Totals for 2.4.18 debian woody: Grand total all directories: 3,704,994 (0.0022%) Total in kernel directory only: 14,058 (0.57%)
So if they're talking specifically of the kernel directory, and if the 80 lines shown are all there are, they're claiming that their decline from $200+ million per year to $60 million per year is entirely due to about one half of one percent of the code.
Off Topic: anyone know a better way than the following to handle large numbers of arguments?
bob@swirl:/usr/local/src/kernel/kernel-source-2. 4. 18$ find * -type f -name '*\.[ch]' | wc
8293 8293 221144 bob@swirl:/usr/local/src/kernel/kernel-sou rce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -1000` | grep total
324215 1061875 8471017 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -2000 | tail -1000` | grep total
551027 1776771 15043435 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -3000 | tail -1000` | grep total
631425 2131416 18512146 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -4000 | tail -1000` | grep total
965346 3486333 29067291 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -5000 | tail -1000` | grep total
581813 2316216 18338130 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -6000 | tail -1000` | grep total
134261 511247 4140371 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -7000 | tail -1000` | grep total
122814 502081 3903002 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -8000 | tail -1000` | grep total
222183 806751 6445312 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | tail -293` | grep total
171910 516407 4303860 total bob@swirl:/usr/local/src/kernel/kernel-sour ce-2.4. 18$ cd kernel/ bob@swirl:/usr/local/src/kernel/kernel-so urce-2.4. 18/kernel$ wc `find * -type f -name '*\.[ch]'` | grep total
14058 43823 343800 total
I somewhat read the NDA and all I saw was stuff about code.
Section 3, like most NDAs, claims everything including but not limited to the word "Pella" etched on SCO's office windows.
3. Definition of Confidential Information. " Confidential Information" means any and all data, technology, research, inventions, intellectual property, trade secrets, know how, computer programs, source code, file names, file trees or extensions, works of authorship, products, processes, methods, customer names, plans, forecasts, prices, business information, financial information, and other information shown or relayed by SCO to RECIPIENT on _______________________ [date].
It is amusing (in the tragic/comic sense) to me that what was originally designed to protect honorable companies from dishonorable people has become a weapon wielded by dishonorable companies. This looks like a pretty standard NDA, and a lot less onerous than most software EULAs (even those from supposedly honorable companies like Ximian).
My favorite piece: 11. Waiver. Any waiver of, or promise not to enforce, any right under this Agreement shall not be enforceable unless evidenced by a writing signed by the Party making said waiver or promise.
Translation - when the blow-dried business weenie / marketroid says, "Oh don't worry about this, it's just for our protection, we would only ever use it defensively." - he's lying. By the time the ink is dry there will be someone checking SourceForge, the USPTO, and the Library of Congress for your name to see what they can hijack.
I think your test was flawed. Please see here for a test showing one popular usage of synchronization (lazy initializers). One million cycles takes about 30 milliseconds longer (a hair more than twice as long) for the synchronized version as for the fastsynched (effectively unsynchronized after the first call) version. Unsynchronized actually came out slower, showing that compiler behaviour is probably a bigger issue.
30 milliseconds per one million cycles. The extra 6 lines of code (and their extra defect opportunities) cannot be justified in any but the most extreme performance optimization situation.
Agreed that this is the most likely outcome in the current legal system. Should it not be the case though, that this action would be protected by the same laws that protect people performing the Heimlich Maneuver?
it appears this academic has missed a crucial piece of the equation in these modern times:
Blatent Product Placement
Did you read the article?
Ms Clayton, who is a movie director and screenwriting lecturer for the University of London and the British Film Council, was commissioned by diet Coke [my emphasis].
I'd rather say she got it exactly right, starting with the product placement in the article about the study about movies.
Filesystem Hierarchy Standard has included the BSDs since 1995 - it's more than just Linux. I'm all in favor of questioning the assumptions to avoid getting caught on a local maxima, and I think this one is a dead end. FHS has some historical baggage, but it also has the strength of years of tempering.
Even if M$ manages to get DRM out there, how riddled with holes will it be?
