Much as I dislike this sort of "picking on the underdog", as I understand dilution, IDG doesn't have much choice. According to the law (again, as I understand, and IANAL) IDG has two options: they can defend their copyright against any and all comers, or they can lose their right to stop anyone from using the copyright -- even blatant abuses (someone else writing and publishing a For Dummies book, for example).
This is exactly the sort of zero-discretion legislation that's making things so much fun in the workplace and school. "Hi, we're human, we can think, but would prefer not!"
And, besides, you have to admit that this is a very polite message. "We understand why you're doing this, but we can't let you do this, and here's the reasons why. Can we be amicable about this?" (Well, amicable for lawyers, anyways.)
A major point of this argument rests on the idea that things are fundamentally probabilities. In probability theory, this concept is blatantly wrong: only unobserved quantities can have probability.
What quantum mechanics rests on is the fact that there is a fundamental limit to our abilities to observe events in the universe; that at some scale, we simply can't make the observation and so the best we can do to describe the situation is describe the probability distribution of the quantity. That doesn't mean that the particle doesn't have an absolute position and momentum or whatever; it just means that we can't know what those values are.
Now, from that perspective, we don't need a "probability mist" hanging over Platonia. We can have an crystalline set of configurations with infinite resolution in measurement. Now time can again be a path traced through this configuration space; what's changed is that the path itself is now described by a probability distribution. For any given instant of time, there are a infinite number of configurations that the Universe could be in, but we simply can't discriminate between them.
As far as I can tell (and, admittedly, I'm much better at probability theory than I am quantum mechanics), this is an alternate explanation that accounts for everything the author brings up, but doesn't require the rejection of time. Which isn't new; philosophers have been banging around the concept of "a frozen instant of time with created memories" for centuries, as far as I know.
A last point (this is way too long, I know): The author claims that a Universe-without-time is better than Universe-with-time because we don't have to explain away the existence of the arrow of time. It might be true that it's hard to explain why our Universe is the way it is (which I don't happen to believe: see the weak anthropic principle), but the author doesn't explain why, of all possible states, Platonia would choose to fall into this highly ordered configuration with such an elaborate "artificial past." To my mind, that's as much of a deus ex machina as the arrow of time itself.
I can't help but think this type of publicity contributes to the flash-in-the-pan image that Linux seems to be constantly fighting against. I mean, where was the news here? Corel is planning a Linux distro, a la Windows-and-Office. Whee. Nobody's buying NT. Whee. Microsoft is going to ship an X server. Whee.
It's as if ZDNet has a Linux-article quota to fill; and so we see yet another content-free Linux Is Good (TM) hype text. If we aren't prepared to buy the Microsoft hype, why should anyone be expected to be swayed by this stuff?
Admittedly, this is a big step beyond "Linux is crud and Unix is dead." Still, maybe the time has come to set the standards for Linux news higher.
If we look at NASA's recent history, we find a lot of distressing items (the Mars observer being lost comes to mind).
This isn't really fair. Most of the time, I'm astonished that any space mission has ever succeeded. The design challenges are overwhelming. You've got a miniscule power budget, a small and inflexible weight allowance, few ways to change things once the mission is under way, and zero first-hand experience with the challenges actually facing you. On one hand, you can't be over-prepared; and on the other hand, you can't take all the precautions you'd like to.
What's changed is the nature of the missions that NASA undertakes. The deep space probes have relatively simple missions they need to undertake, and the manned missions always had the astronauts to handle the complicated tasks and try and patch problems. But the Mars Observer had a fairly complicated protocol to follow, and the Observer itself had to handle the whole procedure. One little hitch that no one predicted, and down the whole mission goes.
I believe that what has changed is that the fire has gone away at NASA. People are treating it as their jobs, not as "I'm so lucky to be here. I'm going to do the best I can."
Well, my experience with the people I've met at JPL is that, in fact, they are dedicated to the space dream, they just aren't being allowed to do what they want. They're scientists and engineers who came in with grand ideas of manned missions to Europa and deep space exploration, only to be told they'll have to be happy with satellites and space shuttles. Maybe if NASA was allowed free rein, or at least freed from incessant micro-management, there could actually be some progress.
