Digging further into the Cox situation, the Cox subscriber said:
I tried to send an email. The email only contained text. The text Cox objected to was "http://my_homebox_IP_number/"
I haven't checked the Cox TOS lately, but don't they prohibit running a home web server like all the other residential internet providers? Hasn't this been the case since for essentially the same length of time that the Internet has been a commercial venture?
I agree with required unbundling: any entity with a controlling power over systems that consume public right-of-ways should be required to unbundle those components to the maximum extent technically feasible. Its impractical for everybody to build infrastructure and unhealthy for the economy for one or a small number of organizations to have a lock-in on the infrastructure that does exist.
I don't agree with the government investment part. That would be better handled by reclaiming the universal service fund. It was corrupted in the '90s to support computers for schools and libraries, but that's not what it was originally for. Originally, the USF was used to pay for the additional cost when a telephone company installed a rural telephone but charged the rural user the same price as the city user. It worked well for telephone service and would work fine for fiber-to-the-home broadband if it was allowed to do so.
Besides, everything the US Government does costs 5 times as much and works half as well as comparable operations by small business. Haven't you figured that out yet?
The best software is the software that, given a reasonable choice, folks choose both choose to write and choose to use. Microkernels are not a new idea, yet few folks have chosen to write them and few have chosen to use the ones that have been written. That speaks for itself.
Besides, what does Andy think, that we're all going to say, "Wow, you're dead on, lets rewrite Linux from scratch with a microkernel?" Linux works. Unless we reach a point where it substantially doesn't (like Windows) there's no value to considering that deep a design change.
The unauthorized publication of the letter, therefore, can expose the publisher to liability. Statutory damages under the US Copyright Act can be as much as $150,000 per occurrence
This is not correct. No statutory damages are due unless the copyright was registered. Further, real damages will be limited to damages due to the verbatim copying of the letter, not damages due to revealing its existence and contents both of which are clearly legal under the first amendment. As a practical matter, this means that the damages aren't likely to amount to much. If the judge said otherwise (and since none of the links seem to be to what the judge actually said, I can't tell) then the judge will be trivially overturned.
Wow, you SO BADLY misunderstand how the law works here.
Licensing it as "GPLv2 or later" has the same kind of effect as licensing it as "GPLv2 or BSD or pay me $100." The RECIPIENT may elect to conform to any offered license. However, the recipient's choice of license creates no obligation on the offerer: he too had the option to choose which license he wanted to conform to.
It actually goes beyond that: The offer of "GPLv2 or later" is outside the scope of the GPL itself. GPLv2 only requires conformance with GPLv2. Likewise, GPLv3 only requires conformance with GPLv3. If I want to fork a "GPLv2 or later" project and create a derivative work, I have the right to release that derivative work as GPLv2 only. You can still fetch the pre-forked code and use it under the terms of GPLv3 but my version and the code I've added is only available as GPLv2.
This is intentional on the part of the FSF. If a big hole was found in one of the GPLs, they wanted projects releasing under "GPLv2 or later" to be able to release their next version as "GPLv3 or later" without having to track down all the authors. That's why they so strongly recommend that you use the "or later" clause.
Note that if the offer had said, "Must conform to GPLv2 and all later GPLs" then I would be required to conform to the terms of all of the licenses instead of my choice. Since GPLv3 conflicts with GPLv2, such conformance is impossible. As a result the offer would be void.
He can in fact DO anything he wants. He just can't UNDO things unilaterally. Its his privilege to cease distributing the code under the GPL. However, he probably* can't unilaterally revoke his PRIOR release of the code under the GPL.
* Courts frown on indefinite contracts and licenses. They can be enforceable but generally must meet more stringent criteria to be legal. Also if he can find a way to void the original grant of license then he doesn't need to revoke it because legally it never existed. For example, if he was under 18 or included a copyrighted work for which he had no permission to grant the license.
Which is just a swell theory until you realize that most distros (I'm lookin' at you, Red Hat) do a lousy job of stabilizing the kernel for release. Even with one that does it well (debian) its still unhelpful if you happen to need to run several distros and several versions of each (as is not uncommon in larger deployments) and would like a single kernel that's reliable on all of them.
Its not entirely impossible that he could make it stick, just unlikely. For example: Was he over 18 at the time he released the code under the GPL? If not, he might not have been competent to enter in to a licensing agreement. If that's the case then the original grant of license under the GPL is void. Technically that's not the same as revoking it, but it has the same effect.