The problem is not whether it works - we all know that DRM is technically impossible (analog hole). The problem is that combined with the DMCA, DRM makes fair use illegal. If Passport were being used for copyright protection, it would be a federal crime to report this security vulnerability.
I use the company's name or the product for my email.
This is what I do now too. But with dictionary spam becoming more popular (my news@ alias gets hit more than any other address), I'm guessing it won't be long before they start hitting hertz@, ebay@, amazon@, etc.
What happens when the customer orders something from Amazon - the purchase confirmation email comes from a non-human address.
This has been my single opposition to challenge response from the first time I heard of it. As I was reading your post, I hit on an idea that borrows from an idea posted by another user.
Put challenge response on my primary email address. Rotate my non-challenge response email address every 30 days.
So Amazon uses wog23t5s@traxel.com, while all my friends use my real address.
Toss in some accepted domains if you want to permanently accept from *.register.com (reversible, of course), and I think it might be a flexible, functional system.
e.g. tulip blubs sell well, but not like they used to.
I'm sure I'm reiterating what a few dozen other posts have already pointed out, but Holland's economic debacle was primarily due to the fact that the value of a tulip is almost entirely based on end-user utility. Information science, on the other hand, deals with the refinement of information. Much of this ability to process information is directed at the acquisition of additional wealth. That is, information science has far higher objective economic value because it directly results in the ability to harvest more wealth.
I would choose to politely disagree. Look at Canada as an example: We are right next to the US, but do we see any effect from these laws?
While it does seem that the DMCA hasn't polluted Canada's legal system yet, it is far from a closed issue. Canada has a stiff media levy, and is considering upping the rate. Moreover, this levy is specifically intended as a levy for making private copies of the music a consumer has already purchased, trampling the notion of archival copies (not to mention fining people for backing up their own data on CDR).
The CCFDA is working with Industry and Heritage Canada to stop the newly purposed levy that would see the cost of a pack of 100 blank CD soar 181 per cent. Currently, a pack of 100 blank CDs includes a levy of $21 on a retail price that averages $50. If the Canadian Private Copying Collective's (CPCC) proposal is approved, this levy could average $59 per pack at an average retail price of $88 plus tax, said Diane Brisebois, president and CEO of the Retail Council of Canada.
with no copy protection... Apple has now done that.
I'm under the impression, given statements like "you can play it on up to 3 computers", that it is DRM'd. This means no playing it in your car CDP or on your stereo.
Block inbound SMTP from AOL. Set a cutoff date, nofity all of your users, and stand firm. The further they get down this path the more accepted this practice will become. Once this practice becomes accepted, it is entirely likely that other protocols will begin being blocked for other reasons (suppose the RIAA suggests that AOL block all connections to non-commercial HTTP servers to avoid piracy liability).
It is not unreasonable to see this as the first step in the stratification of the Internet into corporate sanctioned, generally accessible servers and cordoned-off slums. This is every bit as dangerous, if not more so, than government censorship.
Note: it will anger your users. I know, because it angered my users. The biggest problem was that they did not understand why. I've posted a FAQ to address their most common questions.
If we give an inch now, later they will take a yard. Better to endure a little pain now than to try to stand against the tide in a year or two.
How about a family of operating systems that has managed to capture over 90% of the small computer market?
I actually agree with this 100%. Microsoft got there because they were one of the revolutionaries in the 1980s, when feathered hair and skinny leather piano ties were the rage. But what have they done for me lately?
How about an operating system written as a substitute for massive commercial systems, written initially by one guy, then by a bunch of people collaborating, without direct compensation, via email? (Linux)
How about a system to allow anyone with a computer and a pipe to publish structured hypertext and images for all the world to see? (Mosaic)
How about a system for independent individuals to type to each other in real time? (talk, IM)
How about a system for people without a static IP to share files? (P2P)
How about a system for people to contribute spare CPU cycles to a collective social work? (Distributed.net, SETI@Home, Folding@Home)
The Future:
What's on the edge now that will be huge tomorrow? If I knew that I'd be in angel capital. (speaking of equity, how about online stock trading systems?)
What's on the edge and either hasn't found a niche or isn't sufficiently advanced yet (and may never be)? 3DUIs, Freenet, Complex Adaptive Systems, Face Recognition; and those are less than a cube in the iceberg.