I think the real question this article raises is...if you're setting up a mail server, and you've chosen sendmail as your MTA, why in the heck would you want to run it on NT?
Well, turn it around. The real question is: if you've been told that your server has to be NT, wouldn't you prefer to be running sendmail than IIS?
The sad truth, as some of us are all too aware, is that technical considerations don't always guide the purchasing decisions. There's still a big high-level push to use Microsoft products in many places. Being able to use sendmail, with all the support and testing that it's gone through, very likely makes an NT mailserver that much more stable.
Re:Just nuke the journals? Maybe not. -- follow-up
on
Buffy and Dr. Varnus
·
· Score: 1
Just to answer some of the points that have been raised in reply:
a) Yes, we can burn CD's with the data and the viewers. Why bother? What OS is going to run those viewers in even five years, let alone fifty? How much software broke going from Linux 2.0 to 2.2?
b) For that matter, who's going to have a CD-ROM drive in ten years? Not ten feet from me is a room packed with 200MB tapes written to by a PDP that we can't read anymore, because you simply can't buy a compatible tape drive anymore.
c) If you think ASCII is a solution, talk to someone who doesn't write in English. ASCII is very, very limited. What's the code for the integration symbol? We could just use little JPEGs (which is hardly a standard that I'd bet on being around in fifteen years) for all those special characters, but then our document size blows up, in addition to now needing precise layout capabilities which ASCII text doesn't provide.
Now, all these problems have solutions, and the solutions are worth finding, because complete on-line access to scientific research would be a wonderful thing. What I'm concerened about is that the solutions be found before the journals all fold up. We don't need another Y2K-esque poor planning disaster on our hands.
The bounty model is not new. The corporate world has a name for it: consulting. Have something you need done but can't do? Hire someone outside to do the job for you. They do the work, you pay them money, and that's the end of it.
Instead of contrasting open source with the bounty style, let's look at why the bounty model might be superior to consulting:
Openness: The contractors have to be very explicit about what they want. They can't vaguely describe the project to lure you in, and then change the specs: it's all up on the webpage. What they ask for is what they'll get.
Working conditions: No chance of being treated as a slave employee here. They don't even have to know about you until you've finished the coding! You can work how you want, where you want, and when you want, in addition to being able to pick all your own projects.
Negotiating power: In conventional contracting, your pay is set before you start work -- which, for the coder, is when you are at your negotiating weakest. In the bounty model, you can have the entire product ready before you first approach the customer; which not only means that you have a better idea of the worth of your work, it means that you have the upper hand.
There are probably other benefits. The point here is that the bounty model is a replacement for contracting, not open source. The niche that contracting fills isn't going to disappear; people are still going to need in-house custom data acquistition software written. While bounty coding may have all the disadvantages described in the essay, conventional contracting has them worse. Why should we be against anything that shifts power towards the hackers?
ASCII, unfortunately, doesn't cut it. I can't remember the last time I read an article that didn't either have figures (graphs, charts, photos, whatever -- and these tend to be the most informative parts of the article) or equations; try explaining to a mathematician that he'll have to try typesetting differential geometry in ASCII. We do need a more complicated standard, and all the problems still apply.
Besides, "the dawn of the internet" is a very small time compared to the amount of time we need to keep documents around for. And even on simple text, there's been no established standard. Think of all those poor IBM mainframe people with their EBCDIC documents.
I can't speak for clinical researchers, but I haven't talked to anyone doing basic research who isn't in favour of having everything possible available online. The problem is archiving. Ten, twenty, one hundred years down the line, we still need to be able to get at the document. Computers aren't good at that.
Leaving aside the difficulty of keeping exabytes of data accessible, computers are always in a state of flux, up to and including file formats. Opening a document that was created even two years ago can be an adventure -- imagine what that will be like in a century. We have options, of course: we could take documents over a certain age off-line, and resign ourselves to dealing with a certain format for decades to come. The latter is a lousy choice (from experience); and the former negates all the benefits that Katz is praising.
