Fascinating how the New York Times manages to fill the whole article, on the front- and continuation pages, discussing who released the documents, etc. and never explains who taught Iraq how to build atom bombs.
As the article says:
...the site has posted some documents that weapons experts say are a
danger themselves: detailed accounts of Iraq's secret nuclear research
before the 1991 Persian Gulf war. The documents, the experts say,
constitute a basic guide to building an atom bomb.
Iraq was indeed dangerous before the war: the first
Gulf war in 1991. A suspicious reader might, perhaps, wonder where Iraq learnt
advanced techniques of bomb construction. A likely source would be the country with the most nuclear weapons:
In one of the more stunning incidents, in September 1989, just one year
before the Iraqi military stormed over the Kuwaiti border, U.S. military
officials invited several Iraqi technicians to attend a "detonation
conference" at the Red Lion Inn in Portland, Ore.
The conference -- the Ninth Symposium (International) on Detonation,
was a crash course from the world's experts on how to detonate a
nuclear weapon. Among the named sponsors of the conference were the
Office of Naval Research, the Air Force Armament Laboratory, the
Army Armament Research, Development and Engineering Center, the Army
Ballistic Research Laboratory, Lawrence Livermore National
Laboratory, Los Alamos National Laboratory, the Naval Sea Systems
Command, Naval Surface Warfare Center, Office of Naval Technology
and Sandia National Laboratories, according to the conference
proceedings. [Source: "Made in America", San Francisco Bay Guardian, 25 Feb 1998]
Fun trivia: Who was president of the United States in September 1989?
No doubt, the New York Times just did not know of the above highly
secret information, because it was first revealed only on the floor of
the US House of Representatives. See the Statement of Henry B. Gonzalez,
Chairman, House Committee on Banking, Finance, and Urban Affairs, "Bush
Administration had Acute Knowledge of Iraq's Military Industrialization
Plans," July 27, 1992.
I tip my hat to the New York Times, masters of 'free'-world propaganda.
Can you recommend a good website with some tutorials and examples?
I don't know whether it's good, but suggestions for improvement are welcome for this example/tutorial. A summary of what's there:
This is a sample LaTeX document. You can use it as a start for
your own. We give examples of basic LaTeX writing, including
using a.bib file for handling bibliographies. We explain how
to use the graphicx package to include postscript or PDF
figures and the hyperref package to produce an automatically
hyperlinked document and to add URL's to your document.
http://arxiv.org/abs/physics/0412107 has a paper by me and a friend (who is also a physics professor) on how terrible introductory physics textbooks are. The paper itself is open source, by the way, and the source is available at the link above (click on "other formats").
The paper includes prices and weights for most of the textbooks. For the first version of the paper, about a year ago, we checked the prices and shipping weights by hand at Amazon. For the revised version last month, we wanted to complete the table -- but the already-checked prices had mostly gone up, so we had to throw out all the old data. I therefore wrote the Python script included in the source; I'd include it below but the posting robot complains about junk characters. The script will extract ISBN numbers from stdin (which was our tex file), look them up at Amazon, and give you the prices and weights. I use it track the prices of my (least) favorite books. Not one has got less expensive.
In our survey, the average book price was $152 (and average weight was 6.8 pounds): for boring and often incorrect, unphysical problems and explanations. It's no wonder so many people hate physics, and we have only ourselves to blame if we lose all our funding.
Every physicist should put their (good) textbooks at http://arxiv.org/, where they would be available, sans DRM, to everyone in the world. We are supported by the public; why should the public have to pay twice, the second time in the form of royalties?
He's from a university press, which is not to be confused with 'seller of exorbitant textbooks to university students' (e.g. the $80 calculus text).
The university presses publish mostly monographs with a limited circulation, such as graduate and upper-division undergraduate textbooks or books of no use to anyone except their author's case for promotion. Very few university presses have the marketing and sales muscle to compete against Addison-Wesley, Wiley, Saunders, etc. in the introductory-textbook market.
But I agree with your main point, that those introductory textbooks are absurdly priced (they are also terrible, which is another story).
that was ten years ago; I can only imagine how much it is now
What you feared. In footnote 2 of
this paper, a colleague and I tabulated the prices and weights of the main introductory physics textbooks.
