But if you *are* an Einstein, you would realize what you are good at and stick to that. Einstein did. (Even when he was wrong - he was doing theoretical physics - and being wrong at it which is *far* better than being right at something you are bad at.)
Are teachers paid enough? Check out the Los Angeles Unified School District pay scale and see for yourself. Go to the following link and click "New Salary Scale" Los Angeles School District
Being a physicist, my dream handheld has to be able to understand greek letters, superscripts, subscripts, and many, many mathematical symbols. The only decent way (i.e. no hunting through huge arrays of symbols and clicking the right one with a mouse) I have ever seen is with LaTex - where the greek lower case alpha is \alpha and so on. Even this can be improved on, I am sure, but for the moment, handhelds are useless to me.
Also, I have no idea why so many people are worrying about the power supply or the input mechanism. I would think you would want to think about what you want your handheld to do.
As long as we're dreaming, I think the ideal handheld that people use in the distant future will likely be subdermal and controlled by thought. What people will _want_ to do with a handheld in the distant future is completely beyond me.
The pole over on CNN indicates about 2:1 in favor of dropping the case. I think people are scared that any multiplicity of Microsofts will simply cause incompatability. Most slashdotters realize how much better an OS can be, but most other people do not. All they see is the repeated promises of software writers which have, historically, been dissapointing. Finally, after decades of promise, computing platforms for ordinary people are stabilizing and software is, generally, working. We all know things could be better and want change, but the average user is content with stability - even a stability which means users have to reboot daily. What I think slashdotters do not realize is that the average person using a computer just wants to get a relativly simple task done - word processing, printing, scanning, writing email, reading email, etc.
Republic of Rome is a good example. You play senators in the Roman Republic trying to become famous, or become rich, or just keep Rome from being sacked by the occasional invading army. There are, in essense, three possible outcomes: (1) Rome is sacked. All your characters die. (2) One person becomes so famous that they are made consol for life, and that person wins - but the Republic and all the ideals it stands for dies. Non-winning characters are now powerless 'advisors' to, essentially, a dictator. (3) Rome survives but no one person is so famous as to become Consol for life. Obviously, everyone wants ending number (3), however every time I play the players form two camps: the one side with a slight majority of votes, and the other with a slight minority - and the majority keeps all the honors to themselves - thus making all their character so famous that if they win any more wars, that lone character will become Consol for life. In my experience, at this time, the leading faction splits and either Rome is sacked, or a coalition government pulls it out by, essentially, cutting the most famous character out of public life completely. In fact, we used to bring famous people back once in a while specifically to take unpopular action - such as raise taxes - under the condition that they would not try to go for a solo win. Often, the player could actually have the power to simply bestow an honor on himself and win, but would not because that would end the game. Often these characters were great generals, so we would have to send them out to fight the Punic Wars, or something, and after fighting the first Punic War, they would be totally willing not to win the game (and end the game) because it would be so much fun to fight the second Punic War, instead.
So, in conclusion, we play games because they are fun, thus since winning the game ends the game, they stop the fun. Typically the people having the most fun are winning, so why should they want to end it?
I like games that can end without a "win". In Republic of Rome, it was surviving that last invading army. I'm sure people know of other examples.
This discussion reminds me of a book I read called "Finite and Infinite Games" by James P. Carse. In it, a finite game is a game played to "win" - ending in the termination of the game. An infinite game is played to continue the play at a different "level" or "type" of play - essentially to innovate. This seems the goal of the discussion, not zero/non-zero sum games, which I believe are defined as games in which by helping (or not helping) others, the total rewards of the game may be increased. I have seen 3 posts about Everquest. Although it is a non-zero sum game (you help others - in fact, you can make items for other people and sell it to them for less than they could buy similar items for and more than you can sell to a merchant for - essentially helping both you and them), it (Everquest) is clearly not an infinite game to any extent.
In the book, culture is often refered to as an example of an infinite game, but it is really something you have to get a sense of what the author is talking about. Don't just assume what you mean by "culture" is this guy's definition of infinite. After reading it, I was so caught up in infinite play that I tried to advocate it as a way of life to a couple people I knew. Then my Father pointed out feeding other people was a finite game. That shut me up.
I don't want to "re-define" this topic - I really love the discussions of both zero/non-zero games and what James P. Carse (author of Finite...) would call finite vs. inifinte games. Just wanted to point out a nice book on the subject.
