I think that it's more that the policy has not been enforced, rather than it having been repealed. I know that their are now primary and secondary schools with.edu domains, as well as non-American institutions. I suspect that this is basically keeping with the spirit of the original naming convention, rather than the letter. After all, when the TLDs were first created nobody would have considered that any school but a fairly serious university would even be able to connect to the net, much less want to have their own address. I'm sure that the idea of kindergardeners needing their own address was the last thing on the designers mind when they were coming up with the naming scheme way back when.
Again, restucturing is not always practical or possible. As an example, consider a situation that my Mom might face. She's the chairman of the school board in my home town, and they use an expensive software package to design their schoolbus routes. This is exactly the kind of software which is most likely to involve software patents, since an obscure algorithm may prove to be critical to getting good results. The school district can't really get out of the bussing business, since it contains large rural areas for which bussing is necessary, and they can't readily do the routing by hand because the problem is too tough- which is why the're willing to pay a huge fee for the software in the first place. They certainly can't get out of the business of running a school district!
Now admittedly they're not in a situation where they'll be in immediate trouble if the software becomes unavailable. They could keep using the old bus routes for a year or two while working on a replacement, but that's not necessarily the case for every organization that might be in a similar situation. Sometimes it just isn't possible to restructure your organization to work around a problem like that.
The biggest problem is that you don't always have a real choice in the matter. There are real world cases in which there's a single vendor for a critical piece of software that you desperately need. Even worse, that's likely to become the case more often as software patents become more prevalent, as they give the companies legal monopolies in specific areas. If the company that holds a monopoly on your critical piece of software decides to offer their software only on a rental basis, you have only unappetizing options. You can break the law by writing your own software that violates their patent, rent their software with the odious terms that implies, or do without something that may be critical to your business success. This is the kind of messy decision that happens in the real world of effective monopolies that "free market" appologists choose to ignore when claiming that consumers aren't really stuck.
What do you do when all of your choices are bad? Since the government is fundamentally responsible for this kind of mess by giving away monopolies in the form of software patents shouldn't they be involved in protecting people who wind up being screwed by them?
You can trademark all kinds of things, but the scope is limited. For instance, Apple computer can trademark the word Apple for use in reference to computers while leaving it OK for use in describing fruit. More amazingly, PeptoBismol can actually trademark their color of pink! That's right, you can't make your stomach medicine the same pink color as PeptoBismol, but you could paint your car that color if you particularly felt like it.
Of course, it would be pretty stupid to assume that altering of documents by politicians or other people is anything new.
What do you mean? No president before Clinton would ever do something so terrible as to manipulate records to try to hide his guilt. Honest Dick Nixon would never have stooped so low as to, say, erase 18 1/2 minutes of an incriminating taped conversation in order to cover his ass. Evil Bill Clinton is a complete departure from the behavior of every other president in the history of this fine nation.
Of course the flip side of this is that it's not always possible to tell who will be considered interesting in the future. In many cases, the most interesting use of archives is to look at the work of interesting people while they were working their way up and weren't of broad enough interest to attract major attention. Nobody knew that a 25 year old patent examiner named Albert Einstein was about to become a scientific star, but because we have his personal letters we can find out what he was doing scientifically and personally.
You never know if the next great author might be posting his early, great works to some fan e-mail list because he can't get his foot in the door at a major publisher. Maybe the next great debator is getting started in flamewars on Slashdot. Maybe the next great OS designer is getting into arguments with established academics on USENET. Oh, wait, that already happened, and we can only read the argument because somebody though to archive it. Maybe the next great philosopher who will be mostly ignored for 100 years is already publishing his early thoughts somewhere on the web. You can't always tell what will be valuable to the future until well after the fact, so preserving as much as possible is still a really good idea.
