I think the difference that he's pointing out is that this License appears to require a contract between the distributor and the third party.
The GPL on the other hand is a contract beween the copyright holder and all the people taking advantage of the rights granted by the GPL --- there are no contracts needed between the users and distributors.
Only when you take advantage of the rights granted by the GPL (modification and/or distribution) is there a need for a contract to exist. So there is no contract needed for you to download the software, and use it.
The person that owns the server you downloaded it from would be bound by the GPL, because they are distributing (unless they're the copyright holder), but you would not, until you modify or distribute the code.
If Microsoft sold fish, this deal would be equivalent to being taught how to fill out the MS order form for your next fish consignement, and being given a discount on your first order (but still paying enough to cover the cost of the training and the fish).
If they kept their money, and spent it on training people as GNU/Linux trainers, they would have taught themselves to fish, and would soon be in a possition to export their fishing skills to the world, including the USA.
I know which I'd prefer if I were Mexican.
And of course, this is in a free software ocean that cannot be over-fished. It actually becomes more productive, the more people fish it.
The problem, as I understand it, is that the Univerity of Washinton have an anal attitude to allowing the redistribution of binaries, hence the source only distribution by Debian.
Of course this puts pine into the "non-free" section, so pine is not part of the official Debian distribution, and as such will not get onto Debian CDs until UoW changes their license.
Personally, I'd recomment finding a MUA that's got a sensible licensing policy instead, so you don't have to put up with this nonsense
Sid is, and will continue to be the code name for the unstable "release", on the basis that sid is the deranged kid next door, with a tendency to break all your toys (gedit?)
When the new testing is assembled, it will be assigned a new name, but I've not been keeping track of suggestions for the next one.
In case you're wondering, Bruce Perens (ex. Debian Project Leader, and Pixar employee) started the Toy Story codename thing, and we've already had rex, bo, hamm, slink, buzz, potato and now woody (I think that's the lot)
Sorry for the double post, but I just noticed a quote from that article you mentioned about the Navy running NT, where thet are boasting about how reliable NT is:
Bryan Scurry, the SPAWAR test director, says that during the test the NT-based GCCS-M ran "for more than 1,000 hours, and it passed with an operational availability of over 95 percent. In a couple of instances, that availability hit 98 percent."
98 percent! that's pathetic.
If you were assuming that a battle might take 24 hours from start to finish, you're going to find that you're going to be blind for about half an hour of that battle --- how anyone could quote that as though it were a good thing is mind boggling.
And that's without considering that the more normal 95% is 72 minutes down time in the average day. Enough time to miss the whole of the Perl Harbour attack IIRC.
Impoverished hacker sells his time to Unenlightened corp.
Corp tells hacker to add some feature to a Free Software project.
Hacker does so, assuming that he's going to release his enhancement once written.
Corp then tries to stop him publishing, so they can go proprietary, and make money (largely out of other people's work).
If the code in question is under the BSD license, the hacker has the freedom to get shafted by his employer, and the users have the freedom to get all the normal grief of being tied into a proprietary product, which then gets dropped, or is not properly supported.
If it's a modification to a GPLed work. the employer is obliged to do what the hacker wanted to do in the first place, or discard all his work.
Not that this affects me too much, since I work for myself, but I've certainly used the "I'd love to sign the rights over to you, but this software's GPLed so I'm not allowed to" line to shut up some of the people that have tried to make my work proprietary in the past.
Just because I come up with an idea before you, or more likely just because I came up with an idea after you, but ran to the patent office faster, why should I be allowed to stop you thinking of it too?
The point of the patent system is to encourage industries that have to put in massive investment, and then have a long time to wait before seeing a return, during which time they need a government sanctioned monopoly to avoid being ripped off.
Software tends to monopoly too easily already, without any more help from Government.
Software development requires a few hundred pounds for the machine, and you can get paid enough doing support to leave you with enough spare time for development.
