be studying like a sophomore, but knows it all, and is making nearly $100k/yr, saving up for MIT).
UC Santa Cruz is a great example -- look how they've moved away from just narrative evaluations, because graduate schools want letter grades from students -- and they want a test, in every case, as well.
My memory is all shot to hell, but I sort of figured it was because it was ram-packed with hundreds of ips -- networks, host addresses, etc.
Yes, I meant US. If I can pick and choose any music I want at $2-3 a song, I'd pay to get the digital versions. I'd want an assurance I could re-download them, of course, if I lost them or something. Maybe have the songs I'd paid for tied to a login. But the convenience and ability to select song-by-song is worth it. It would be nice, but not necessary, to have a whole-cd discount available.
I'm happy to pay for the music. As a consumer, there's 2 things I want out of digital music (maybe 3):
I want it now, conveniently, and I want to get it again if I lose a hard drive without paying for it again
I want to be able to buy only what I want. One song from an artist shouldn't cost $14.
I want artists to get more money, so more artists are supported, and more people are encouraged to go for it
One does have to wonder, what are the record companies good for, in this scenario? But whatever. I want my music digital, I want to pick and choose songs, and if I get those things, I'll happily pay for them. (I'd gladly pay $2-3 a song)
Some people enjoy 'roleplaying' game in the sense of Diablo II -- action, in which you play a character. But Diablo is no more a roleplaying game than is Tomb Raider, where you 'play' Lara Croft. A true RPG, where people assume an alternate persona, requires a certain consistency, an effort, at building up a set of mannerisms, a coherent scheme of mores for an alternate personality.
If one person builds up a personality, and another just 'takes over', they will very likely be unable to play the character in the same manner. They know nothing of the characters history, their associations. They will compromise the integrity of the character, and therefore contribute to the compromise of the game.
Viewed in perhaps another way, if you played D&D with a group of friends, who accepted your character as 'real' for the purposes of their gameplay, would you expect them to welcome another person playing that same character, who had no real idea about them, because they paid you for the privelege? Undoubtedly not.
Having not played Everquest, I can't say whether this is a valid concern for them in particular, but having recognized the issue, and heard debate in general about whether people have a right to all their virtual property, I have considered the question. I don't believe they necessarily have the right, if it is made clear up front that it is not acceptable.
The scarcity of items/characters in such a game is inevitable given larger demand than supply (almost a given, if people have something to try to do in a game), and a closed system. Diablo II is an excellent example. Too many players, not enough of certain items, closed system. But unlike an RPG, diablo 2 does not suffer a loss from the sale of items.
I wonder how the players and sellers would feel if sony/verant simply said: "ok, they are your property, feel free to sell. And, by the way, we're now offering characters and items, for an additional charge, for you to use in the game." And they could simply sell for less than whatever anyone could fetch on ebay, with a swift descent to a value of 0.
Disadvantages:
UDP is easier to spoof, because you don't need to predict sequence numbers. But random spoofing would require you KNOW when someone else was starting up a request -- ie, you'd have to be sniffing. If you were, it would be easy to hijack the tcp connection. This is why you don't hear complaints about spoofing of DNS very often. (The DNS exploits that involve cache poisoning are NOT examples of UDP spoofing)
limited bandwidth clients -- I'm asking about this. good question
Spoofing: not an issue, again. All you have to do is send a cookie back, and wait for a response before passing it on. Some sort of exchange is required anyhow, to confirm their join. The dead conversation thing isn't really an issue -- it would render YOU unable to hear, but anyone else talking would still be functional. Your final scenario can't be helped -- you can always DOS someone else, short of not knowing who they are -- but most people unleash DOS now trying to disrupt a channel, hack ops on it, etc. -- and that won't be effective in this scenario, so there's a diminished motivation.
Nothing popular yet, but at least one very talented software engineer I know of wants to create a DNS-based client-to-client chat service that would allow for a total distributed chat architecture, so that you could never DoS a server, only a single client. There would be no such thing as "ops", and no need. Clientserver chat protocols had their day, and were good in their day. Time to change models.