Agreed, rebinding caps lock -> ctrl is a must. And type ctrl-[ instead of escape, since the escape key is usually hard to reach and anyway its position varies from keyboard to keyboard. ctrl-[ is also easier to reach than the alt key is (important for alt-x in Emacs).
"in this case there are specific statutes at work"
True. According to the US copyright code, if you grant a non-exclusive license (e.g. a license using the GPL) and then transfer the copyright to someone else (e.g. sell it to a company), the license survives the transfer only if "evidenced by a written instrument signed by the owner of the rights
licensed or such owner's duly authorized agent" [17 USC 205(e) http://www.bitlaw.com/source/17usc/205.html]. Is the GPL a signed, written instrument? It's written, but where's the signature (digital or otherwise)?
So I would not like to rely on how a judge would rule.
The FSF does not worry about revocation or transfers for GNU software because it requires contributors to transfer copyright to the FSF, and the FSF will not revoke the GPL (though they may relicense under the GPLv3 when that license is drafted and ready). Probably the FSF's incorporation charter would forbid it from doing something dubious anyway, e.g. revoking the GPL license and making a project purely closed-source.
But should others, who use the GPL without getting copyright assignments, worry? I do a bit and lean towards contract-based licenses. The main argument against them is that they are mostly incompatible with the GPL, under which most free software is licensed.
"Mr. Moglen is a copyright professor at Columbia Law School"
He is a professor of law and legal history, not copyright. However, I have no doubt that Prof. Moglen knows more about copyright law than I or than most people who read and post to slashdot know. However, he is busy and not easily available for these discussions, and the same goes for others with similar expertise. Furthermore, it is an issue on which the copyright experts are divided. Partly because there is so little case law on the subject. Most licenses are not to the public at large, and are about limiting rights, whereas the GPL and its cousins grant huge rights to everyone, and courts have not faced
the new issues that this situation creates. Are these licenses gifts? Courts usually refuse to enforce gifts, so they will not force a copyright holder to stick by his or her license. Is it a contract? If so, was it validly formed (offer, acceptance, consideration)? If not, did the licensee reasonably rely on the contract terms (e.g. in forming a company), which might now be revoked and thereby destroy the company?
Without court cases and law review articles addressing these issues (or sections in Nimmer on Copyright), one must think through these issues oneself and guess how courts will rule, and try to cover the bigger loopholes -- even if there is a significant chance of being wrong.
I agree with you that contracts cannot be unilaterally changed (unless, perhaps, the contract explicitly allows it). But one of the authors of the GPL, Prof. Eben Moglen, says repeatedly that the GPL is not a contract, it's a plain copyright license. Not much prevents you from revoking a plain (bare) license, even one that you say is perpetual. If you put up a sign saying "you can walk through my back yard forever", you can take down the sign later and revoke the permission -- because it is not a contract.
If, however, you make a contract with your neighbors on the issue, e.g. they pay you $100 right now, and in exchange may walk through your back yard forever, then you cannot revoke the permission without risk of being sued for breach of contract.
Similarly, if you post a program on your website and say "Here, you may, until the sun burns out, copy this program, make derivative works from it, and distribute them, as long as you
provide source code, etc." [i.e. implement the GPL], you can later take it down and revoke your permission. Then people will no longer be able to distribute their derivative works. Too bad for them. They should have made a contract with you: e.g. paid you $5 for a written license allowing all of the above.
I am not a lawyer, but that's what I learnt from reading Larry Rosen's informative book _Open Source Licensing_, as well as a few snippets on the subject of contracts.
I'd be interested to know of any corrections or additions to the above. It's an interesting issue that divides the experts in free-software licensing: whether to use a contract-based license (the Open Software License and many others) or a bare license (such as the GPL).
For many reasons, institutional libraries should pay the page charges
for their institutional users rather than pay proprietary publishers
for a license to read a paper.
Once a library ends a non-open-access electronic subscription, it
no longer can access articles published while it paid for
access. Same if the publisher goes out of business. With a paper
subscription, the library at least has printed copies from when it
subscribed.