Also, if you want to keep your methods secret, you simply keep the source to yourself.
This has been repeatedly demonstrated in our industry.
Patents are not needed here. The only people that want them are the people that have already arrived, and they want them so they can raise the drawbridge on the next generation of newcomers.
... just ISPs with really, really bad(missing entirely) DNS.
I agree, DNS facism can be quite useful, but unfortunately one such incompetent ISP is BTOpenwoe, who are responsible for a massive number of UK dial-up & DSL lines.
Check this out:
sheikh:~$ host 213.122.0.74
Name: host213-122-0-74.in-addr.btopenworld.com
Address: 213.122.0.74
sheikh:~$ host host213-122-0-74.in-addr.btopenworld.com
host213-122-0-74.in-addr.btopenworld.com does not exist (Authoritative answer)
sheikh:~$
and they've been told about repeatedly, bat apparently cannot be arsed to fix it.
Suggest that he uses one of the several authentication tricks, such as POP before SMTP (where the server will only accept relay mail from IP addresses that have had a successful POP authentication in the last 5 minutes) to limit the relay.
If he still ignores you, Submit his IP to ordb.com --- at least that way I won't have to see the spam that evenually starts pouring through his server.
OK, so the patent looks like total nonsense, and they're probably hoping to bully people into settling, but just think what would happen if the won.
It could pretty much shut down the Web in the USA, or at least severely disrupt it, by forcing people to pay BT tax, or come up with some disgusting work-around.
Of course, this would not apply to most of the rest of the world, because we still haven't quite given into the lobbying of the WTO and USPO types that want the rest of the world to adopt US patent silliness.
So briefly we'd have BT holding the USA to ransom.
Pretty soon after that the public outcry against the USPO might finally reach the ears of enough US politicians, to result in the USPO being beaten up in the manner that they've clearly deserved for years.
The final result being that people would have their eyes opened to the stupidity of allowing patents on software and business methods.
So, join the "BT for Web supremacy" campaign today!
Ask him "What if NSI use their warped Terms & Conditions to arbritrarily steal our domain, and give it to a friend of theirs?"
Personally, I prefer being with a registrar who states that I own the domain, and they're just looking after it for me, rather than the other way round.
The only way I ever got them to do anything was by listening to their phone system for an hour or two (I'm not joking, and that was at international phone rates, several times over the years) until someone picked up the phone, and tried to do the Dogbert-esque "How may I disconnect you" thing.
With a just a little bit more persistence they would finally react.
I transfered to gandi (which went totally smoothly, and only took about 3 days to happen, and costs about half what NSI charge) and have never had a problem with them. They have a web interface that works, they have a mail system that means that a human gets to see your queries within less than a day, and replies --- In short, they do a proper job.
Why the bloody hell NSI are still allowed to have anything to do with domain administration is totally beyond me. It's about time that someone brought a class action against them for wasting tens of thousands of hours of other peoples time.
Further examples of their uselessness:
They took to returning every request I sent, saying that it must come from one of the domain contacts. The only address I ever used was listed as all three contacts. The bounce would helpfully inform me that the domain contacts had been informed --- I never saw a single such message --- several phonecalls did nothing to rectify this
They switched my account to PGP authentication at some point, without associating a key with the account (or at least, no key I've ever used. I tried them all), thus locking me out.
They also managed to look after a mail they sent to me for 2 months, in a queue on their web server, before delivering it to their mail server, and then (about 5 seconds later) finally to me.
Use gandi, use anyone --- just run screaming from NSI and your problems will be solved.
He ask US$ 25 for hour and send an invoice for 6,300 US$, for 219 hours in one month and half.
I promise him to be paid.
So, let me get this straight. You are saying that he presented you with an invoice, and that you accepted the invoice as being justified, and promised to pay him. Is that correct?
If so, why have you not paid him?
Then, he send a 6,300 US$ invoice and all the software he wrote was available for free to everyone!
