The only explanation for RMS agreeing with the analysis I take issue with is that he's being inconsistent. This would not surprise me in the slightest.
You forget one minor detail: there's nothing stopping others from taking the original BSD code and hacking it up to be compatible with your hypothetical MS/TCP, then releasing it freely.
Actually, I argue that M$ using the BSD TCP stack is a feature, not a bug. If they hadn't, we'd be dealing with their buggy networking stack for years to come, and things would be in even a worse mess than they are now.
The key is that true freedom must necessarily include the freedom to do things that piss other people off, so long as you do not actually harm them. This is where RMS's redefinition of "free" falls short: he defines it such that people are no longer free to piss him off. This destroys it as a useful word.
We've run repeated tests, and Hercules under Cygwin is no slower than Hercules under Linux on the same hardware. This surprised me at first, but it's been shown enough times that I accept it.
We could have a pointer to the Cygwin source, true - but that pointer must be to the same site where the software using it is stored. This doesn't do my disk space any good, nor my maintenance overhead...It's simply not worth it to me to mess with.
The distinction is both necessary and important only to the FSF and its minions. The rest of the world has moved on, by conclusively demonstrating that they simply don't share the FSF's concerns.
If you want to honour the GPL, then if you give a copy of the software to a friend on a CD you are required to either maintain a copy of the src for 3 years and provide it to your friend upon request, or ensure that the src is transferred to your friend when you hand them the CD. This is exactly why the open-source package Hercules does not distribute the Cygwin binaries with the Windows version, but instead directs the user to the Cygwin site. We've found that Cygwin is enough of a moving target that it's best to install the version that Hercules was compiled against, and that means we'd have to keep around several different source code packages for years just to satisfy the GPL. In this case, we're actively working on getting away from Cygwin just so we don't have to mess with the installation hassles. If Cygwin were licensed under some other license, even one that required that the user be pointed at a place where he could download the source, we wouldn't have this problem.
The GPL actively inhibits software reuse, one of RMS's original goals for the license. Of course, he never lets goals stand in the way of being simon-pure on ideology...
The requirement you say isn't listed equates to "freedom only applies to non-programmers". This is why I refuse to accept the FSF's redefinition of the word "free" as anything approaching authoritative, and one reason why it has utterly failed in its attempt to get its redefinition widely accepted.
They tried to get positive karma for their political position by wrapping it in the flag, and they've failed. It's time that they recognized this and went on to something that might actually do some good.
This analysis is a long way too simplistic. Even the holy RMS says the BSD license is "free", and yet one of the biggest arguments "free software" zealots raise against it is that nasty eeeeeevil vendors (the "free software" zealots never met a vendor they liked) can "take it private".
Of course, this is not the case. Nobody can take BSD-licensed code out of the freely available world. They can add their own extensions to it, but that's a different story.
Not even RMS says that the GPL is the only "free" license.
Most objections seem to be to the challenge/response mechanism. I'm persuaded that that would only be use in a tiny minority of cases by this system.
A bigger problem is the wide range of prerequisites: Java 1.4, JavaMail, Apache with modssl and mod-auth-external, Postfix 2.1. If you're not running x86 or x86-64, forget it. (Or Solaris, but who runs that?:-)
When I read Just a Geek, I was struck by the similarity between the reaction to your 25 August 2001 WWdN post (A Look Inside My Mind) and the reaction to the Slashdot and Fark stories about my TRON costume. The 26 August 2001 post (My Velouria) was something I could have written, at least in tone if not in details, if I were half the writer you are.
What would you tell the anonymous, average geek who has Internet fame suddenly dropped on him? How would you help him get through the flood of negativity and get to the point where he can ignore the naysayers and just have fun?
Most chromakey technology these days uses a truly hideous green that's less likely to conflict with clothing or props. I don't know which was used in this movie, though.
All of the scenes in TRON inside the computer were shot on a bare black set with the computer imagery filled in later. This was done in 1982. The actors talk about how hard that was in the making-of video in the collector's edition box set.
They shared OS/360 with everyone at the same time. OS/360 wasn't copyrighted, and was distributed in source form - in fact, to install it, you had to start with a small pregenerated system and completely assemble everything from the ground up. OS/360 was open source when open source wasn't cool - it was just the way everybody did things.
The RCA "compatible" mainframe was compatible only at the problem program level. OS programs (supervisor state code) were markedly different because the I/O subsystem wasn't even close to being compatible.
'It's time for a solution to the P2P conflict that pays artists, not lawyers.'
The problem here is that Congress is full of lawyers bent on doing things that amount to full-employment programs for lawyers and accountants. A program like this one that would have the effect of reducing lawsuits has no chance at all.