A truly wonderful example of this kind of thing are the early works of JRR Tolkein. The early history of the Silmarillion is absolutely fascinating and a wonderful example of the development of a literary theme. That's a work that wasn't published for over 50 years after it was started, but some of the earliest drafts still exist. Because those drafts are available, it's possible to see how it developed. Will the same thing happen when authors write everything in Word and write over old versions every time they change anything? How about if they're still very careful about keeping copies of early drafts but the formats change so much that they can't be read anymore?
Yes. This is the infamous "viral" nature of the GPL. GPLed software may not be distributed as a composite work with non-GPLed software without a special exception being written into the license. That's where the "But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License," come into play. Since the core of KDE, including Qt, is distributed as a coherent unit (and simply breaking Qt off into a separate tarball isn't enough) then the whole needs to be GPLed. As an alternative, though, you may include a special exception saying that it's OK to distribute the GPLed stuff with this specific non-GPLed stuff- and RMS did say that he felt that this was implied by the original authors distributing it that way- but that only applies if you have the right to change the licensing terms, i.e. you own the copyright. If you've incorporated somebody else's GPLed code, you'll have to get their permission to add the permission to distribute with the non-GPLed stuff. Of course this is now rather academic, since Qt is now available under the GPL.
Since when exactly has the GPL allowed its use in
commercial products?
Always. Think about commercial products like RedHat, Mandrake, etc. They're selling GPLed code and nobody even blinks. The GPL specifies only that source must be made available for the program and derived works. IOW, as long Sun, HP, etc. are willing to send the source, they're free to redistribute Gnome or KDE, and charge whatever they damn well feel like for it, to their hearts' content.
My general impression is that the big limitation is when you talk about linking to a shared library. If the library is narrowly enough shared that it must be distributed with the program that depends on it, then the program must be GPLed if the library is. If the library is LGPLed, though, then programs that depend on it and are distributed with it can be released under any license. There's some uncertainty, though, primarily as a result of this bit of section 2 of the GPL:
These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
There's obviously some wriggle room there, but my general impression is that this means that the whole core of Gnome or KDE must be licensed under the GPL if any part of it is; this is why RMS had problems with the GPLed KDE depending on a non-GPLed Qt. OTOH, other packages that are not distributed as part of the core but assume that it's already running on your machine need not be GPLed. This is a big win for a company like Sun or HP, since it means that they can distribute separate optional commercial software packages that depend on GPLed packages that are part of the base OS (like GTK or Qt). Equally important is this other bit of section 2:
In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program)
on a volume of a storage or distribution medium does not bring the other work under the scope of this License.
IOW, it's perfectly OK to put a GPLed program onto a CD containing Solaris, HPUX, etc. so long as it's clearly a separate program and you offer the source. There's no need to distribute GPLed chunks of your code separately from non-GPLed sections, just so long as it's clear what things fall under which license.
I mean, they'd probably rebuild Linux to B1 or better in the Orange Book.
Actually, one of the interesting points that they make is that Orange Book standards are not the be-all end-all of computer security. To quote:
The TCSEC provides a narrow definition of mandatory security which is tightly coupled to the multi-level security policy of the Department
of Defense. This has become the commonly understood definition for mandatory security. However, this definition is insufficient to meet the needs of either the Department of Defense or private industry as it ignores critical properties such as intransitivity and dynamic separation of duty. This paper instead uses the more general notion of mandatory security... in which a mandatory security policy is considered
to be any security policy where the definition of the policy logic and the assignment of security attributes is tightly controlled by a system security policy administrator.
The Orange Book was designed to implement the military's system of data security, but there are other potential security models that depend on mandatory access controls. To take this into account, the NSA researchers designed a much more flexible system in which the kernel implements some very generic mandatory access control structures but the details of the security model are substantially configurable. That means that you can implement an Orange Book B-class security model, but that's not the only security model that's available.
Of course there's still a very serious need to make those systems secure. Not being on the Internet does reduce your risk of being hax0red by skript kiddiez, but being a high profile, high value target attracts other kinds of attackers. You can bet that just about every unfriendly power out there is trying to get access to Intelink, either by infiltrating a mole or suborning someone who already has access. The number of potential attackers may be lower, but their dedication, skill, and support is likely to be a lot higher than random kiddiez.