...Sorry. I don't pay in these conditions.
So you're saying that you refuse to pay him because he published the work that you paid him to do?
Are you really trying to say that you didn't realise that a Free Software programmer was going to do this?
If you are saying that as his employer, you should own all the code he wrote in your employ, then that is possibly true (once you pay him), but that does not stop him from publishing the code, because it was in the form of modifications to an existing GPL/QPL code base.
It doesn't matter that he is the original author. When he works for you, he is either acting as himself, in which case he owns the new code, or he is acting as FT, in which case FT may own the new code, but it is bound by the GPL/QPL to license the modifications in the same way.
Either way, he gets the right to publish the whole work when he reverts to acting as himself on the several hours a day that he didn't bill you for.
Re:why use a dvd? just setup a package mirror
on
Debian On DVD
·
· Score: 1
or even simpler, and more efficient, just use squid.
If you add a line something like this to your/etc/squid.conf:
refresh_pattern debian.org/.*\.deb$ 129600 100% 129600
so that it keeps.debs for a long time (about 3 months, or until it runs out of disk) then set all your machines to use the cache (http_proxy in the environment, or a setting in apt.conf).
You still get the benefit of downloading any particular package file only once, but you don't waste bandwidth on downloading packages nobody installs, you don't have to have as much disk space, and there is no admin overhead after telling squid how much disk to use.
Debian based in Germany?
on
Debian On DVD
·
· Score: 1
you seem to have got it mixed up with SuSE.
Debian is based just off the coast of Greenland, on average;-)
Also, Debian's version is currently 2.2, soon to be 3.0, so I'm not sure where you got 5.x from.
IBM, the great supporter of Linux, is also engaged in obtaining many software patents.
If you are being generous to IBM, and other people like the that stand on both sides of the Patent vs. Free Software divide, you can believe that they are doing this in order to defend themselves from the patents of others.
On that basis, you cannot insist that they simply give up Patents on technologies that are to be included in a future standard.
That being the case, perhaps it would be better to make the RAND process define "reasonable" to be that they may only ask for a license fee from people who first attempt to extract a license fee from them, for any patent, whether related or not.
In other words, the patents accepted into the RAND process can only be used to extract fees as a tit-for-tat response. Otherwise, they should be freely usable.
Let's say you wrote a open source program... Aren't you entitled to use the DMCA against them?
The short answer is "No".
Firstly, in order to qualify as Free Software, or Open Source, the license must grant rights that would include people using the code for things of which I don't approve.
Secondly, I don't live or work in the USA, and I'm not a US citizen. I live in the UK, and so am subject to UK law, which doesn't as yet include a DMCA-like law.
supposed to be uncopyable...
You are arguing that lack of technical competence of a solution can be compensated for by resorting to law?
In an equivalent situation, if a company was selling door locks that could be easily picked, would it make sense to stop people talking about that fact, in order to protect the lock company's market? Would you be happy to have a defective door lock on your house, as long as nobody was allowed to talk about it? Personally, I'd prefer for a competent locksmith to be allowed to publish an analysis of what is wrong with the lock, so that the public could make an informed purchasing decision.
You seem to think that making the discussion illegal would stop criminals discussing it, but they are criminals.
You also seem to think that because a tool can be applied to criminal activities, that all users of that tool must have criminal intent. How about knives?
Just because bin laden ordered his troops from Afghanistan is he innocent...
I think you'll find that conspiracy to mass-murder is illegal, even in Afghanistan.
Dmitry Sklyarov committed no crime in Russia.
Compairing mass-murder to reverse-engineering is a bit much, but I suppose I said totalitarian first, so I'll let it pass.
What he did isn't against the law in the place where he did it.
It wouldn't have been against the law in most of the world.
I think the use of the word totalitarian was perfectly appropriate, if you work on the assumption that a "totalitarian corporation" is one that attempts to maintain absolute control over the circumstances in which their product is used.