Microsoft's Leach emphasized that this change in positioning doesn't negate that "customer-defined XML schema support is a feature of Pro." On the other hand...
Cool, they've actually appointed a corporate leach. Perhaps that explains why MS Office came out with XML support after it was released in OpenOffice.
You're referencing the basic WinCE shared source license. The one they're talking about is the premium redistribution license, which you can read here.
Use and Restrictions
The program authorizes companies to:
* Develop and distribute derivative works of the source code within Windows CE-based OEM devices.
* Create derivative works of the source code to improve, optimize and debug Windows CE.
* Use the source code as a reference to develop enhancements to Windows CE.
This same thing has been said before, typically with a reference to Sourceforge or Freshmeat. The reply remains the same:
Quantity and quality are not the same thing. Think of Sourceforge as the primordial ooze from which life is occasionally sparked. Simply putting all the primordial ooze into one pile will not make a great application. Software development is not like the output from an industrial machine, developers are not interchangable. The Mythical Man Month pointed this out decades ago - applying more developers to a project has decreasing returns. Furthermore, in a system where noone is compelled to work, the belief that you can increase focused development by reducing the number of projects is dubious at best.
So that explains why limiting the number of projects in Sourceforge would probably not help. What abou t the flipside? Is there any benefit to the large number of projects? I would argue yes. While there may be some projects which never released anything (I think I created at least one of these), primordial ooze is not supposed to be pretty.
I worked on CipherCore for a few months, then realized that cryptography is hard and decided to move on. While it is not a secure crypto system, and should not be used, I still occasionally get email from people who are poking through the source code to see how the JCE (Java Cryptography Extension) works. While CipherCore is not a good product, it might help some more motivated developer to build something similar and better, or at least to get his or her feet wet.
Reducing the number of projects would only reduce the volume of primordial ooze. It is not logically consistent to believe that it would necessarily lead to more focused development on the remaining projects, but it is a certainty that it would reduce the amount of example code (good, bad, and ugly) available to experience.
While I agree that the congressman's comments seem to reflect poor judgement in terms of how to phase in punishment, is it not appropriate to begin enforcing the law against those who break it? When ideas like the DMCA or SSSCA (CBDTPA) are proposed I am offended because they punish me for things that I do not do. Isn't enforcing the existing law exactly the right approach?
As a counter-proposal how about this:
Step 1: For six months, send letters to violators stating that the executive branch will begin enforcing the law.
Step 2: For six months, enforce the law lightly. Give a few kids a week a summons and (if found guilty of willful infringement) probation.
Step 3: Gradually increase the punishment to the natural level for the law in question. Start adding public service and small fines, and gradually ramp up.
This is much the way that traffic law enforcement devices with cameras are phased in, and as a person living in a city with red-light enforcement traffic cameras, I have seen it work.
If you were moving to a new schema you would have to write scripts to handle the RDB to RDB import right? Same thing here. You set up the mappings from the existing data structure to the new data structure, and let it rip.
The big difference though is that if you're going RDB to RDB and using an OO front end, you have to do the RDB to RDB map for the one time import and the O-R map for continued use (including maintenance). If you're migrating to prevalence, you write the O-R mapping for one time use, and then you're done. It should wind up being less expensive to migrate to prevalence unless the schema you're retiring is extremely hard to O-R map and the new schema is extremely easy.
Mind you, I'm still a died in the wool RDBMS advocate myself, but a friend of mine who is a prevalence true believer has explained a lot of the upside to me.
Slashdot Mirror
Totals for 2.4.18 debian woody:
. 4. 18$ find * -type f -name '*\.[ch]' | wcu rce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -2000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -3000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -4000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -5000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -6000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -7000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | head -8000 | tail -1000` | grep totalr ce-2.4. 18$ wc `find * -type f -name '*\.[ch]' | tail -293` | grep totalr ce-2.4. 18$ cd kernel/o urce-2.4. 18/kernel$ wc `find * -type f -name '*\.[ch]'` | grep total
Grand total all directories: 3,704,994 (0.0022%)
Total in kernel directory only: 14,058 (0.57%)
So if they're talking specifically of the kernel directory, and if the 80 lines shown are all there are, they're claiming that their decline from $200+ million per year to $60 million per year is entirely due to about one half of one percent of the code.