Random kernel lockups with a 9-way IDE software RAID on a 1.2ghz Athlon under 2.6.17 and 2.6.18 is the first thing that jumps to mind. Massive interrupt error counts when using more than one channel on Promise IDE controllers on the same.
I haven't tried it on 2.6.>20 and I don't plan to until someone does like was done with 2.6.16 and declares a stable version that will continue to receive bug fixes but not destabilizing new features.
I understand that there are some folks who find it immensely entertaining to try out the latest, greatest kernel but I'm not one of them. I have better things to do with my time.
When is the next "stable" release, a la 2.6.16? I got fed up with the earlier 2.6 kernels and have been sticking with 2.6.16 until there is another release in which the devs make another serious attempt at stability.
creationism (-izem) n. Theol. 1 the doctrine that God creates a new soul for every human being born: opposed to traducianism 2 the doctrine that ascribes the origin of matter, species, etc. to acts of creation by God
But surely the authors of dictionary.reference.com have it right because, hey, its reference.com!
It's possible to think a particular church hierarchy is despicable without finding Christianity itself objectionable. It's unfortunate that you don't see the distinction, particularly given the overwhelming evidence of the existence of groups which have splintered from the church over the years as a result of similar frustrations.
You offered a German proverb, I'll swap you an English one: you threw the baby out with the bathwater.
The laws regarding communications from the White House are completely asinine.
On the one hand, its illegal to undertake political activities using government paid resources. This means that every White House employee that engages in politics must have a separate PC on a separate network with a separate email account. This is the White House. Almost every employee has a political role.
On the other hand, the law compels the White House to keep a record of all official activities. Thus every official email must be saved by the White House system administrators.
Somehow every White House employee is supposed to magically know whether that call to the lobbyist about trying to get the administration's favorite bill passed is an official function or a political function. Not only that, everyone who sends email to someone at the White House is supposed to know which of that individual's email addresses the message should go to so that it can be properly archived.
To cap it all off: the political committees who run the "political" email systems (RNC, DNC) try to stay to a 30-day retention policy after which emails not deliberately saved are deleted. They get sued constantly so if they didn't, the legal discovery across backup tapes would be destructively expensive.
And God help you if you get it wrong because the Congress certainly won't.
Government accountability is a joke. There's so much accountability that it wraps around to zero.
The "int" declaration in the for statement is not so much a C thing, but I won't quibble. Adding/subtracting instead of using a temp variable is novel. Here's another one:
In the interview, on paper without access to a computer, complete this C function without making any additional function calls (i.e. no strlen, no malloc).
char *reverse (char *buf) {/* Reverse the contents of the string buf. E.g. abcde becomes edcba. */ }
Try it. Sure you know how to do it. Any freshman CS major should know how to do it. But on the spot its harder than you expect. It may not identify you as a good programmer, but its remarkably effective at weeding out the poor ones. Its also pretty good at making them feel foolish for having wasted your time with a bogus resume.
Bear in mind that the hand-counted precincts tend to be small while the Diebold automated precincts tend to be large. Large precincts tend to be urban and suburban while small precincts tend to be rural. Urban and suburban voters have a long and well understood history of voting differently than rural voters.
I'm not excusing the use of the crappy Diebold machines. They should be banned. But I am saying that fraud is the least plausible explanation for the difference in the voting pattern.
Okay, they get some credit for including the Atari 400 keyboard. That thing was useless.
I take issue with the complaint about the C64 keyboard. The only serious problem with the C64 keyboard was its integration with the computer so that every bang of a key sent a nice little shockwave into the electronics. The extra symbols were on the edge of the key and printed in a different color. It took about 5 minutes before the operator learned to ignore them. They were, however, extremely helpful to the software developers that wanted to use those symbols. I also don't recall having any trouble missing the backspace key and hitting clear/home. I can see how I might if I had previously been used to a long backspace key, but I wasn't previously used to one.
Slashdot Mirror
Digging further into the Cox situation, the Cox subscriber said:
I tried to send an email. The email only contained text. The text Cox
objected to was "http://my_homebox_IP_number/"
I haven't checked the Cox TOS lately, but don't they prohibit running a home web server like all the other residential internet providers? Hasn't this been the case since for essentially the same length of time that the Internet has been a commercial venture?
I agree with required unbundling: any entity with a controlling power over systems that consume public right-of-ways should be required to unbundle those components to the maximum extent technically feasible. Its impractical for everybody to build infrastructure and unhealthy for the economy for one or a small number of organizations to have a lock-in on the infrastructure that does exist.