It seems entirely likely that if the DMCA is not overturned, that it will eventually be used to prevent people from quoting e-books in reviews, if the review is bad --- i.e. suppression of negative comment.
Just because Adobe and other DMCA supporters don't control all aspects of life in the USA, doesn't mean that they cannot be considered to have totalitarian ambitions towards those aspects they do control IMO. Using the FBI, as attack dogs, against law-abiding foreign nationals just goes to reinforce the impression.
The DMCA goes beyond traditional copyright law, and strips people in the USA of their rights of fair use. We still have those rights in the rest of the world, so writing software that allows us to excercise those rights is totally legitimate.
You seem to be unaware that copyright law allows for a number of circumstances in which copying is permitted, for instance, here is a description of some forms of copying allowed WRT computer programs.
So, in the free world (i.e. places like Russia, where the DMCA is not law) you are entirely within your rights to write, and use such software, just as you were in the USA prior to the DMCA.
It strikes me that the USA are retroactively enforcing this law outside its jurisdiction.
From where I'm standing, this looks pretty similar to a totalitarian regime arresting foreign journalists for criticising the regime. The only difference is that the authorities are acting on behalf of totalitarian corporations, rather than a dictator.
In fact there is no clause in the GPL that states that you must give the modified source to the original author if you redistribute a modified version.
The correct analogy would be:
---
An author publishes a book on fishing, and allows the people that get the book to modify, and republish the book without reference to the original author (pretty generous so far)
The only restriction being that if you choose to distribute your improved version of the book, the people that get a copy also get the same right to modify it that you just took advantage of in order to create your version. (still seems pretty fair to me)
---
So there's no obligation to the original author, and you don't have to send them a copy.
If you don't like having to dilute the rights you have over the page you added, simple write the whole book from scratch.
The idea that adding a single page to the book should allow you to seize greater control of the resulting work than the original author had in the original seems a little churlish to me.
Slashdot Mirror
I think the difference that he's pointing out is that this License appears to require a contract between the distributor and the third party.
The GPL on the other hand is a contract beween the copyright holder and all the people taking advantage of the rights granted by the GPL --- there are no contracts needed between the users and distributors.
Only when you take advantage of the rights granted by the GPL (modification and/or distribution) is there a need for a contract to exist. So there is no contract needed for you to download the software, and use it.
The person that owns the server you downloaded it from would be bound by the GPL, because they are distributing (unless they're the copyright holder), but you would not, until you modify or distribute the code.
If Microsoft sold fish, this deal would be equivalent to being taught how to fill out the MS order form for your next fish consignement, and being given a discount on your first order (but still paying enough to cover the cost of the training and the fish).
If they kept their money, and spent it on training people as GNU/Linux trainers, they would have taught themselves to fish, and would soon be in a possition to export their fishing skills to the world, including the USA.
I know which I'd prefer if I were Mexican.
And of course, this is in a free software ocean that cannot be over-fished. It actually becomes more productive, the more people fish it.
just type:
apt-get install [PACKAGE_NAME]
answer zero or maybe more configuration questions, and you're done.
If you prefer pointy-clicky things, there are several that allow you to acheive the same result with a few mouse clicks.
The problem, as I understand it, is that the Univerity of Washinton have an anal attitude to allowing the redistribution of binaries, hence the source only distribution by Debian.
Of course this puts pine into the "non-free" section, so pine is not part of the official Debian distribution, and as such will not get onto Debian CDs until UoW changes their license.
Personally, I'd recomment finding a MUA that's got a sensible licensing policy instead, so you don't have to put up with this nonsense
Coming up after Woody is Sid.
No, sid will not be released.
Sid is, and will continue to be the code name for the unstable "release", on the basis that sid is the deranged kid next door, with a tendency to break all your toys (gedit?)
When the new testing is assembled, it will be assigned a new name, but I've not been keeping track of suggestions for the next one.