Off Topic: anyone know a better way than the following to handle large numbers of arguments?
bob@swirl:/usr/local/src/kernel/kernel-source-2
8293 8293 221144
bob@swirl:/usr/local/src/kernel/kernel-so
324215 1061875 8471017 total
bob@swirl:/usr/local/src/kernel/kernel-sou
551027 1776771 15043435 total
bob@swirl:/usr/local/src/kernel/kernel-sou
631425 2131416 18512146 total
bob@swirl:/usr/local/src/kernel/kernel-sou
965346 3486333 29067291 total
bob@swirl:/usr/local/src/kernel/kernel-sou
581813 2316216 18338130 total
bob@swirl:/usr/local/src/kernel/kernel-sou
134261 511247 4140371 total
bob@swirl:/usr/local/src/kernel/kernel-sou
122814 502081 3903002 total
bob@swirl:/usr/local/src/kernel/kernel-sou
222183 806751 6445312 total
bob@swirl:/usr/local/src/kernel/kernel-sou
171910 516407 4303860 total
bob@swirl:/usr/local/src/kernel/kernel-sou
bob@swirl:/usr/local/src/kernel/kernel-s
14058 43823 343800 total
I somewhat read the NDA and all I saw was stuff about code.
Section 3, like most NDAs, claims everything including but not limited to the word "Pella" etched on SCO's office windows.
3. Definition of Confidential Information. " Confidential Information" means any and all data, technology, research, inventions, intellectual property, trade secrets, know how, computer programs, source code, file names, file trees or extensions, works of authorship, products, processes, methods, customer names, plans, forecasts, prices, business information, financial information, and other information shown or relayed by SCO to RECIPIENT on _______________________ [date].
It is amusing (in the tragic/comic sense) to me that what was originally designed to protect honorable companies from dishonorable people has become a weapon wielded by dishonorable companies. This looks like a pretty standard NDA, and a lot less onerous than most software EULAs (even those from supposedly honorable companies like Ximian).
My favorite piece:
11. Waiver. Any waiver of, or promise not to enforce, any right under this Agreement shall not be enforceable unless evidenced by a writing signed by the Party making said waiver or promise.
Translation - when the blow-dried business weenie / marketroid says, "Oh don't worry about this, it's just for our protection, we would only ever use it defensively." - he's lying. By the time the ink is dry there will be someone checking SourceForge, the USPTO, and the Library of Congress for your name to see what they can hijack.
1: Synchronization.
This is slow. Really slow.
I think your test was flawed. Please see here for a test showing one popular usage of synchronization (lazy initializers). One million cycles takes about 30 milliseconds longer (a hair more than twice as long) for the synchronized version as for the fastsynched (effectively unsynchronized after the first call) version. Unsynchronized actually came out slower, showing that compiler behaviour is probably a bigger issue.
30 milliseconds per one million cycles. The extra 6 lines of code (and their extra defect opportunities) cannot be justified in any but the most extreme performance optimization situation.
If they cracked it, this would be very illegal.
Agreed that this is the most likely outcome in the current legal system. Should it not be the case though, that this action would be protected by the same laws that protect people performing the Heimlich Maneuver?
it appears this academic has missed a crucial piece of the equation in these modern times:
Blatent Product Placement
Did you read the article?
Ms Clayton, who is a movie director and screenwriting lecturer for the University of London and the British Film Council, was commissioned by diet Coke [my emphasis].
I'd rather say she got it exactly right, starting with the product placement in the article about the study about movies.
GoboLinux Rethinks The Linux Filesystem
Filesystem Hierarchy Standard has included the BSDs since 1995 - it's more than just Linux. I'm all in favor of questioning the assumptions to avoid getting caught on a local maxima, and I think this one is a dead end. FHS has some historical baggage, but it also has the strength of years of tempering.
Should samples be protected by copyright, or should artists/musicians have the right to manipulate the old into the new?
You're goddamm skippy they should be. If they want music to enter the public domain, let them fight the psychotic duration of copyright.