Printing them on paper solves a lot of these problems. (For most people, journals are freely accessible, if not convenient: it's just a matter of finding the closest university library.) The ideal case would be that E-biomed also published its articles in journal form, or that everything that went to E-biomed was also published elsewhere. The problem is that journals are very, very expensive to publish. They're not high profile enough to attract good advertising budgets, and the subscription charges are already too high. Most bio journals already charge the author a massive per page cost to have the article published. If subscriptions to these journals end, then I don't know where they could make up the revenue.
E-biomed is good, but we can't simply let the journals vanish. They do fill a role. If freely accessible information is good, then information freely accessible for centuries is better. For a change, let's try to plan this move all the way through.
Slashdot Mirror
This is exactly the sort of zero-discretion legislation that's making things so much fun in the workplace and school. "Hi, we're human, we can think, but would prefer not!"
And, besides, you have to admit that this is a very polite message. "We understand why you're doing this, but we can't let you do this, and here's the reasons why. Can we be amicable about this?" (Well, amicable for lawyers, anyways.)
What quantum mechanics rests on is the fact that there is a fundamental limit to our abilities to observe events in the universe; that at some scale, we simply can't make the observation and so the best we can do to describe the situation is describe the probability distribution of the quantity. That doesn't mean that the particle doesn't have an absolute position and momentum or whatever; it just means that we can't know what those values are.
Now, from that perspective, we don't need a "probability mist" hanging over Platonia. We can have an crystalline set of configurations with infinite resolution in measurement. Now time can again be a path traced through this configuration space; what's changed is that the path itself is now described by a probability distribution. For any given instant of time, there are a infinite number of configurations that the Universe could be in, but we simply can't discriminate between them.
As far as I can tell (and, admittedly, I'm much better at probability theory than I am quantum mechanics), this is an alternate explanation that accounts for everything the author brings up, but doesn't require the rejection of time. Which isn't new; philosophers have been banging around the concept of "a frozen instant of time with created memories" for centuries, as far as I know.
A last point (this is way too long, I know): The author claims that a Universe-without-time is better than Universe-with-time because we don't have to explain away the existence of the arrow of time. It might be true that it's hard to explain why our Universe is the way it is (which I don't happen to believe: see the weak anthropic principle), but the author doesn't explain why, of all possible states, Platonia would choose to fall into this highly ordered configuration with such an elaborate "artificial past." To my mind, that's as much of a deus ex machina as the arrow of time itself.
It's as if ZDNet has a Linux-article quota to fill; and so we see yet another content-free Linux Is Good (TM) hype text. If we aren't prepared to buy the Microsoft hype, why should anyone be expected to be swayed by this stuff?
Admittedly, this is a big step beyond "Linux is crud and Unix is dead." Still, maybe the time has come to set the standards for Linux news higher.
This isn't really fair. Most of the time, I'm astonished that any space mission has ever succeeded. The design challenges are overwhelming. You've got a miniscule power budget, a small and inflexible weight allowance, few ways to change things once the mission is under way, and zero first-hand experience with the challenges actually facing you. On one hand, you can't be over-prepared; and on the other hand, you can't take all the precautions you'd like to.
What's changed is the nature of the missions that NASA undertakes. The deep space probes have relatively simple missions they need to undertake, and the manned missions always had the astronauts to handle the complicated tasks and try and patch problems. But the Mars Observer had a fairly complicated protocol to follow, and the Observer itself had to handle the whole procedure. One little hitch that no one predicted, and down the whole mission goes.
I believe that what has changed is that the fire has gone away at NASA. People are treating it as their jobs, not as "I'm so lucky to be here. I'm going to do the best I can."
Well, my experience with the people I've met at JPL is that, in fact, they are dedicated to the space dream, they just aren't being allowed to do what they want. They're scientists and engineers who came in with grand ideas of manned missions to Europa and deep space exploration, only to be told they'll have to be happy with satellites and space shuttles. Maybe if NASA was allowed free rein, or at least freed from incessant micro-management, there could actually be some progress.
Well, turn it around. The real question is: if you've been told that your server has to be NT, wouldn't you prefer to be running sendmail than IIS?