Just a little more on the book: although it is a philosophy book, and I believe it is fairly modern, it is written very simply (nothing like most recent philosophy I have read). Even though I kind of dissed it's conclusions above (to live as an infinite player) I truely appreciate it as a totally different framework in which to see our actions (and our games and competitions - even the little ones we play like who gets out of the airplane first or who gets the closest "best" parking spot. Just think of all the little competitions we do every day. After reading it I asked myself about so many of these little things: should I even be concerned about these things?
To take it a step further: it allowed me to see the signes of when I was "competing" with another person: who knows the most obscure Unix commands, who knows the speed of light to more digits, who can recite the most Python (Monte) lines flawlessly, have I just gotten into a game of who's wife is better (or worse!), or who's got a better job/car/home. The frequency of this kind of interaction was really stunning to me. Of course I'm sure most of you are smarter than I and have already realized all this at a far younger age (another game...) but if you are interested in this kind of discusion about games, it is not a bad book to look at. It's $9.50 (US) at Amazon, likely free at your local library.
Cheers,
Greg
Linux will win with longer time scale experiments
on
Red Hat Finishes Last
·
· Score: 2
I'm surprised no mention was made of the limited time for evaluation of the products in these experiments. It is obvious to me many of the things the IDG.com reviewers were considering as highlights of the MS operating system would get really annoying after a long time, but allow people to get started quickly. (Here I refer to the "Moving into Management" section of the article). Although for something like a personal operating system, this may make a lot of sense, for a network operating system, however, I would expect them to realize the point and click interface in the MS OS is going to become a pain in the rear after a few months, and they will be wishing fondly for the text based tools they mention once (and don't comment on further) in Red Hat. Again, in the "Handling the Staples: File and Print" section, they fall into a similar trap. How many times will the average network admin want to start the "print administration wizard" and start clicking things before they start wishing for a faster, more precise technique? Below, the reviewers call the ASCII file configuration "a serious drawback" to Linux, and I agree it is a drawback - at the beginning when one is learning - but an advantage later. The most obvious quality of a network OS from a users point of view is stability ("is my email available to me, or not?"), and this is not mentioned in the review. The experiments needed to address this question correctly likely excede the time scale this review covers. (Perhaps one can simulate a month's activity by applying a large load on the network for a short span of time, but this seems very speculative at best.) The results of such an experiment seem necessary to me to conclude which is the "King of the network operating systems". (Well, since one of these American beers (Coors?) is the King of Beers perhaps the title is somewhat in jest?)
Having briefly read Dr. Schulman's paper from the aforementioned web site, one important aspect seems to be left out of current discussion. He says only "Here I use that framework [whole bunch of references] to show that small interaction does not destroy the arrows [of time]." Later he says: "Can this yield causal paradoxes? [Insert example here]... In principle such signals could be exchanged and paradoxes avoided as discussed in [a reference]. It is also possible that such an interaction would violate the small coupling assumption. At this stage I draw no conclusion." It seems to me he is talking about sending a few photons to transmit a signal from the reverse time region to the "normal" time region - and he's saying that may be too much! Much less taking something like a spacecraft or a person from one region to the other. Schulman's work does not say anything about such "larger" interactions. What is a small interaction? You have to read his paper. He has an interaction parameter in his simulaitions, but I can't figure out what it means in terms of a force or an energy and I can't compare it to, say, the energy of a photon or the mass-energy of a person. Perhaps a person more familiar with Schulman's termonology could comment?
For completeness, the URL of the paper is : http://xxx.lanl.gov/cond-mat/9911101 - for MS win users, you can click on 'other formats' and 'create PDF' to view the paper if you have Acrobat affiliated with your web browser.
I'm mostly wondering what various posters actually do for a living. Being a physicist, I don't think my experiences are in line with what most people have posted here. The origional question asks about "high tech", which I guess is likely in this forum to be connected to programming or sys admin or some specialized, arcane things I know little of. So, for the uninformed, please indicate what kind of work you are doing. In physics, it is very common to have a salaried position. I know some physicists who have specialized skills and consult for companies (or even on projects within the university...) and what strikes me about the difference between the positions is the need for marketing savvy. Self promotion is very important in both, but in consulting work, a more mercenary attitude is prevelant (in my experience). Most physicists would be very happy to take a perminant position for US$40-50k/year - as long as they were able to continue their physics research. However, their is a glut of PhD physicists - at least for the academic positions available. Red-Hat-esque ratios (say on the order of 500-1) of applicants for jobs exist even at the community college level (no research, just a teaching position). I offer this as an example of how different a "high tech" job market and job experiences might be. Don't take this as a complaint. I am very glad I have the opportunity to do fundamental physics research and it may be the most important work I ever do (as a job, of course: people are always more important than ideas.)