Even if M$ manages to get DRM out there, how riddled with holes will it be?
The problem is not whether it works - we all know that DRM is technically impossible (analog hole). The problem is that combined with the DMCA, DRM makes fair use illegal. If Passport were being used for copyright protection, it would be a federal crime to report this security vulnerability.
I use the company's name or the product for my email.
This is what I do now too. But with dictionary spam becoming more popular (my news@ alias gets hit more than any other address), I'm guessing it won't be long before they start hitting hertz@, ebay@, amazon@, etc.
What happens when the customer orders something from Amazon - the purchase confirmation email comes from a non-human address.
This has been my single opposition to challenge response from the first time I heard of it. As I was reading your post, I hit on an idea that borrows from an idea posted by another user.
Put challenge response on my primary email address.
Rotate my non-challenge response email address every 30 days.
So Amazon uses wog23t5s@traxel.com, while all my friends use my real address.
Toss in some accepted domains if you want to permanently accept from *.register.com (reversible, of course), and I think it might be a flexible, functional system.
e.g. tulip blubs sell well, but not like they used to.
I'm sure I'm reiterating what a few dozen other posts have already pointed out, but Holland's economic debacle was primarily due to the fact that the value of a tulip is almost entirely based on end-user utility. Information science, on the other hand, deals with the refinement of information. Much of this ability to process information is directed at the acquisition of additional wealth. That is, information science has far higher objective economic value because it directly results in the ability to harvest more wealth.
I would choose to politely disagree. Look at Canada as an example: We are right next to the US, but do we see any effect from these laws?
While it does seem that the DMCA hasn't polluted Canada's legal system yet, it is far from a closed issue. Canada has a stiff media levy, and is considering upping the rate. Moreover, this levy is specifically intended as a levy for making private copies of the music a consumer has already purchased, trampling the notion of archival copies (not to mention fining people for backing up their own data on CDR).
Here's the proposed "Private Copying" legislation.
The CCFDA is working with Industry and Heritage Canada to stop the newly purposed levy that would see the cost of a pack of 100 blank CD soar 181 per cent. Currently, a pack of 100 blank CDs includes a levy of $21 on a retail price that averages $50. If the Canadian Private Copying Collective's (CPCC) proposal is approved, this levy could average $59 per pack at an average retail price of $88 plus tax, said Diane Brisebois, president and CEO of the Retail Council of Canada.
with no copy protection ... Apple has now done that.
I'm under the impression, given statements like "you can play it on up to 3 computers", that it is DRM'd. This means no playing it in your car CDP or on your stereo.
By blocking the dynamic DSL IPs, AOL has removed itself from the global SMTP network.
Block inbound SMTP from AOL. Set a cutoff date, nofity all of your users, and stand firm. The further they get down this path the more accepted this practice will become. Once this practice becomes accepted, it is entirely likely that other protocols will begin being blocked for other reasons (suppose the RIAA suggests that AOL block all connections to non-commercial HTTP servers to avoid piracy liability).
It is not unreasonable to see this as the first step in the stratification of the Internet into corporate sanctioned, generally accessible servers and cordoned-off slums. This is every bit as dangerous, if not more so, than government censorship.
Note: it will anger your users. I know, because it angered my users. The biggest problem was that they did not understand why. I've posted a FAQ to address their most common questions.
If we give an inch now, later they will take a yard. Better to endure a little pain now than to try to stand against the tide in a year or two.
So has Jeff applied for a patent on this yet?
(I'm gonna get flamed for this...)
How about a family of operating systems that has managed to capture over 90% of the small computer market?
I actually agree with this 100%. Microsoft got there because they were one of the revolutionaries in the 1980s, when feathered hair and skinny leather piano ties were the rage. But what have they done for me lately?
Recent History:
How about an operating system written as a substitute for massive commercial systems, written initially by one guy, then by a bunch of people collaborating, without direct compensation, via email? (Linux)
How about a system to allow anyone with a computer and a pipe to publish structured hypertext and images for all the world to see? (Mosaic)
How about a system for independent individuals to type to each other in real time? (talk, IM)
How about a system for people without a static IP to share files? (P2P)
How about a system for people to contribute spare CPU cycles to a collective social work? (Distributed.net, SETI@Home, Folding@Home)
The Future:
What's on the edge now that will be huge tomorrow? If I knew that I'd be in angel capital. (speaking of equity, how about online stock trading systems?)