We complain loudly about conflict of interest by legislators and regulators, while ignoring the biggest one of all: that lawyers write laws. I believe that being a practicing attorney should bar one from being eligible to serve in Congress in much the same way as being an insurance company executive, as a practical matter if not a legal one, bars one from serving as an insurance regulator.
It's still a magic bullet against one particular postal spammer at a time, which is reasonable when the sender is paying for the mail. I have yet to have single mailer continue to send me mail after being hit with that form. The penalties are, after all, quite stiff.
Wrong. The difference is that flyers through the USPS are paid for entirely by the sender. Spam is paid for as much, or more, by the recipient as the sender. Thus, spam is theft.
Yes, you do have the right to prevent a sender from sending you postal spam. Check out USPS Form 1500: it allows you to obtain an order prohibiting a sender from sending you anything, at your sole discretion. The form, and the law behind it, were originally intended to stop porn and adult solicitations, but the US Supreme Court has ruled that the only one who gets to have an opinion on whether the solicitation being complained about is porn is the recipient - and the USPS must accept any such forms it gets, as long as the requirements of the law (primarily, that it have been opened and is complete) are met.
Unlike the You CAN-SPAM Act, this decision by the FTC shows that they have two clues to rub together. There's no guarantee that spammers would adhere to the list..witness the fact that telephone spammers are moving their operations offshore to evade the do-not-call list.
The only way to stop spam is to hammer the advertisers. Follow the money. Penalize the folks who benefit. No other law-based solution will work.
Supposedly, he didn't sacrifice bandwidth to get the smaller size - in fact, that's a central point of the article: he's apparently found an answer to the "bandwidth, size, efficiency, pick two" problem.
Mark Pilgrim's article perpetuates the myth that only the FSF's definition of "free" software is adequate to ensure that freely available software remains so. I wish people would quit perpetuating this lie. That's an attribute of any software under a license that fits the Open Source Definition. No matter what, a program released under an OSD-compliant license is now, and will always and forever remain, freely available in source form, and will always and forever remain freely modifiable and redistributable.
Slashdot Mirror
Oh, really? How do you explain this entry in the FSF page on "categories of free and non-free software", then? Or this page on the BSD license, where it explicitly lists it as "free"? (While excoriating the advertising clause, which has been invalidated.)
The only explanation for RMS agreeing with the analysis I take issue with is that he's being inconsistent. This would not surprise me in the slightest.
You forget one minor detail: there's nothing stopping others from taking the original BSD code and hacking it up to be compatible with your hypothetical MS/TCP, then releasing it freely.
Actually, I argue that M$ using the BSD TCP stack is a feature, not a bug. If they hadn't, we'd be dealing with their buggy networking stack for years to come, and things would be in even a worse mess than they are now.
The key is that true freedom must necessarily include the freedom to do things that piss other people off, so long as you do not actually harm them. This is where RMS's redefinition of "free" falls short: he defines it such that people are no longer free to piss him off. This destroys it as a useful word.
We've run repeated tests, and Hercules under Cygwin is no slower than Hercules under Linux on the same hardware. This surprised me at first, but it's been shown enough times that I accept it.
We could have a pointer to the Cygwin source, true - but that pointer must be to the same site where the software using it is stored. This doesn't do my disk space any good, nor my maintenance overhead...It's simply not worth it to me to mess with.
The distinction is both necessary and important only to the FSF and its minions. The rest of the world has moved on, by conclusively demonstrating that they simply don't share the FSF's concerns.
If you want to honour the GPL, then if you give a copy of the software to a friend on a CD you are required to either maintain a copy of the src for 3 years and provide it to your friend upon request, or ensure that the src is transferred to your friend when you hand them the CD.
This is exactly why the open-source package Hercules does not distribute the Cygwin binaries with the Windows version, but instead directs the user to the Cygwin site. We've found that Cygwin is enough of a moving target that it's best to install the version that Hercules was compiled against, and that means we'd have to keep around several different source code packages for years just to satisfy the GPL. In this case, we're actively working on getting away from Cygwin just so we don't have to mess with the installation hassles. If Cygwin were licensed under some other license, even one that required that the user be pointed at a place where he could download the source, we wouldn't have this problem.
The GPL actively inhibits software reuse, one of RMS's original goals for the license. Of course, he never lets goals stand in the way of being simon-pure on ideology...
Just how is the Sun license designed to divide and conquer the community? I missed that when I read it. Please enlighten me.
The requirement you say isn't listed equates to "freedom only applies to non-programmers". This is why I refuse to accept the FSF's redefinition of the word "free" as anything approaching authoritative, and one reason why it has utterly failed in its attempt to get its redefinition widely accepted.
They tried to get positive karma for their political position by wrapping it in the flag, and they've failed. It's time that they recognized this and went on to something that might actually do some good.