And, of course, there's more to national security than keeping top secret military secrets from the prying eyes of the baddies. The long term economic health of the country is critical for national security, and that means helping companies that need security to get it. The NYSE, for instance, needs to have a lot of its critical systems exposed to the net, since their whole purpose is to send out critical information. It would be no good at all if they were broken into by morons intent on vandalizing the computers, and really, really bad if they were cracked by somebody with some subtlety and bad intentions, though I'm pretty confident that they're already running something more secure than Apache/Linux or IIS/Win2000. And, of course, that's just one example. Corporate espionage is a real potential problem, as is large scale credit card fraud, both of which could be carried out by cracking the right computers.
What could possibly motivate them to cooperate with an open source
effort, if not to compromise its security?
Gee, I don't know could it be:
The NSA is chartered to protect the communications security of the United States...
I don't know about you, but I think that pretty clearly covers the idea of new, higher security versions of existing software. Remember that SE Linux isn't really about encryption, but about adding a better security architecture to the system. That means helping to make the system cracker-proof, not making its communications more secure. They still have plenty of room to intercept and decode the other guys' communications even if they can't crack his boxen anymore.
People sometimes attack PGP because it is mostly used by criminals and beast bearded dirty GNU hippies.
Just shows how much they know. No true dirty GNU hippie would dream about using PGP with it's restrictive licensing. Everyone knows that the one true personal privacy program is GNU Privacy Guard, and it's the only one that GNU hippies would consider using. Furthermore, it's Free Software, so you can avoid the problems described above by hacking the source to use a different default location for your key ring that the attacker won't know. See the true power of Free Software in action! GNU forever!
Printed journals are good. They just need more availability, more participation, and a reduced cost. While Internet
distribution sounds great, we have a current system that needs improvement. It's more beneficial... not to mention easier
and cheaper... to improve the printed journals. Plus, we need the journals. Everyone currently reads journals, it's easier
for most of the scientific community at this point to deal with printed journals, and I think the scientific community would
suffer if we let journals degrade further or disappear entirely. So let's think of that first.
I think that you're 100% wrong. The problem is that print journals are bulky, inconvenient, and expensive. The size and inconvenience are essential problems of the print format; you just can't reduce 10,000 pages of journal- which is what top publications like Science and Nature are churning out per year now- into a format that can be easily accessed and archived while keeping it on paper. Furthermore, the need to keep actual physical copies in libraries means that only one person can read a given issue (or group of issues once they've been bound into volumes) at a time, and the librarians have a mess trying to reshelve the things. And a huge chunk of the cost associated with those journals is not the cost of doing the reviews and typesetting- the reviews are typically done pro bono by other scientists and much of the design is done by the author- but with the actual physical production and distribution of paper copies.
Electronic distribution solves many of those problems at a stroke. There's no need to keep shelves full of dead trees, or to print and ship the things all over the world, which is a big cost and effort saving for all involved. Searching for articles will also be much easier. Even the argument that there won't be much real savings in terms of paper because people will print out the articles anyway is flawed because people already copy article from the library for their own personal archives anyway.
The greatly reduced costs will be very beneficial overall. You'll probably be able to pay for fixed costs like editing and typesetting by charging authors a small fee (which many journals already do) and per-view costs can be paid for by banner ads. Anyone who's seen top journals will understand that this is exactly the kind of targeted advertizing that should pay very well, more than enough to cover costs. References can be replaced by actual links to the articles being referenced, which would be a huge win. Publishers could probably also make decent money by selling an annual CD-ROM version of the whole journal for people who want to be able to access it off-line. Journals could even start having separate "passed review" and "under review" sections so that the newest, most exciting work could be available (with obvious caveats) before the review process was 100% complete. Overall it would be a huge gain.