In case you're wondering, Bruce Perens (ex. Debian Project Leader, and Pixar employee) started the Toy Story codename thing, and we've already had rex, bo, hamm, slink, buzz, potato and now woody (I think that's the lot)
That'll be fixed in the next release of release-manager.deb
Sorry for the double post, but I just noticed a quote from that article you mentioned about the Navy running NT, where thet are boasting about how reliable NT is:
Bryan Scurry, the SPAWAR test director, says that during the test the NT-based GCCS-M ran "for more than 1,000 hours, and it passed with an operational availability of over 95 percent. In a couple of instances, that availability hit 98 percent."
98 percent! that's pathetic.
If you were assuming that a battle might take 24 hours from start to finish, you're going to find that you're going to be blind for about half an hour of that battle --- how anyone could quote that as though it were a good thing is mind boggling.
And that's without considering that the more normal 95% is 72 minutes down time in the average day. Enough time to miss the whole of the Perl Harbour attack IIRC.
Actually the Navy Battleships computer system is running WinNT
... Oh Fuck! The bloody thing crashed! We're dead meat!"
Yeah, gives a whole new meaning to Blue Screen of Death when your warship's network crashes.
"Captain, I've got incoming missiles on a bearing of
Not that this affects me too much, since I work for myself, but I've certainly used the "I'd love to sign the rights over to you, but this software's GPLed so I'm not allowed to" line to shut up some of the people that have tried to make my work proprietary in the past.
Just because I come up with an idea before you, or more likely just because I came up with an idea after you, but ran to the patent office faster, why should I be allowed to stop you thinking of it too?
The point of the patent system is to encourage industries that have to put in massive investment, and then have a long time to wait before seeing a return, during which time they need a government sanctioned monopoly to avoid being ripped off.
Software tends to monopoly too easily already, without any more help from Government.
Software development requires a few hundred pounds for the machine, and you can get paid enough doing support to leave you with enough spare time for development.
Also, if you want to keep your methods secret, you simply keep the source to yourself.
This has been repeatedly demonstrated in our industry.
Patents are not needed here. The only people that want them are the people that have already arrived, and they want them so they can raise the drawbridge on the next generation of newcomers.
... just ISPs with really, really bad(missing entirely) DNS.
I agree, DNS facism can be quite useful, but unfortunately one such incompetent ISP is BTOpenwoe, who are responsible for a massive number of UK dial-up & DSL lines.
Check this out:
sheikh:~$ host 213.122.0.74
Name: host213-122-0-74.in-addr.btopenworld.com
Address: 213.122.0.74
sheikh:~$ host host213-122-0-74.in-addr.btopenworld.com
host213-122-0-74.in-addr.btopenworld.com does not exist (Authoritative answer)
sheikh:~$
and they've been told about repeatedly, bat apparently cannot be arsed to fix it.
Suggest that he uses one of the several authentication tricks, such as POP before SMTP (where the server will only accept relay mail from IP addresses that have had a successful POP authentication in the last 5 minutes) to limit the relay.
If he still ignores you, Submit his IP to ordb.com --- at least that way I won't have to see the spam that evenually starts pouring through his server.
OK, so the patent looks like total nonsense, and they're probably hoping to bully people into settling, but just think what would happen if the won.
It could pretty much shut down the Web in the USA, or at least severely disrupt it, by forcing people to pay BT tax, or come up with some disgusting work-around.
Of course, this would not apply to most of the rest of the world, because we still haven't quite given into the lobbying of the WTO and USPO types that want the rest of the world to adopt US patent silliness.
So briefly we'd have BT holding the USA to ransom.
Pretty soon after that the public outcry against the USPO might finally reach the ears of enough US politicians, to result in the USPO being beaten up in the manner that they've clearly deserved for years.
The final result being that people would have their eyes opened to the stupidity of allowing patents on software and business methods.
So, join the "BT for Web supremacy" campaign today!
all is fair in love and war, my friend.
What about the Geneva convention?