The page charge is a one-time cost whereas the license fee is
paid by every library on the planet that subscribes to the journal.
With digital rights management (DRM), it might be paid each time a
user reads an article. Economists tell us that goods should,
in a free market, sell for marginal cost. The marginal cost to allow
someone to download a PDF file is nearly zero. Here is an estimate. A
T1 line, in the US, costs perhaps US$1000/month and offers about
0.2MB/s of bandwidth (megabytes not megabits), figures I got from
http://computer.howstuffworks.com/question372.htm. A month has
roughly 20 million seconds, so the cost per megabyte is given by:
($1000 / month) x (1 month/20 million sec) x
(1 sec / 0.2MB) x (100 cents / 1$)
That comes out to (1/40) cents/MB. Being conservative, say that
the full T1 bandwidth is only 25% used and that website administration
adds another factor of 2 to the cost, so multiply the above cost by 8.
The result is still only 0.2 cents/MB. For a typical 0.1MB pdf
article, it is 0.02 cents per download. Even if an article is
downloaded 10000 times (most authors would love to be so frequently
read!), the total cost is $2. It is a tiny cost compared to reviewing
articles, convering them to a decent publication format
(e.g. converting to TeX/LaTeX or dealing with absurd figure formats).
Along the lines of charging marginal cost, the open-access page
charges are probably the publisher's marginal cost (or only a bit
higher than it). Fine! Scientists would like helpful service
providers, such as open-access publishers, to stay in business for a
long time, and should be willing to support them.
Which brings up a third reason for open-access publication:
archiving. All publishers (except perhaps the Royal Society in
London, which has published for over 300 years) will go out of
business. What happens to their archive of pdf (or whatever format)
articles? With open-access publishing, other repositories can mirror
all the articles (thereby also redistributing the already tiny
bandwidth costs). Replication is the best way to preserve data.
Similarly, if the files need to be converted to a new format (PDF
version 25, SVG, or whatever), anyone can do so with open-access
articles.
A fourth reason, related to archiving, is search indexing. If
articles are freely available, auto-indexes such as
http://scholar.google.com/ will pick it up. If the article is
guarded by subscription passwords, it will not be indexed unless the
publisher submits the articles to the index. And they may be too
busy, or they may be out of business (see reason three above).
As an example of how proprietary publishers can act, Gordon and
Breach (GandB), since bought by Taylor and Francis, sued the American
Institute of Physics (AIP) in the US, Germany, France, Switzerland
because the AIP published a price comparison, and Gordon and Breach's
journals came at or near the bottom of the table (i.e. most
expensive). GandB claimed that the table was false advertising under
the Lanham Act. http://barschall.stanford.edu/ has trial
judgements and transcript, as well as the original articles and
pricing tables in dispute.
The AIP won the case almost completely, but it cost them millions
of dollars. And that cost has affected the thinking of research
librarians
Slashdot Mirror
Fascinating how the New York Times manages to fill the whole article, on the front- and continuation pages, discussing who released the documents, etc. and never explains who taught Iraq how to build atom bombs. As the article says:
Iraq was indeed dangerous before the war: the first Gulf war in 1991. A suspicious reader might, perhaps, wonder where Iraq learnt advanced techniques of bomb construction. A likely source would be the country with the most nuclear weapons:
Fun trivia: Who was president of the United States in September 1989?
No doubt, the New York Times just did not know of the above highly secret information, because it was first revealed only on the floor of the US House of Representatives. See the Statement of Henry B. Gonzalez, Chairman, House Committee on Banking, Finance, and Urban Affairs, "Bush Administration had Acute Knowledge of Iraq's Military Industrialization Plans," July 27, 1992.
I tip my hat to the New York Times, masters of 'free'-world propaganda.
Can you recommend a good website with some tutorials and examples?
I don't know whether it's good, but suggestions for improvement are welcome for this example/tutorial. A summary of what's there:
http://arxiv.org/abs/physics/0412107 has a paper by me and a friend (who is also a physics professor) on how terrible introductory physics textbooks are. The paper itself is open source, by the way, and the source is available at the link above (click on "other formats").