It doesn't work that way. You can always sue a company in the U.S. for violation of U.S. laws, even if they're based overseas. The only way they'd get an advantage by moving elsewhere is if they actually stopped doing business in the U.S.; otherwise the plaintiffs could always get their judgment from their local subsidiary or sezing their local assets, which they'd need to do business here. Since MS most definitely does want to do business in the U.S., they'll eventually have to fight these suits.
There's already a use [version] pragma that requires that the Perl in use be [version] or higher. IOW, use 6.0.0 would not prevent the code from working in a hypothetical perl7.
Which just means that you're not looking. In fact, there is currently a reference to an article on writing Gnome panel applets in the Perl slashbox here on slashdot. Your lack of knowledge does not mean that it isn't being done.
I also get the impression that even if using warnings and strict is the default behavior, it will still be possible to turn it off if that's desirable. IIRC somebody suggested that there be something like a use loose pragma that would turn off default strict behavior, and it's already possible to selectively turn off warnings for particular blocks of code. I don't think that anyone has seriously suggested making Perl a Bondage and Discipline language where everything must be done just so.
Well, there is a difference between replacing your software and replacing your hardware. One of the advantages of switching to Linux is that (theoretically, at least) you could download Debian off the web, burn it to a CD, and install it on every computer in your company for only a hair over the cost of installation. If you did it well, that installation could be handled very easily, too, with the computers booting off a floppy and automagically downloading your choice of packages over the company network. Replacing all of your Windows PCs with Macs is going to be a bit pricier.
But that implies something pretty harsh. If you assume that you'd need a new copy every time you replaced a piece of equipment that qualified you for an OEM license, you wouldn't be able to keep using your license after replacing a $40 CD-ROM drive or a $150 hard drive. That sounds way, way over the top.
There are critical differences, though. Fundamentally, though, the MS license doesn't spell out your rights. The rights that it's supposedly spelling out are rights that you would already have under the doctrine of first sale; instead it's detailing rights that you ought to have but don't because of the License "agreement". That means that GPL violators and MS license enforcers are both fundamentally doing the same thing- taking away user rights- and someone can be perfectly consistent in railing against both.
The city has a team to cheer for. Remember that in most cases the decision to build a stadium is made either by a city that lacks a team or by one where the team is making, at least implicitly, a threat to move. Don't underestimate the value that many people put on having sports teams to cheer for. One might equally well ask why a city should spend its money on a museum, symphony hall, or similar public structure. Most of those places also charge admission and are used by only a fraction of the population, but they're viewed as valuable additions to the community because some people want to use them and they're unlikely to be built otherwise. Not everyone in the city may go to the pro sports games, but the people who do actually get something from the stadium construction.
It depends on the hit list. If it were a list specifically saying "I want to kill these people", it would probably not be protected- especially if a search of his home turned up any additional evidence of plans and/or ability to carry out their murders. If it were a list of "Bullies and Worthless People at My School" it should (by analogy with this abortion ruling) be perfectly fine. Of course that's only in theory. In practice, the author might very well be arrested for making threats, have his home searched, his computer and anything vaguely weaponlike in his home seized, and his life made generally miserable. Even if he eventually won a lawsuit defending his right to publish such a list that would be small consolation for the suffering he experienced. Vindication may be something, but it's not as good as never being accused in the first place.
You are incorrect. The Schenck v. United States decision in which Holmes used the "Fire in a Crowded Theater" metaphor was unanimous in support of restricting some speech. OTOH, the reqirement for restricting speech is quite high; the words must be "used in such circumstances and are of such a nature as to create a clear and present danger that they will bring about the substantive evils that Congress has a right to prevent." IOW, they must be directly related to some wrong, that wrong must be something that the legislative body has a right to concern itself with, and they must create an immediate danger of that wrong. In the current case, it appears, the court has ruled that speech that is not a direct threat but only gives information that somebody interested in doing harm might use does not constitute an immediate danger and thus may not be restricted.