Oh, silly me, I was forgetting, that's not something that USA feels compelled to comply with any more, is it?
So. Software is free. No money to you....
Bzzzt! Wrong.
Yesterday in 5 hours I made enough to pay my mortgage and all my monthly bills for the month.
How much of that was license fees? Answer: None.
It was for supporting Free Software --- My entire income comes from working with Free Software in one capacity or another.
I manage to charge enough of my time to: pay my mortgage, pay my bills, fund the UK Debian mirror, go on several skiing holidays a year, etc. etc.
OK, I'm never going to be a millionaire, but that's not exactly unusual for a programmer.
The myth that it is imposible to make a living from Free Software is patently absurd. Please don't attempt to peddle it again.
Thanks, Phil.
Ask him "What if NSI use their warped Terms & Conditions to arbritrarily steal our domain, and give it to a friend of theirs?"
Personally, I prefer being with a registrar who states that I own the domain, and they're just looking after it for me, rather than the other way round.
The only way I ever got them to do anything was by listening to their phone system for an hour or two (I'm not joking, and that was at international phone rates, several times over the years) until someone picked up the phone, and tried to do the Dogbert-esque "How may I disconnect you" thing.
With a just a little bit more persistence they would finally react.
I transfered to gandi (which went totally smoothly, and only took about 3 days to happen, and costs about half what NSI charge) and have never had a problem with them. They have a web interface that works, they have a mail system that means that a human gets to see your queries within less than a day, and replies --- In short, they do a proper job.
Why the bloody hell NSI are still allowed to have anything to do with domain administration is totally beyond me. It's about time that someone brought a class action against them for wasting tens of thousands of hours of other peoples time.
Further examples of their uselessness:
Use gandi, use anyone --- just run screaming from NSI and your problems will be solved.
He ask US$ 25 for hour and send an invoice for 6,300 US$, for 219 hours in one month and half.
I promise him to be paid.
So, let me get this straight. You are saying that he presented you with an invoice, and that you accepted the invoice as being justified, and promised to pay him. Is that correct?
If so, why have you not paid him?
Then, he send a 6,300 US$ invoice and all the software he wrote was available for free to everyone!
...Sorry. I don't pay in these conditions.
So you're saying that you refuse to pay him because he published the work that you paid him to do?
Are you really trying to say that you didn't realise that a Free Software programmer was going to do this?
If you are saying that as his employer, you should own all the code he wrote in your employ, then that is possibly true (once you pay him), but that does not stop him from publishing the code, because it was in the form of modifications to an existing GPL/QPL code base.
It doesn't matter that he is the original author. When he works for you, he is either acting as himself, in which case he owns the new code, or he is acting as FT, in which case FT may own the new code, but it is bound by the GPL/QPL to license the modifications in the same way.
Either way, he gets the right to publish the whole work when he reverts to acting as himself on the several hours a day that he didn't bill you for.
or even simpler, and more efficient, just use squid.
/etc/squid.conf:
.debs for a long time (about 3 months, or until it runs out of disk) then set all your machines to use the cache (http_proxy in the environment, or a setting in apt.conf).
If you add a line something like this to your
refresh_pattern debian.org/.*\.deb$ 129600 100% 129600
so that it keeps
You still get the benefit of downloading any particular package file only once, but you don't waste bandwidth on downloading packages nobody installs, you don't have to have as much disk space, and there is no admin overhead after telling squid how much disk to use.
you seem to have got it mixed up with SuSE.
;-)
Debian is based just off the coast of Greenland, on average
Also, Debian's version is currently 2.2, soon to be 3.0, so I'm not sure where you got 5.x from.
IBM, the great supporter of Linux, is also engaged in obtaining many software patents.
If you are being generous to IBM, and other people like the that stand on both sides of the Patent vs. Free Software divide, you can believe that they are doing this in order to defend themselves from the patents of others.
On that basis, you cannot insist that they simply give up Patents on technologies that are to be included in a future standard.