The sad truth, as some of us are all too aware, is that technical considerations don't always guide the purchasing decisions. There's still a big high-level push to use Microsoft products in many places. Being able to use sendmail, with all the support and testing that it's gone through, very likely makes an NT mailserver that much more stable.
Just to answer some of the points that have been raised in reply:
a) Yes, we can burn CD's with the data and the viewers. Why bother? What OS is going to run those viewers in even five years, let alone fifty? How much software broke going from Linux 2.0 to 2.2?
b) For that matter, who's going to have a CD-ROM drive in ten years? Not ten feet from me is a room packed with 200MB tapes written to by a PDP that we can't read anymore, because you simply can't buy a compatible tape drive anymore.
c) If you think ASCII is a solution, talk to someone who doesn't write in English. ASCII is very, very limited. What's the code for the integration symbol? We could just use little JPEGs (which is hardly a standard that I'd bet on being around in fifteen years) for all those special characters, but then our document size blows up, in addition to now needing precise layout capabilities which ASCII text doesn't provide.
Now, all these problems have solutions, and the solutions are worth finding, because complete on-line access to scientific research would be a wonderful thing. What I'm concerened about is that the solutions be found before the journals all fold up. We don't need another Y2K-esque poor planning disaster on our hands.
Instead of contrasting open source with the bounty style, let's look at why the bounty model might be superior to consulting:
Openness: The contractors have to be very explicit about what they want. They can't vaguely describe the project to lure you in, and then change the specs: it's all up on the webpage. What they ask for is what they'll get.
Working conditions: No chance of being treated as a slave employee here. They don't even have to know about you until you've finished the coding! You can work how you want, where you want, and when you want, in addition to being able to pick all your own projects.
Negotiating power: In conventional contracting, your pay is set before you start work -- which, for the coder, is when you are at your negotiating weakest. In the bounty model, you can have the entire product ready before you first approach the customer; which not only means that you have a better idea of the worth of your work, it means that you have the upper hand.
There are probably other benefits. The point here is that the bounty model is a replacement for contracting, not open source. The niche that contracting fills isn't going to disappear; people are still going to need in-house custom data acquistition software written. While bounty coding may have all the disadvantages described in the essay, conventional contracting has them worse. Why should we be against anything that shifts power towards the hackers?
ASCII, unfortunately, doesn't cut it. I can't remember the last time I read an article that didn't either have figures (graphs, charts, photos, whatever -- and these tend to be the most informative parts of the article) or equations; try explaining to a mathematician that he'll have to try typesetting differential geometry in ASCII. We do need a more complicated standard, and all the problems still apply.
Besides, "the dawn of the internet" is a very small time compared to the amount of time we need to keep documents around for. And even on simple text, there's been no established standard. Think of all those poor IBM mainframe people with their EBCDIC documents.
I can't speak for clinical researchers, but I haven't talked to anyone doing basic research who isn't in favour of having everything possible available online. The problem is archiving. Ten, twenty, one hundred years down the line, we still need to be able to get at the document. Computers aren't good at that.
Leaving aside the difficulty of keeping exabytes of data accessible, computers are always in a state of flux, up to and including file formats. Opening a document that was created even two years ago can be an adventure -- imagine what that will be like in a century. We have options, of course: we could take documents over a certain age off-line, and resign ourselves to dealing with a certain format for decades to come. The latter is a lousy choice (from experience); and the former negates all the benefits that Katz is praising.
Printing them on paper solves a lot of these problems. (For most people, journals are freely accessible, if not convenient: it's just a matter of finding the closest university library.) The ideal case would be that E-biomed also published its articles in journal form, or that everything that went to E-biomed was also published elsewhere. The problem is that journals are very, very expensive to publish. They're not high profile enough to attract good advertising budgets, and the subscription charges are already too high. Most bio journals already charge the author a massive per page cost to have the article published. If subscriptions to these journals end, then I don't know where they could make up the revenue.
E-biomed is good, but we can't simply let the journals vanish. They do fill a role. If freely accessible information is good, then information freely accessible for centuries is better. For a change, let's try to plan this move all the way through.