In the biography of Reagan, on A&E, a US cable television channel, it was put fairly bluntly that the US and the USSR leaderships were willing to completely disarm all nuclear weapons - as long as Reagan was willing to stop the SDI, which he wasn't. The web site for A&E's biography program is http://www.biography.com/, unfortunately it does not contain transcripts of the programs, most likely because they are selling copies of the show for about US$15. I believe the program I was watching was: Ronald Reagan: Role of a Lifetime, (however there seems to be a second biography program about him, so I might be wrong.) I recall the meeting was in Poland, but I could easily be wrong. It wasn't the most famous meeting, and it wasn't one where any big results were attained.
Since E.S.Raymond's definition of hack doesn't specifically mention computers, here is a list of hacks in theoretical physics (and perhaps applied math). I cannot compare these to other hacks (say, the wheel or something) in importance, but these do effect my daily life, and probably yours, too.
In no particular order: - Calculus (by Newton and Leibnetz) - Faraday's law of induction (relates time changing magnetic fields to an induced electric field, and vica versa) - Maxwell relating three equations (including Faraday's law (above)), fixing a bug in Ampere's law, adding a simple fourth law and discovering that light is an electro-magnetic wave. - The Schrodinger equation - part of the basis for quantum mechanics. A wave equation quite similar to the relation Maxwell found for light (however it relates the time derivitive of the wave to the second space derivitive, whereas Maxwell found for light, the second time derivitive relates to the second space derivitive.) Of course there were many, necessairy, steps between these events, and many great 'hacks' which are far greater intellectual achievements, but which also don't impact daily life outside the physics labs (or do so in a very indirect way - i.e. if gravity was different, the orbits of planets would be unstable, planets wouldn't form, etc.) A couple of the inventions I would attribute to fairly direct consequences of these works would be the transistor and the (electrical) generator. At the time of each of these hacks, many people had the opportunity to make them. Today, to do fundamental physics research most of us need expensive equipment, software and years and years of training; so the real, exciting, frontier of physics is closed to most people, which I find sad. (On a side note, I recomend visiting the Michel Faraday museum in London if you get the chance. You can see the equipment he made and worked with. The surroundings look like a cross between what I would think a blacksmith and a glassblower would use. When I visited in 1994 I was the only person there at the time!)
I agree: whatever Red Hat does with it's newfound wealth will help the open source community. What greater benefit could we ask than an open source windows emulator? Certainly the task in daunting. Even after years of development, we can be sure M$ will change their code to make it incompatable with whatever the open source community comes up with. However, Red Hat now has the money and can certainly get the talent to do it. (What is the ratio of applications to jobs at Red Hat? More than 100 to 1?) The project would have to be supported indefinatly, and that's exactly why it will have to be open source. Any propriatary software to compete with M$ won't sell - and for good reason: if the competetor to M$ goes out of business, the software will die with them. And we can all imagine what M$ would do to such a potential competitor. So why should Red Hat do it? In the short term, the media interest it would create would benefit the stock further. In the long term, the financial gains from sales of books, support contracts, etc. would be enormous. (The recent paper by E.S. Raymond, The Magic Cauldron http://www.tuxedo.org/~esr/writings/magic-cauldr on/ addresses just this question.) I think Red Hat is the first of many Linux distributions which will have this unique opportunity. My question is only: are they the ones who will actually do it? If so, I think I'll check into their stock, whatever it's current value - the potential is virtually unlimited.
Currently, in physics at least, some of the most prestegious journals (Physical Review Letters http://prl.aps.org/ to name one) has a limit of 4 pages. This makes it very difficult to write articles that are accessable to the non-expert. With science getting more specialized, I feel the move to electronic publishing to remove artificial lengths like these would be of great value. Writing and understanding articles would become vastly simpler - and thus reviewing articles could also become faster and easier. It is generally acknowledged that Phys. Rev. Letters are incomprehensible to researchers not already very familiar with the subject of the article. One difficulty with existing pre-print servers is that articles placed there are formatted for publication in journals where physical space is a very precious commodity. A last point: with web publication, data graphs are represented digitally, and thus it is easier to recover the data points themselves. However, there is so much more room available in web based publishing that people could actually include their raw data (something almost unheard of now if you have more than a hundred or so points.)