This analysis is a long way too simplistic. Even the holy RMS says the BSD license is "free", and yet one of the biggest arguments "free software" zealots raise against it is that nasty eeeeeevil vendors (the "free software" zealots never met a vendor they liked) can "take it private".
Of course, this is not the case. Nobody can take BSD-licensed code out of the freely available world. They can add their own extensions to it, but that's a different story.
Not even RMS says that the GPL is the only "free" license.
What's this "we" stuff, kemosabe?
Speak for yourself.
Most objections seem to be to the challenge/response mechanism. I'm persuaded that that would only be use in a tiny minority of cases by this system.
:-)
A bigger problem is the wide range of prerequisites: Java 1.4, JavaMail, Apache with modssl and mod-auth-external, Postfix 2.1. If you're not running x86 or x86-64, forget it. (Or Solaris, but who runs that?
Yup, that is indeed me.
Merciful $DEITY, I'd hope not... because I'd hope Wil would do one hell of a lot better than WFS managed.
When I read Just a Geek, I was struck by the similarity between the reaction to your 25 August 2001 WWdN post (A Look Inside My Mind) and the reaction to the Slashdot and Fark stories about my TRON costume. The 26 August 2001 post (My Velouria) was something I could have written, at least in tone if not in details, if I were half the writer you are.
What would you tell the anonymous, average geek who has Internet fame suddenly dropped on him? How would you help him get through the flood of negativity and get to the point where he can ignore the naysayers and just have fun?
Most chromakey technology these days uses a truly hideous green that's less likely to conflict with clothing or props. I don't know which was used in this movie, though.
All of the scenes in TRON inside the computer were shot on a bare black set with the computer imagery filled in later. This was done in 1982. The actors talk about how hard that was in the making-of video in the collector's edition box set.
They shared OS/360 with everyone at the same time. OS/360 wasn't copyrighted, and was distributed in source form - in fact, to install it, you had to start with a small pregenerated system and completely assemble everything from the ground up. OS/360 was open source when open source wasn't cool - it was just the way everybody did things.
The RCA "compatible" mainframe was compatible only at the problem program level. OS programs (supervisor state code) were markedly different because the I/O subsystem wasn't even close to being compatible.
They're sensible about everything but spam...and they're wrong enough about that that it cancels out the rest of their good karma, for me.
'It's time for a solution to the P2P conflict that pays artists, not lawyers.'
The problem here is that Congress is full of lawyers bent on doing things that amount to full-employment programs for lawyers and accountants. A program like this one that would have the effect of reducing lawsuits has no chance at all.
We complain loudly about conflict of interest by legislators and regulators, while ignoring the biggest one of all: that lawyers write laws. I believe that being a practicing attorney should bar one from being eligible to serve in Congress in much the same way as being an insurance company executive, as a practical matter if not a legal one, bars one from serving as an insurance regulator.
It's still a magic bullet against one particular postal spammer at a time, which is reasonable when the sender is paying for the mail. I have yet to have single mailer continue to send me mail after being hit with that form. The penalties are, after all, quite stiff.
Joe-jobs are easy to detect. Spamfighters do this all the time.
Wrong. The difference is that flyers through the USPS are paid for entirely by the sender. Spam is paid for as much, or more, by the recipient as the sender. Thus, spam is theft.
Yes, you do have the right to prevent a sender from sending you postal spam. Check out USPS Form 1500: it allows you to obtain an order prohibiting a sender from sending you anything, at your sole discretion. The form, and the law behind it, were originally intended to stop porn and adult solicitations, but the US Supreme Court has ruled that the only one who gets to have an opinion on whether the solicitation being complained about is porn is the recipient - and the USPS must accept any such forms it gets, as long as the requirements of the law (primarily, that it have been opened and is complete) are met.
Unlike the You CAN-SPAM Act, this decision by the FTC shows that they have two clues to rub together. There's no guarantee that spammers would adhere to the list..witness the fact that telephone spammers are moving their operations offshore to evade the do-not-call list.
The only way to stop spam is to hammer the advertisers. Follow the money. Penalize the folks who benefit. No other law-based solution will work.
Supposedly, he didn't sacrifice bandwidth to get the smaller size - in fact, that's a central point of the article: he's apparently found an answer to the "bandwidth, size, efficiency, pick two" problem.
According to his bio on qrz.com (he's K1DFT), he hangs out on 160 CW. Me, I just want one to hang on my car for my FT-857. ...Jay, K5ZC
Mark Pilgrim's article perpetuates the myth that only the FSF's definition of "free" software is adequate to ensure that freely available software remains so. I wish people would quit perpetuating this lie. That's an attribute of any software under a license that fits the Open Source Definition. No matter what, a program released under an OSD-compliant license is now, and will always and forever remain, freely available in source form, and will always and forever remain freely modifiable and redistributable.
Why do people keep spreading this falsehood?