Would you feel the same way if you worked in a job where you felt threatened enough that you had taken steps to prevent your name and number from being in the phone book? If you knew there were people who advocated killing you because of what you do? How about if the people who put your name and home address on their web site had previously advocated murdering people in your line of work but didn't specifically mention it in this case?
The point is that in this case the people who put the names on the web site were harassers and terrorists. Some of them had been jailed for bombing clinics in the past, and many of them had previously advocated murdering abortion providers. Just because they didn't explicitly threaten or advocate violence on those web sites doesn't meant that they didn't advocate or even participate in violence in other forums. Can I avoid the law just by having two separate web sites that don't directly link to each other, one of which advocates murdering abortion providers and the other of which lists their names?
Slashdot Mirror
What kind of computer company makes you retire at age 65? What a silly number. They should make it 64 instead.
I think that it's more that the policy has not been enforced, rather than it having been repealed. I know that their are now primary and secondary schools with .edu domains, as well as non-American institutions. I suspect that this is basically keeping with the spirit of the original naming convention, rather than the letter. After all, when the TLDs were first created nobody would have considered that any school but a fairly serious university would even be able to connect to the net, much less want to have their own address. I'm sure that the idea of kindergardeners needing their own address was the last thing on the designers mind when they were coming up with the naming scheme way back when.
Again, restucturing is not always practical or possible. As an example, consider a situation that my Mom might face. She's the chairman of the school board in my home town, and they use an expensive software package to design their schoolbus routes. This is exactly the kind of software which is most likely to involve software patents, since an obscure algorithm may prove to be critical to getting good results. The school district can't really get out of the bussing business, since it contains large rural areas for which bussing is necessary, and they can't readily do the routing by hand because the problem is too tough- which is why the're willing to pay a huge fee for the software in the first place. They certainly can't get out of the business of running a school district!
Now admittedly they're not in a situation where they'll be in immediate trouble if the software becomes unavailable. They could keep using the old bus routes for a year or two while working on a replacement, but that's not necessarily the case for every organization that might be in a similar situation. Sometimes it just isn't possible to restructure your organization to work around a problem like that.
The biggest problem is that you don't always have a real choice in the matter. There are real world cases in which there's a single vendor for a critical piece of software that you desperately need. Even worse, that's likely to become the case more often as software patents become more prevalent, as they give the companies legal monopolies in specific areas. If the company that holds a monopoly on your critical piece of software decides to offer their software only on a rental basis, you have only unappetizing options. You can break the law by writing your own software that violates their patent, rent their software with the odious terms that implies, or do without something that may be critical to your business success. This is the kind of messy decision that happens in the real world of effective monopolies that "free market" appologists choose to ignore when claiming that consumers aren't really stuck.
What do you do when all of your choices are bad? Since the government is fundamentally responsible for this kind of mess by giving away monopolies in the form of software patents shouldn't they be involved in protecting people who wind up being screwed by them?
You can trademark all kinds of things, but the scope is limited. For instance, Apple computer can trademark the word Apple for use in reference to computers while leaving it OK for use in describing fruit. More amazingly, PeptoBismol can actually trademark their color of pink! That's right, you can't make your stomach medicine the same pink color as PeptoBismol, but you could paint your car that color if you particularly felt like it.
What do you mean? No president before Clinton would ever do something so terrible as to manipulate records to try to hide his guilt. Honest Dick Nixon would never have stooped so low as to, say, erase 18 1/2 minutes of an incriminating taped conversation in order to cover his ass. Evil Bill Clinton is a complete departure from the behavior of every other president in the history of this fine nation.
Of course the flip side of this is that it's not always possible to tell who will be considered interesting in the future. In many cases, the most interesting use of archives is to look at the work of interesting people while they were working their way up and weren't of broad enough interest to attract major attention. Nobody knew that a 25 year old patent examiner named Albert Einstein was about to become a scientific star, but because we have his personal letters we can find out what he was doing scientifically and personally.