Slashdot Mirror
But if you *are* an Einstein, you would realize what you are good at and stick to that. Einstein did. (Even when he was wrong - he was doing theoretical physics - and being wrong at it which is *far* better than being right at something you are bad at.)
Cheers,
Greg
Are teachers paid enough? Check out the Los Angeles Unified School District pay scale and see for yourself. Go to the following link and click "New Salary Scale"
Los Angeles School District
Being a physicist, my dream handheld has to be able to understand greek letters, superscripts, subscripts, and many, many mathematical symbols. The only decent way (i.e. no hunting through huge arrays of symbols and clicking the right one with a mouse) I have ever seen is with LaTex - where the greek lower case alpha is \alpha and so on. Even this can be improved on, I am sure, but for the moment, handhelds are useless to me.
Also, I have no idea why so many people are worrying about the power supply or the input mechanism. I would think you would want to think about what you want your handheld to do.
As long as we're dreaming, I think the ideal handheld that people use in the distant future will likely be subdermal and controlled by thought. What people will _want_ to do with a handheld in the distant future is completely beyond me.
Greg G. Wood
The pole over on CNN indicates about 2:1 in favor of dropping the case. I think people are scared that any multiplicity of Microsofts will simply cause incompatability. Most slashdotters realize how much better an OS can be, but most other people do not. All they see is the repeated promises of software writers which have, historically, been dissapointing. Finally, after decades of promise, computing platforms for ordinary people are stabilizing and software is, generally, working. We all know things could be better and want change, but the average user is content with stability - even a stability which means users have to reboot daily. What I think slashdotters do not realize is that the average person using a computer just wants to get a relativly simple task done - word processing, printing, scanning, writing email, reading email, etc.
Cheers,
Greg
Republic of Rome is a good example. You play senators in the Roman Republic trying to become famous, or become rich, or just keep Rome from being sacked by the occasional invading army. There are, in essense, three possible outcomes: (1) Rome is sacked. All your characters die. (2) One person becomes so famous that they are made consol for life, and that person wins - but the Republic and all the ideals it stands for dies. Non-winning characters are now powerless 'advisors' to, essentially, a dictator. (3) Rome survives but no one person is so famous as to become Consol for life. Obviously, everyone wants ending number (3), however every time I play the players form two camps: the one side with a slight majority of votes, and the other with a slight minority - and the majority keeps all the honors to themselves - thus making all their character so famous that if they win any more wars, that lone character will become Consol for life. In my experience, at this time, the leading faction splits and either Rome is sacked, or a coalition government pulls it out by, essentially, cutting the most famous character out of public life completely. In fact, we used to bring famous people back once in a while specifically to take unpopular action - such as raise taxes - under the condition that they would not try to go for a solo win. Often, the player could actually have the power to simply bestow an honor on himself and win, but would not because that would end the game. Often these characters were great generals, so we would have to send them out to fight the Punic Wars, or something, and after fighting the first Punic War, they would be totally willing not to win the game (and end the game) because it would be so much fun to fight the second Punic War, instead.
So, in conclusion, we play games because they are fun, thus since winning the game ends the game, they stop the fun. Typically the people having the most fun are winning, so why should they want to end it?
I like games that can end without a "win". In Republic of Rome, it was surviving that last invading army. I'm sure people know of other examples.
Cheers,
Greg
This discussion reminds me of a book I read called "Finite and Infinite Games" by James P. Carse. In it, a finite game is a game played to "win" - ending in the termination of the game. An infinite game is played to continue the play at a different "level" or "type" of play - essentially to innovate. This seems the goal of the discussion, not zero/non-zero sum games, which I believe are defined as games in which by helping (or not helping) others, the total rewards of the game may be increased. I have seen 3 posts about Everquest. Although it is a non-zero sum game (you help others - in fact, you can make items for other people and sell it to them for less than they could buy similar items for and more than you can sell to a merchant for - essentially helping both you and them), it (Everquest) is clearly not an infinite game to any extent.
In the book, culture is often refered to as an example of an infinite game, but it is really something you have to get a sense of what the author is talking about. Don't just assume what you mean by "culture" is this guy's definition of infinite. After reading it, I was so caught up in infinite play that I tried to advocate it as a way of life to a couple people I knew. Then my Father pointed out feeding other people was a finite game. That shut me up.
I don't want to "re-define" this topic - I really love the discussions of both zero/non-zero games and what James P. Carse (author of Finite...) would call finite vs. inifinte games. Just wanted to point out a nice book on the subject.