What's on the edge and either hasn't found a niche or isn't sufficiently advanced yet (and may never be)? 3DUIs, Freenet, Complex Adaptive Systems, Face Recognition; and those are less than a cube in the iceberg.
Microsoft's Leach emphasized that this change in positioning doesn't negate that "customer-defined XML schema support is a feature of Pro." On the other hand...
Cool, they've actually appointed a corporate leach. Perhaps that explains why MS Office came out with XML support after it was released in OpenOffice.
You're referencing the basic WinCE shared source license. The one they're talking about is the premium redistribution license, which you can read here.
Use and Restrictions
The program authorizes companies to:
* Develop and distribute derivative works of the source code within Windows CE-based OEM devices.
* Create derivative works of the source code to improve, optimize and debug Windows CE.
* Use the source code as a reference to develop enhancements to Windows CE.
This same thing has been said before, typically with a reference to Sourceforge or Freshmeat. The reply remains the same:
Quantity and quality are not the same thing. Think of Sourceforge as the primordial ooze from which life is occasionally sparked. Simply putting all the primordial ooze into one pile will not make a great application. Software development is not like the output from an industrial machine, developers are not interchangable. The Mythical Man Month pointed this out decades ago - applying more developers to a project has decreasing returns. Furthermore, in a system where noone is compelled to work, the belief that you can increase focused development by reducing the number of projects is dubious at best.
So that explains why limiting the number of projects in Sourceforge would probably not help. What abou t the flipside? Is there any benefit to the large number of projects? I would argue yes. While there may be some projects which never released anything (I think I created at least one of these), primordial ooze is not supposed to be pretty.
I worked on CipherCore for a few months, then realized that cryptography is hard and decided to move on. While it is not a secure crypto system, and should not be used, I still occasionally get email from people who are poking through the source code to see how the JCE (Java Cryptography Extension) works. While CipherCore is not a good product, it might help some more motivated developer to build something similar and better, or at least to get his or her feet wet.
Reducing the number of projects would only reduce the volume of primordial ooze. It is not logically consistent to believe that it would necessarily lead to more focused development on the remaining projects, but it is a certainty that it would reduce the amount of example code (good, bad, and ugly) available to experience.
While I agree that the congressman's comments seem to reflect poor judgement in terms of how to phase in punishment, is it not appropriate to begin enforcing the law against those who break it? When ideas like the DMCA or SSSCA (CBDTPA) are proposed I am offended because they punish me for things that I do not do. Isn't enforcing the existing law exactly the right approach?
As a counter-proposal how about this:
Step 1:
For six months, send letters to violators stating that the executive branch will begin enforcing the law.
Step 2:
For six months, enforce the law lightly. Give a few kids a week a summons and (if found guilty of willful infringement) probation.
Step 3:
Gradually increase the punishment to the natural level for the law in question. Start adding public service and small fines, and gradually ramp up.
This is much the way that traffic law enforcement devices with cameras are phased in, and as a person living in a city with red-light enforcement traffic cameras, I have seen it work.
the necessity for rational grounding for [DRM] technologies
hahahahahhaahhahaha
Rational Grounding:
1. The only possible solution is to not give information to people you do not trust with it.
2. Once you accept item 1, there is no item 2.
If you were moving to a new schema you would have to write scripts to handle the RDB to RDB import right? Same thing here. You set up the mappings from the existing data structure to the new data structure, and let it rip.
The big difference though is that if you're going RDB to RDB and using an OO front end, you have to do the RDB to RDB map for the one time import and the O-R map for continued use (including maintenance). If you're migrating to prevalence, you write the O-R mapping for one time use, and then you're done. It should wind up being less expensive to migrate to prevalence unless the schema you're retiring is extremely hard to O-R map and the new schema is extremely easy.
Mind you, I'm still a died in the wool RDBMS advocate myself, but a friend of mine who is a prevalence true believer has explained a lot of the upside to me.