You never know if the next great author might be posting his early, great works to some fan e-mail list because he can't get his foot in the door at a major publisher. Maybe the next great debator is getting started in flamewars on Slashdot. Maybe the next great OS designer is getting into arguments with established academics on USENET. Oh, wait, that already happened, and we can only read the argument because somebody though to archive it. Maybe the next great philosopher who will be mostly ignored for 100 years is already publishing his early thoughts somewhere on the web. You can't always tell what will be valuable to the future until well after the fact, so preserving as much as possible is still a really good idea.
A truly wonderful example of this kind of thing are the early works of JRR Tolkein. The early history of the Silmarillion is absolutely fascinating and a wonderful example of the development of a literary theme. That's a work that wasn't published for over 50 years after it was started, but some of the earliest drafts still exist. Because those drafts are available, it's possible to see how it developed. Will the same thing happen when authors write everything in Word and write over old versions every time they change anything? How about if they're still very careful about keeping copies of early drafts but the formats change so much that they can't be read anymore?
Yes. This is the infamous "viral" nature of the GPL. GPLed software may not be distributed as a composite work with non-GPLed software without a special exception being written into the license. That's where the "But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License," come into play. Since the core of KDE, including Qt, is distributed as a coherent unit (and simply breaking Qt off into a separate tarball isn't enough) then the whole needs to be GPLed. As an alternative, though, you may include a special exception saying that it's OK to distribute the GPLed stuff with this specific non-GPLed stuff- and RMS did say that he felt that this was implied by the original authors distributing it that way- but that only applies if you have the right to change the licensing terms, i.e. you own the copyright. If you've incorporated somebody else's GPLed code, you'll have to get their permission to add the permission to distribute with the non-GPLed stuff. Of course this is now rather academic, since Qt is now available under the GPL.
Always. Think about commercial products like RedHat, Mandrake, etc. They're selling GPLed code and nobody even blinks. The GPL specifies only that source must be made available for the program and derived works. IOW, as long Sun, HP, etc. are willing to send the source, they're free to redistribute Gnome or KDE, and charge whatever they damn well feel like for it, to their hearts' content.
My general impression is that the big limitation is when you talk about linking to a shared library. If the library is narrowly enough shared that it must be distributed with the program that depends on it, then the program must be GPLed if the library is. If the library is LGPLed, though, then programs that depend on it and are distributed with it can be released under any license. There's some uncertainty, though, primarily as a result of this bit of section 2 of the GPL:
There's obviously some wriggle room there, but my general impression is that this means that the whole core of Gnome or KDE must be licensed under the GPL if any part of it is; this is why RMS had problems with the GPLed KDE depending on a non-GPLed Qt. OTOH, other packages that are not distributed as part of the core but assume that it's already running on your machine need not be GPLed. This is a big win for a company like Sun or HP, since it means that they can distribute separate optional commercial software packages that depend on GPLed packages that are part of the base OS (like GTK or Qt). Equally important is this other bit of section 2:
IOW, it's perfectly OK to put a GPLed program onto a CD containing Solaris, HPUX, etc. so long as it's clearly a separate program and you offer the source. There's no need to distribute GPLed chunks of your code separately from non-GPLed sections, just so long as it's clear what things fall under which license.
Actually, one of the interesting points that they make is that Orange Book standards are not the be-all end-all of computer security. To quote:
The Orange Book was designed to implement the military's system of data security, but there are other potential security models that depend on mandatory access controls. To take this into account, the NSA researchers designed a much more flexible system in which the kernel implements some very generic mandatory access control structures but the details of the security model are substantially configurable. That means that you can implement an Orange Book B-class security model, but that's not the only security model that's available.
Of course there's still a very serious need to make those systems secure. Not being on the Internet does reduce your risk of being hax0red by skript kiddiez, but being a high profile, high value target attracts other kinds of attackers. You can bet that just about every unfriendly power out there is trying to get access to Intelink, either by infiltrating a mole or suborning someone who already has access. The number of potential attackers may be lower, but their dedication, skill, and support is likely to be a lot higher than random kiddiez.