Just a little more on the book: although it is a philosophy book, and I believe it is fairly modern, it is written very simply (nothing like most recent philosophy I have read). Even though I kind of dissed it's conclusions above (to live as an infinite player) I truely appreciate it as a totally different framework in which to see our actions (and our games and competitions - even the little ones we play like who gets out of the airplane first or who gets the closest "best" parking spot. Just think of all the little competitions we do every day. After reading it I asked myself about so many of these little things: should I even be concerned about these things?
To take it a step further: it allowed me to see the signes of when I was "competing" with another person: who knows the most obscure Unix commands, who knows the speed of light to more digits, who can recite the most Python (Monte) lines flawlessly, have I just gotten into a game of who's wife is better (or worse!), or who's got a better job/car/home. The frequency of this kind of interaction was really stunning to me. Of course I'm sure most of you are smarter than I and have already realized all this at a far younger age (another game...) but if you are interested in this kind of discusion about games, it is not a bad book to look at. It's $9.50 (US) at Amazon, likely free at your local library.
Cheers,
Greg
I'm surprised no mention was made of the limited time for evaluation of the products in these experiments.
It is obvious to me many of the things the IDG.com reviewers were considering as highlights of the MS operating system would get really annoying after a long time, but allow people to get started quickly. (Here I refer to the "Moving into Management" section of the article). Although for something like a personal operating system, this may make a lot of sense, for a network operating system, however, I would expect them to realize the point and click interface in the MS OS is going to become a pain in the rear after a few months, and they will be wishing fondly for the text based tools they mention once (and don't comment on further) in Red Hat.
Again, in the "Handling the Staples: File and Print" section, they fall into a similar trap. How many times will the average network admin want to start the "print administration wizard" and start clicking things before they start wishing for a faster, more precise technique? Below, the reviewers call the ASCII file configuration "a serious drawback" to Linux, and I agree it is a drawback - at the beginning when one is learning - but an advantage later.
The most obvious quality of a network OS from a users point of view is stability ("is my email available to me, or not?"), and this is not mentioned in the review. The experiments needed to address this question correctly likely excede the time scale this review covers. (Perhaps one can simulate a month's activity by applying a large load on the network for a short span of time, but this seems very speculative at best.) The results of such an experiment seem necessary to me to conclude which is the "King of the network operating systems".
(Well, since one of these American beers (Coors?) is the King of Beers perhaps the title is somewhat in jest?)
Having briefly read Dr. Schulman's paper from the aforementioned web site, one important aspect seems to be left out of current discussion. He says only "Here I use that framework [whole bunch of references] to show that small interaction does not destroy the arrows [of time]." Later he says: "Can this yield causal paradoxes? [Insert example here] ... In principle such signals could be exchanged and paradoxes avoided as discussed in [a reference]. It is also possible that such an interaction would violate the small coupling assumption. At this stage I draw no conclusion."
It seems to me he is talking about sending a few photons to transmit a signal from the reverse time region to the "normal" time region - and he's saying that may be too much! Much less taking something like a spacecraft or a person from one region to the other. Schulman's work does not say anything about such "larger" interactions.
What is a small interaction? You have to read his paper. He has an interaction parameter in his simulaitions, but I can't figure out what it means in terms of a force or an energy and I can't compare it to, say, the energy of a photon or the mass-energy of a person. Perhaps a person more familiar with Schulman's termonology could comment?
For completeness, the URL of the paper is : http://xxx.lanl.gov/cond-mat/9911101 - for MS win users, you can click on 'other formats' and 'create PDF' to view the paper if you have Acrobat affiliated with your web browser.
I'm mostly wondering what various posters actually do for a living. Being a physicist, I don't think my experiences are in line with what most people have posted here. The origional question asks about "high tech", which I guess is likely in this forum to be connected to programming or sys admin or some specialized, arcane things I know little of. So, for the uninformed, please indicate what kind of work you are doing.
In physics, it is very common to have a salaried position. I know some physicists who have specialized skills and consult for companies (or even on projects within the university...) and what strikes me about the difference between the positions is the need for marketing savvy.
Self promotion is very important in both, but in consulting work, a more mercenary attitude is prevelant (in my experience). Most physicists would be very happy to take a perminant position for US$40-50k/year - as long as they were able to continue their physics research. However, their is a glut of PhD physicists - at least for the academic positions available. Red-Hat-esque ratios (say on the order of 500-1) of applicants for jobs exist even at the community college level (no research, just a teaching position).