That being the case, perhaps it would be better to make the RAND process define "reasonable" to be that they may only ask for a license fee from people who first attempt to extract a license fee from them, for any patent, whether related or not.
In other words, the patents accepted into the RAND process can only be used to extract fees as a tit-for-tat response. Otherwise, they should be freely usable.
What do you think of this idea?
Let's say you wrote a open source program ... Aren't you entitled to use the DMCA against them?
...
The short answer is "No".
Firstly, in order to qualify as Free Software, or Open Source, the license must grant rights that would include people using the code for things of which I don't approve.
Secondly, I don't live or work in the USA, and I'm not a US citizen. I live in the UK, and so am subject to UK law, which doesn't as yet include a DMCA-like law.
supposed to be uncopyable
You are arguing that lack of technical competence of a solution can be compensated for by resorting to law?
In an equivalent situation, if a company was selling door locks that could be easily picked, would it make sense to stop people talking about that fact, in order to protect the lock company's market? Would you be happy to have a defective door lock on your house, as long as nobody was allowed to talk about it? Personally, I'd prefer for a competent locksmith to be allowed to publish an analysis of what is wrong with the lock, so that the public could make an informed purchasing decision.
You seem to think that making the discussion illegal would stop criminals discussing it, but they are criminals.
You also seem to think that because a tool can be applied to criminal activities, that all users of that tool must have criminal intent. How about knives?
Just because bin laden ordered his troops from Afghanistan is he innocent...
I think you'll find that conspiracy to mass-murder is illegal, even in Afghanistan.
Dmitry Sklyarov committed no crime in Russia.
Compairing mass-murder to reverse-engineering is a bit much, but I suppose I said totalitarian first, so I'll let it pass.
What he did isn't against the law in the place where he did it.
It wouldn't have been against the law in most of the world.
I think the use of the word totalitarian was perfectly appropriate, if you work on the assumption that a "totalitarian corporation" is one that attempts to maintain absolute control over the circumstances in which their product is used.
It seems entirely likely that if the DMCA is not overturned, that it will eventually be used to prevent people from quoting e-books in reviews, if the review is bad --- i.e. suppression of negative comment.
Just because Adobe and other DMCA supporters don't control all aspects of life in the USA, doesn't mean that they cannot be considered to have totalitarian ambitions towards those aspects they do control IMO. Using the FBI, as attack dogs, against law-abiding foreign nationals just goes to reinforce the impression.
The DMCA goes beyond traditional copyright law, and strips people in the USA of their rights of fair use. We still have those rights in the rest of the world, so writing software that allows us to excercise those rights is totally legitimate.
You seem to be unaware that copyright law allows for a number of circumstances in which copying is permitted, for instance, here is a description of some forms of copying allowed WRT computer programs.
So, in the free world (i.e. places like Russia, where the DMCA is not law) you are entirely within your rights to write, and use such software, just as you were in the USA prior to the DMCA.
It strikes me that the USA are retroactively enforcing this law outside its jurisdiction.
From where I'm standing, this looks pretty similar to a totalitarian regime arresting foreign journalists for criticising the regime. The only difference is that the authorities are acting on behalf of totalitarian corporations, rather than a dictator.
In fact there is no clause in the GPL that states that you must give the modified source to the original author if you redistribute a modified version.
The correct analogy would be:
---
An author publishes a book on fishing, and allows the people that get the book to modify, and republish the book without reference to the original author (pretty generous so far)
The only restriction being that if you choose to distribute your improved version of the book, the people that get a copy also get the same right to modify it that you just took advantage of in order to create your version. (still seems pretty fair to me)
---
So there's no obligation to the original author, and you don't have to send them a copy.
If you don't like having to dilute the rights you have over the page you added, simple write the whole book from scratch.
The idea that adding a single page to the book should allow you to seize greater control of the resulting work than the original author had in the original seems a little churlish to me.