And, of course, there's more to national security than keeping top secret military secrets from the prying eyes of the baddies. The long term economic health of the country is critical for national security, and that means helping companies that need security to get it. The NYSE, for instance, needs to have a lot of its critical systems exposed to the net, since their whole purpose is to send out critical information. It would be no good at all if they were broken into by morons intent on vandalizing the computers, and really, really bad if they were cracked by somebody with some subtlety and bad intentions, though I'm pretty confident that they're already running something more secure than Apache/Linux or IIS/Win2000. And, of course, that's just one example. Corporate espionage is a real potential problem, as is large scale credit card fraud, both of which could be carried out by cracking the right computers.
Gee, I don't know could it be:
I don't know about you, but I think that pretty clearly covers the idea of new, higher security versions of existing software. Remember that SE Linux isn't really about encryption, but about adding a better security architecture to the system. That means helping to make the system cracker-proof, not making its communications more secure. They still have plenty of room to intercept and decode the other guys' communications even if they can't crack his boxen anymore.
Just shows how much they know. No true dirty GNU hippie would dream about using PGP with it's restrictive licensing. Everyone knows that the one true personal privacy program is GNU Privacy Guard, and it's the only one that GNU hippies would consider using. Furthermore, it's Free Software, so you can avoid the problems described above by hacking the source to use a different default location for your key ring that the attacker won't know. See the true power of Free Software in action! GNU forever!
I think that you're 100% wrong. The problem is that print journals are bulky, inconvenient, and expensive. The size and inconvenience are essential problems of the print format; you just can't reduce 10,000 pages of journal- which is what top publications like Science and Nature are churning out per year now- into a format that can be easily accessed and archived while keeping it on paper. Furthermore, the need to keep actual physical copies in libraries means that only one person can read a given issue (or group of issues once they've been bound into volumes) at a time, and the librarians have a mess trying to reshelve the things. And a huge chunk of the cost associated with those journals is not the cost of doing the reviews and typesetting- the reviews are typically done pro bono by other scientists and much of the design is done by the author- but with the actual physical production and distribution of paper copies.
Electronic distribution solves many of those problems at a stroke. There's no need to keep shelves full of dead trees, or to print and ship the things all over the world, which is a big cost and effort saving for all involved. Searching for articles will also be much easier. Even the argument that there won't be much real savings in terms of paper because people will print out the articles anyway is flawed because people already copy article from the library for their own personal archives anyway.
The greatly reduced costs will be very beneficial overall. You'll probably be able to pay for fixed costs like editing and typesetting by charging authors a small fee (which many journals already do) and per-view costs can be paid for by banner ads. Anyone who's seen top journals will understand that this is exactly the kind of targeted advertizing that should pay very well, more than enough to cover costs. References can be replaced by actual links to the articles being referenced, which would be a huge win. Publishers could probably also make decent money by selling an annual CD-ROM version of the whole journal for people who want to be able to access it off-line. Journals could even start having separate "passed review" and "under review" sections so that the newest, most exciting work could be available (with obvious caveats) before the review process was 100% complete. Overall it would be a huge gain.
It doesn't work that way. You can always sue a company in the U.S. for violation of U.S. laws, even if they're based overseas. The only way they'd get an advantage by moving elsewhere is if they actually stopped doing business in the U.S.; otherwise the plaintiffs could always get their judgment from their local subsidiary or sezing their local assets, which they'd need to do business here. Since MS most definitely does want to do business in the U.S., they'll eventually have to fight these suits.
There's already a use [version] pragma that requires that the Perl in use be [version] or higher. IOW, use 6.0.0 would not prevent the code from working in a hypothetical perl7.
Which just means that you're not looking. In fact, there is currently a reference to an article on writing Gnome panel applets in the Perl slashbox here on slashdot. Your lack of knowledge does not mean that it isn't being done.