I offer this as an example of how different a "high tech" job market and job experiences might be. Don't take this as a complaint. I am very glad I have the opportunity to do fundamental physics research and it may be the most important work I ever do (as a job, of course: people are always more important than ideas.)
In the biography of Reagan, on A&E, a US cable television channel, it was put fairly bluntly that the US and the USSR leaderships were willing to completely disarm all nuclear weapons - as long as Reagan was willing to stop the SDI, which he wasn't.
The web site for A&E's biography program is http://www.biography.com/, unfortunately it does not contain transcripts of the programs, most likely because they are selling copies of the show for about US$15.
I believe the program I was watching was: Ronald Reagan: Role of a Lifetime, (however there seems to be a second biography program about him, so I might be wrong.) I recall the meeting was in Poland, but I could easily be wrong. It wasn't the most famous meeting, and it wasn't one where any big results were attained.
Since E.S.Raymond's definition of hack doesn't specifically mention computers, here is a list of hacks in theoretical physics (and perhaps applied math). I cannot compare these to other hacks (say, the wheel or something) in importance, but these do effect my daily life, and probably yours, too.
In no particular order:
- Calculus (by Newton and Leibnetz)
- Faraday's law of induction (relates time changing magnetic fields to an induced electric field, and vica versa)
- Maxwell relating three equations (including Faraday's law (above)), fixing a bug in Ampere's law, adding a simple fourth law and discovering that light is an electro-magnetic wave.
- The Schrodinger equation - part of the basis for quantum mechanics. A wave equation quite similar to the relation Maxwell found for light (however it relates the time derivitive of the wave to the second space derivitive, whereas Maxwell found for light, the second time derivitive relates to the second space derivitive.)
Of course there were many, necessairy, steps between these events, and many great 'hacks' which are far greater intellectual achievements, but which also don't impact daily life outside the physics labs (or do so in a very indirect way - i.e. if gravity was different, the orbits of planets would be unstable, planets wouldn't form, etc.)
A couple of the inventions I would attribute to fairly direct consequences of these works would be the transistor and the (electrical) generator.
At the time of each of these hacks, many people had the opportunity to make them. Today, to do fundamental physics research most of us need expensive equipment, software and years and years of training; so the real, exciting, frontier of physics is closed to most people, which I find sad.
(On a side note, I recomend visiting the Michel Faraday museum in London if you get the chance. You can see the equipment he made and worked with. The surroundings look like a cross between what I would think a blacksmith and a glassblower would use. When I visited in 1994 I was the only person there at the time!)
I agree: whatever Red Hat does with it's newfound wealth will help the open source community. What greater benefit could we ask than an open source windows emulator? Certainly the task in daunting. Even after years of development, we can be sure M$ will change their code to make it incompatable with whatever the open source community comes up with.r on/ addresses just this question.)
However, Red Hat now has the money and can certainly get the talent to do it. (What is the ratio of applications to jobs at Red Hat? More than 100 to 1?) The project would have to be supported indefinatly, and that's exactly why it will have to be open source. Any propriatary software to compete with M$ won't sell - and for good reason: if the competetor to M$ goes out of business, the software will die with them. And we can all imagine what M$ would do to such a potential competitor.
So why should Red Hat do it? In the short term, the media interest it would create would benefit the stock further. In the long term, the financial gains from sales of books, support contracts, etc. would be enormous. (The recent paper by E.S. Raymond, The Magic Cauldron
http://www.tuxedo.org/~esr/writings/magic-cauld
I think Red Hat is the first of many Linux distributions which will have this unique opportunity. My question is only: are they the ones who will actually do it? If so, I think I'll check into their stock, whatever it's current value - the potential is virtually unlimited.
Currently, in physics at least, some of the most prestegious journals (Physical Review Letters http://prl.aps.org/ to name one) has a limit of 4 pages. This makes it very difficult to write articles that are accessable to the non-expert. With science getting more specialized, I feel the move to electronic publishing to remove artificial lengths like these would be of great value. Writing and understanding articles would become vastly simpler - and thus reviewing articles could also become faster and easier. It is generally acknowledged that Phys. Rev. Letters are incomprehensible to researchers not already very familiar with the subject of the article.
One difficulty with existing pre-print servers is that articles placed there are formatted for publication in journals where physical space is a very precious commodity.
A last point: with web publication, data graphs are represented digitally, and thus it is easier to recover the data points themselves. However, there is so much more room available in web based publishing that people could actually include their raw data (something almost unheard of now if you have more than a hundred or so points.)