The paper includes prices and weights for most of the textbooks. For the first version of the paper, about a year ago, we checked the prices and shipping weights by hand at Amazon. For the revised version last month, we wanted to complete the table -- but the already-checked prices had mostly gone up, so we had to throw out all the old data. I therefore wrote the Python script included in the source; I'd include it below but the posting robot complains about junk characters. The script will extract ISBN numbers from stdin (which was our tex file), look them up at Amazon, and give you the prices and weights. I use it track the prices of my (least) favorite books. Not one has got less expensive.
In our survey, the average book price was $152 (and average weight was 6.8 pounds): for boring and often incorrect, unphysical problems and explanations. It's no wonder so many people hate physics, and we have only ourselves to blame if we lose all our funding.
Every physicist should put their (good) textbooks at http://arxiv.org/, where they would be available, sans DRM, to everyone in the world. We are supported by the public; why should the public have to pay twice, the second time in the form of royalties?
He's from a university press, which is not to be confused with 'seller of exorbitant textbooks to university students' (e.g. the $80 calculus text).
The university presses publish mostly monographs with a limited circulation, such as graduate and upper-division undergraduate textbooks or books of no use to anyone except their author's case for promotion. Very few university presses have the marketing and sales muscle to compete against Addison-Wesley, Wiley, Saunders, etc. in the introductory-textbook market.
But I agree with your main point, that those introductory textbooks are absurdly priced (they are also terrible, which is another story).
that was ten years ago; I can only imagine how much it is now
What you feared. In footnote 2 of this paper, a colleague and I tabulated the prices and weights of the main introductory physics textbooks.
Agreed, rebinding caps lock -> ctrl is a must. And type ctrl-[ instead of escape, since the escape key is usually hard to reach and anyway its position varies from keyboard to keyboard. ctrl-[ is also easier to reach than the alt key is (important for alt-x in Emacs).
"in this case there are specific statutes at work"
True. According to the US copyright code, if you grant a non-exclusive license (e.g. a license using the GPL) and then transfer the copyright to someone else (e.g. sell it to a company), the license survives the transfer only if "evidenced by a written instrument signed by the owner of the rights licensed or such owner's duly authorized agent" [17 USC 205(e) http://www.bitlaw.com/source/17usc/205.html]. Is the GPL a signed, written instrument? It's written, but where's the signature (digital or otherwise)? So I would not like to rely on how a judge would rule.
The FSF does not worry about revocation or transfers for GNU software because it requires contributors to transfer copyright to the FSF, and the FSF will not revoke the GPL (though they may relicense under the GPLv3 when that license is drafted and ready). Probably the FSF's incorporation charter would forbid it from doing something dubious anyway, e.g. revoking the GPL license and making a project purely closed-source.
But should others, who use the GPL without getting copyright assignments, worry? I do a bit and lean towards contract-based licenses. The main argument against them is that they are mostly incompatible with the GPL, under which most free software is licensed.
"Mr. Moglen is a copyright professor at Columbia Law School"
He is a professor of law and legal history, not copyright. However, I have no doubt that Prof. Moglen knows more about copyright law than I or than most people who read and post to slashdot know. However, he is busy and not easily available for these discussions, and the same goes for others with similar expertise. Furthermore, it is an issue on which the copyright experts are divided. Partly because there is so little case law on the subject. Most licenses are not to the public at large, and are about limiting rights, whereas the GPL and its cousins grant huge rights to everyone, and courts have not faced the new issues that this situation creates. Are these licenses gifts? Courts usually refuse to enforce gifts, so they will not force a copyright holder to stick by his or her license. Is it a contract? If so, was it validly formed (offer, acceptance, consideration)? If not, did the licensee reasonably rely on the contract terms (e.g. in forming a company), which might now be revoked and thereby destroy the company?
Without court cases and law review articles addressing these issues (or sections in Nimmer on Copyright), one must think through these issues oneself and guess how courts will rule, and try to cover the bigger loopholes -- even if there is a significant chance of being wrong.