I also get the impression that even if using warnings and strict is the default behavior, it will still be possible to turn it off if that's desirable. IIRC somebody suggested that there be something like a use loose pragma that would turn off default strict behavior, and it's already possible to selectively turn off warnings for particular blocks of code. I don't think that anyone has seriously suggested making Perl a Bondage and Discipline language where everything must be done just so.
Well, there is a difference between replacing your software and replacing your hardware. One of the advantages of switching to Linux is that (theoretically, at least) you could download Debian off the web, burn it to a CD, and install it on every computer in your company for only a hair over the cost of installation. If you did it well, that installation could be handled very easily, too, with the computers booting off a floppy and automagically downloading your choice of packages over the company network. Replacing all of your Windows PCs with Macs is going to be a bit pricier.
But that implies something pretty harsh. If you assume that you'd need a new copy every time you replaced a piece of equipment that qualified you for an OEM license, you wouldn't be able to keep using your license after replacing a $40 CD-ROM drive or a $150 hard drive. That sounds way, way over the top.
There are critical differences, though. Fundamentally, though, the MS license doesn't spell out your rights. The rights that it's supposedly spelling out are rights that you would already have under the doctrine of first sale; instead it's detailing rights that you ought to have but don't because of the License "agreement". That means that GPL violators and MS license enforcers are both fundamentally doing the same thing- taking away user rights- and someone can be perfectly consistent in railing against both.
The city has a team to cheer for. Remember that in most cases the decision to build a stadium is made either by a city that lacks a team or by one where the team is making, at least implicitly, a threat to move. Don't underestimate the value that many people put on having sports teams to cheer for. One might equally well ask why a city should spend its money on a museum, symphony hall, or similar public structure. Most of those places also charge admission and are used by only a fraction of the population, but they're viewed as valuable additions to the community because some people want to use them and they're unlikely to be built otherwise. Not everyone in the city may go to the pro sports games, but the people who do actually get something from the stadium construction.
It depends on the hit list. If it were a list specifically saying "I want to kill these people", it would probably not be protected- especially if a search of his home turned up any additional evidence of plans and/or ability to carry out their murders. If it were a list of "Bullies and Worthless People at My School" it should (by analogy with this abortion ruling) be perfectly fine. Of course that's only in theory. In practice, the author might very well be arrested for making threats, have his home searched, his computer and anything vaguely weaponlike in his home seized, and his life made generally miserable. Even if he eventually won a lawsuit defending his right to publish such a list that would be small consolation for the suffering he experienced. Vindication may be something, but it's not as good as never being accused in the first place.
You are incorrect. The Schenck v. United States decision in which Holmes used the "Fire in a Crowded Theater" metaphor was unanimous in support of restricting some speech. OTOH, the reqirement for restricting speech is quite high; the words must be "used in such circumstances and are of such a nature as to create a clear and present danger that they will bring about the substantive evils that Congress has a right to prevent." IOW, they must be directly related to some wrong, that wrong must be something that the legislative body has a right to concern itself with, and they must create an immediate danger of that wrong. In the current case, it appears, the court has ruled that speech that is not a direct threat but only gives information that somebody interested in doing harm might use does not constitute an immediate danger and thus may not be restricted.
Would you feel the same way if you worked in a job where you felt threatened enough that you had taken steps to prevent your name and number from being in the phone book? If you knew there were people who advocated killing you because of what you do? How about if the people who put your name and home address on their web site had previously advocated murdering people in your line of work but didn't specifically mention it in this case?
The point is that in this case the people who put the names on the web site were harassers and terrorists. Some of them had been jailed for bombing clinics in the past, and many of them had previously advocated murdering abortion providers. Just because they didn't explicitly threaten or advocate violence on those web sites doesn't meant that they didn't advocate or even participate in violence in other forums. Can I avoid the law just by having two separate web sites that don't directly link to each other, one of which advocates murdering abortion providers and the other of which lists their names?