I don't agree with the government investment part. That would be better handled by reclaiming the universal service fund. It was corrupted in the '90s to support computers for schools and libraries, but that's not what it was originally for. Originally, the USF was used to pay for the additional cost when a telephone company installed a rural telephone but charged the rural user the same price as the city user. It worked well for telephone service and would work fine for fiber-to-the-home broadband if it was allowed to do so.
Besides, everything the US Government does costs 5 times as much and works half as well as comparable operations by small business. Haven't you figured that out yet?
So, Windows, then?
On the desktop that sure seems to be the case. On the other hand, Linux is doing quite well in some large server niches despite Windows' best efforts.
Also, it's fair to note that Apple hasn't given up the ghost just yet and they're still using a Microkernel the last I heard.
The best software is the software that, given a reasonable choice, folks choose both choose to write and choose to use. Microkernels are not a new idea, yet few folks have chosen to write them and few have chosen to use the ones that have been written. That speaks for itself.
Besides, what does Andy think, that we're all going to say, "Wow, you're dead on, lets rewrite Linux from scratch with a microkernel?" Linux works. Unless we reach a point where it substantially doesn't (like Windows) there's no value to considering that deep a design change.
I thought you Canadians already paid for casual copying with a surcharge on blank CDs? Sounds like double-billing to me.
The unauthorized publication of the letter, therefore, can expose the publisher to liability. Statutory damages under the US Copyright Act can be as much as $150,000 per occurrence
This is not correct. No statutory damages are due unless the copyright was registered. Further, real damages will be limited to damages due to the verbatim copying of the letter, not damages due to revealing its existence and contents both of which are clearly legal under the first amendment. As a practical matter, this means that the damages aren't likely to amount to much. If the judge said otherwise (and since none of the links seem to be to what the judge actually said, I can't tell) then the judge will be trivially overturned.
That's a great point. If I could mod you up, I would.
Wow, you SO BADLY misunderstand how the law works here.
Licensing it as "GPLv2 or later" has the same kind of effect as licensing it as "GPLv2 or BSD or pay me $100." The RECIPIENT may elect to conform to any offered license. However, the recipient's choice of license creates no obligation on the offerer: he too had the option to choose which license he wanted to conform to.
It actually goes beyond that: The offer of "GPLv2 or later" is outside the scope of the GPL itself. GPLv2 only requires conformance with GPLv2. Likewise, GPLv3 only requires conformance with GPLv3. If I want to fork a "GPLv2 or later" project and create a derivative work, I have the right to release that derivative work as GPLv2 only. You can still fetch the pre-forked code and use it under the terms of GPLv3 but my version and the code I've added is only available as GPLv2.
This is intentional on the part of the FSF. If a big hole was found in one of the GPLs, they wanted projects releasing under "GPLv2 or later" to be able to release their next version as "GPLv3 or later" without having to track down all the authors. That's why they so strongly recommend that you use the "or later" clause.
Note that if the offer had said, "Must conform to GPLv2 and all later GPLs" then I would be required to conform to the terms of all of the licenses instead of my choice. Since GPLv3 conflicts with GPLv2, such conformance is impossible. As a result the offer would be void.
Seriously, under what legal theory is this proceeding?
Without knowing any background besides the linked info, I'd guess its one of the following:
The "you wankers didn't respect me so now you can suck wind" theory, or
The "I sold the code and they made me do this to get paid" theory.
Both theories have been very well tested in court. Very well tested.
He can in fact DO anything he wants. He just can't UNDO things unilaterally. Its his privilege to cease distributing the code under the GPL. However, he probably* can't unilaterally revoke his PRIOR release of the code under the GPL.
* Courts frown on indefinite contracts and licenses. They can be enforceable but generally must meet more stringent criteria to be legal. Also if he can find a way to void the original grant of license then he doesn't need to revoke it because legally it never existed. For example, if he was under 18 or included a copyrighted work for which he had no permission to grant the license.
Which is just a swell theory until you realize that most distros (I'm lookin' at you, Red Hat) do a lousy job of stabilizing the kernel for release. Even with one that does it well (debian) its still unhelpful if you happen to need to run several distros and several versions of each (as is not uncommon in larger deployments) and would like a single kernel that's reliable on all of them.
Its not entirely impossible that he could make it stick, just unlikely. For example: Was he over 18 at the time he released the code under the GPL? If not, he might not have been competent to enter in to a licensing agreement. If that's the case then the original grant of license under the GPL is void. Technically that's not the same as revoking it, but it has the same effect.