I agree with you that contracts cannot be unilaterally changed (unless, perhaps, the contract explicitly allows it). But one of the authors of the GPL, Prof. Eben Moglen, says repeatedly that the GPL is not a contract, it's a plain copyright license. Not much prevents you from revoking a plain (bare) license, even one that you say is perpetual. If you put up a sign saying "you can walk through my back yard forever", you can take down the sign later and revoke the permission -- because it is not a contract.
If, however, you make a contract with your neighbors on the issue, e.g. they pay you $100 right now, and in exchange may walk through your back yard forever, then you cannot revoke the permission without risk of being sued for breach of contract.
Similarly, if you post a program on your website and say "Here, you may, until the sun burns out, copy this program, make derivative works from it, and distribute them, as long as you provide source code, etc." [i.e. implement the GPL], you can later take it down and revoke your permission. Then people will no longer be able to distribute their derivative works. Too bad for them. They should have made a contract with you: e.g. paid you $5 for a written license allowing all of the above.
I am not a lawyer, but that's what I learnt from reading Larry Rosen's informative book _Open Source Licensing_, as well as a few snippets on the subject of contracts. I'd be interested to know of any corrections or additions to the above. It's an interesting issue that divides the experts in free-software licensing: whether to use a contract-based license (the Open Software License and many others) or a bare license (such as the GPL).
For many reasons, institutional libraries should pay the page charges for their institutional users rather than pay proprietary publishers for a license to read a paper.
Once a library ends a non-open-access electronic subscription, it no longer can access articles published while it paid for access. Same if the publisher goes out of business. With a paper subscription, the library at least has printed copies from when it subscribed.
The page charge is a one-time cost whereas the license fee is paid by every library on the planet that subscribes to the journal. With digital rights management (DRM), it might be paid each time a user reads an article. Economists tell us that goods should, in a free market, sell for marginal cost. The marginal cost to allow someone to download a PDF file is nearly zero. Here is an estimate. A T1 line, in the US, costs perhaps US$1000/month and offers about 0.2MB/s of bandwidth (megabytes not megabits), figures I got from http://computer.howstuffworks.com/question372.htm. A month has roughly 20 million seconds, so the cost per megabyte is given by:
That comes out to (1/40) cents/MB. Being conservative, say that the full T1 bandwidth is only 25% used and that website administration adds another factor of 2 to the cost, so multiply the above cost by 8. The result is still only 0.2 cents/MB. For a typical 0.1MB pdf article, it is 0.02 cents per download. Even if an article is downloaded 10000 times (most authors would love to be so frequently read!), the total cost is $2. It is a tiny cost compared to reviewing articles, convering them to a decent publication format (e.g. converting to TeX/LaTeX or dealing with absurd figure formats).
Along the lines of charging marginal cost, the open-access page charges are probably the publisher's marginal cost (or only a bit higher than it). Fine! Scientists would like helpful service providers, such as open-access publishers, to stay in business for a long time, and should be willing to support them.
Which brings up a third reason for open-access publication: archiving. All publishers (except perhaps the Royal Society in London, which has published for over 300 years) will go out of business. What happens to their archive of pdf (or whatever format) articles? With open-access publishing, other repositories can mirror all the articles (thereby also redistributing the already tiny bandwidth costs). Replication is the best way to preserve data. Similarly, if the files need to be converted to a new format (PDF version 25, SVG, or whatever), anyone can do so with open-access articles.
A fourth reason, related to archiving, is search indexing. If articles are freely available, auto-indexes such as http://scholar.google.com/ will pick it up. If the article is guarded by subscription passwords, it will not be indexed unless the publisher submits the articles to the index. And they may be too busy, or they may be out of business (see reason three above).
As an example of how proprietary publishers can act, Gordon and Breach (GandB), since bought by Taylor and Francis, sued the American Institute of Physics (AIP) in the US, Germany, France, Switzerland because the AIP published a price comparison, and Gordon and Breach's journals came at or near the bottom of the table (i.e. most expensive). GandB claimed that the table was false advertising under the Lanham Act. http://barschall.stanford.edu/ has trial judgements and transcript, as well as the original articles and pricing tables in dispute.
The AIP won the case almost completely, but it cost them millions of dollars. And that cost has affected the thinking of research librarians