Random kernel lockups with a 9-way IDE software RAID on a 1.2ghz Athlon under 2.6.17 and 2.6.18 is the first thing that jumps to mind. Massive interrupt error counts when using more than one channel on Promise IDE controllers on the same.
I haven't tried it on 2.6.>20 and I don't plan to until someone does like was done with 2.6.16 and declares a stable version that will continue to receive bug fixes but not destabilizing new features.
I understand that there are some folks who find it immensely entertaining to try out the latest, greatest kernel but I'm not one of them. I have better things to do with my time.
When is the next "stable" release, a la 2.6.16? I got fed up with the earlier 2.6 kernels and have been sticking with 2.6.16 until there is another release in which the devs make another serious attempt at stability.
On the other hand, allowing folks to log on to root directly from a remote console is begging to be hacked. Even telnet didn't allow that.
If you issue laptops to everyone (instead of desktops) the users won't have much choice about powering down at night.
Webster's circa 1995 has it as:
creationism (-izem)
n.
Theol.
1 the doctrine that God creates a new soul for every human being born: opposed to traducianism
2 the doctrine that ascribes the origin of matter, species, etc. to acts of creation by God
But surely the authors of dictionary.reference.com have it right because, hey, its reference.com!
There is no such thing as sound method, in truth.
That's called "Nihilism." It's a particularly nasty branch of philosophy that made Nietzsche a famous man.
It's possible to think a particular church hierarchy is despicable without finding Christianity itself objectionable. It's unfortunate that you don't see the distinction, particularly given the overwhelming evidence of the existence of groups which have splintered from the church over the years as a result of similar frustrations.
You offered a German proverb, I'll swap you an English one: you threw the baby out with the bathwater.
The laws regarding communications from the White House are completely asinine.
On the one hand, its illegal to undertake political activities using government paid resources. This means that every White House employee that engages in politics must have a separate PC on a separate network with a separate email account. This is the White House. Almost every employee has a political role.
On the other hand, the law compels the White House to keep a record of all official activities. Thus every official email must be saved by the White House system administrators.
Somehow every White House employee is supposed to magically know whether that call to the lobbyist about trying to get the administration's favorite bill passed is an official function or a political function. Not only that, everyone who sends email to someone at the White House is supposed to know which of that individual's email addresses the message should go to so that it can be properly archived.
To cap it all off: the political committees who run the "political" email systems (RNC, DNC) try to stay to a 30-day retention policy after which emails not deliberately saved are deleted. They get sued constantly so if they didn't, the legal discovery across backup tapes would be destructively expensive.
And God help you if you get it wrong because the Congress certainly won't.
Government accountability is a joke. There's so much accountability that it wraps around to zero.
Why am I returning anything if I am reverting the string in place?
Because I want to see how you react to a condition that doesn't matter.
The "int" declaration in the for statement is not so much a C thing, but I won't quibble. Adding/subtracting instead of using a temp variable is novel. Here's another one:
char *reverse (char *buf) {
char *front, *back, temp;
for (front=back=buf; *back; back++);
back--;
while (back>front) {
temp=*back;
*back=*front;
*front=temp;
back--;
front++;
}
return buf;
}
In the interview, on paper without access to a computer, complete this C function without making any additional function calls (i.e. no strlen, no malloc).
/* Reverse the contents of the string buf. E.g. abcde becomes edcba. */
char *reverse (char *buf) {
}
Try it. Sure you know how to do it. Any freshman CS major should know how to do it. But on the spot its harder than you expect. It may not identify you as a good programmer, but its remarkably effective at weeding out the poor ones. Its also pretty good at making them feel foolish for having wasted your time with a bogus resume.
Bear in mind that the hand-counted precincts tend to be small while the Diebold automated precincts tend to be large. Large precincts tend to be urban and suburban while small precincts tend to be rural. Urban and suburban voters have a long and well understood history of voting differently than rural voters.
I'm not excusing the use of the crappy Diebold machines. They should be banned. But I am saying that fraud is the least plausible explanation for the difference in the voting pattern.
Okay, they get some credit for including the Atari 400 keyboard. That thing was useless.
I take issue with the complaint about the C64 keyboard. The only serious problem with the C64 keyboard was its integration with the computer so that every bang of a key sent a nice little shockwave into the electronics. The extra symbols were on the edge of the key and printed in a different color. It took about 5 minutes before the operator learned to ignore them. They were, however, extremely helpful to the software developers that wanted to use those symbols. I also don't recall having any trouble missing the backspace key and hitting clear/home. I can see how I might if I had previously been used to a long backspace key, but I wasn't previously used to one.