But the Open Source mindset that is required to have a successful project can't also agree with the idea of censorship in the first place.
Last I saw, the Junkbuster was free software. As I'm pretty sure you know, its sole purpose is to block sites you don't want to see. It can easily be combined with a firewall prohibiting outbound connections to port 80 to force a machine's users (i.e. your kids) to rely on it for web browsing. Sounds like a net filter to me.
No two censors agree on what should be censored,
Here's the thing: I'm pretty sure that any Scientologist will agree that xenu.net should be blocked. I'm pretty sure that much of PFLAG thinks that godhatesfags.com should be blocked. I'm pretty sure that there are a lot of other small groups who agree about blocking/unblocking decisions, and are willing to spend the time to write their own filter.
They can each have their own filter.
I understand that the Scientologists actually have their own special net filter, to keep their flock focused on acceptable thoughts (or whatever). That's fine. They chose to be Scientologists (dumbasses... but I digress).
I use the Junkbuster to block ad.doubleclick.net. That's also fine. I chose not to be tracked or advertised to. I've also imposed that decision on my girlfriend, who chooses to use my box to access the net, and on anybody else who would (hypothetically) want and get an account on my machine.
Both cases are voluntary censorship, and perfectly valid.
For that matter, as far as the software itself goes, I share many goals with the Scientologists. We both want the software to be an effective blocking agent. We just disagree on what to block. For that reason, we can collaborate on building open-source blocking software, and use different blocklists.
An "Open Source mindset" should be able to encompass a group of interested individuals who choose to block their own access to a bunch of sites. The problem comes in when they block other people's access to the sites. Junkbuster can be just as restrictive as any other Censorware program, if you use it that way. It can be very effective in ensuring that your public library will be safe from http://hottits.com (and http://aclu.org). As long as you can choose your own filter, though, there's no problem -- and since (by definition) you can't control what free software is used for, you'll see open-source filter programs developed for voluntary self-censorship, and used (by jerks) to censor others.
Good question. Trivially, you're in America -- the local cops can still come up and knock on your door, and haul your ass off. They can't get at your Swedish data, though, without asking somebody in Switzerland. (OTOH, we all know how good European cops are about following orders from US IP interests... Norwegians in particular.)
I'm not a lawyer, though, so what do I know? At any rate, I think this isn't so much a legal question than a practical question of 'whose cops patrol my neighborhood?' --
The 4DOS/4NT/4OS2 command.com/cmd.exe replacements from JP Software have offered tab completion for many years. So there is no excuse for M$ to have left out this useful function for so long.
How's this for an excuse: "Average people will use our new, spiffy Windows 3.1 point-n-click interface to select files. Only power users will use the command line now... and power users aren't a significant portion of our target market for DOS-Win3.1."
Yes, it sucks, but it's the logic of the (overrated) free market: if we (think we) won't get an acceptable ROI, we won't do it. It's not a valid excuse, but an excuse nonetheless... --
So you're saying instead of just buying a product I need to write a product for me? That's just stupid.
If there just happens to be a proprietary product out there that meets your needs, go ahead and buy it. Just go in with your eyes open:
You're at ProprietaryCorp's mercy if there's a problem, or if you want a new feature implemented. You don't have source; you have to ask them kindly to do it for you -- they may or may not, depending on the negotiation. If you had gone with an open solution, you at least have the option of getting somebody else to do the customization (or at least threaten PropCorp with going elsewhere).
Your code is only being improved by PropCorp's coders. Will they find and fix as many bugs as the free software community would? Will their programmers write clean code, especially if nobody outside PropCorp will ever read it?
What if PropCorp goes out of business, or gets bought out by your arch-rival?
Paranoia mode on: You know what PropSolution says it does. What exactly does it do? Is it secure? Does it give J Random Employee at PropCorp a back door -- or is its security model just broken? If it uses crypto, what kind of crypto does it use, and how reliable is the implementation?
It's late here. I could go on, but then again, I could sleep -- and I'm preaching to the choir.:)
Now, if PropSolution[TM] does not exactly fit your needs, what then? If you need a special foobar for your company, you can either pay a PropCorp to code it into their proprietary product (if they'll give you the time of day) -- or you can pay a contractor (or your own coding team, if you have one) to code it into an open product. Either way, if you want something new, you're paying somebody. --
They're claiming that through the device of the shrink wrap license, you have entered into a contract with them where you agree not to reverse engineer the software.
In the Santa Clara trade secret case, yes. They're even trying to claim that the mere existence of a shrinkwrap license somewhere binds everybody to that license, even if the person in question has never seen the license text.
If you include the east coast case, they're claiming more than that... They're claiming that the DMCA + CSS combo effectively puts the software and data in a cheap cardboard box, with a 'video out' jack. If you try to open the box to exercise your fair use rights to the data inside, you violate the DMCA first, before any fair use questions enter into it -- the DMCA prohibits opening the box without permission, and if there's no other way to get fair use, well, life's tough. And they claim that this isn't "interoperability," but "piracy," so that loophole is moot; they are trying to protect movies through code, not the code itself -- so they claim. That's what's at stake in the east coast case.
Pay no attention to the code behind that curtain! You could be committing a copyright violation! --
Something smells putrid, even... I didn't even mention that the GPL is of no use whatsoever against the claims of a third party with a patent. They never agreed to the GPL, they have the patent covering the GPL'd work, the GPL doesn't apply to them. The only benefit in the GPL in this case is that it encourages prior-use precedents through earlier code release.
A while ago I tried to put together a framework for explicit patent grants alongside GPL redistribution... and then I realized that that kind of agreement won't bind a third party.
This sort of legal crap is why software patents are a Bad Idea[TM]. --
The GPL does state that any patented code you use must be licensed free for everyone's use.
Yes, in the preamble... Unfortunately, in the license itself, it says:
Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted...
Clause 7 only says that a patent license can't restrict the modification/distribution requirements of the GPL without violating the GPL. It doesn't forbid using a patent to restrict use of the program.
I am not a lawyer, but as far as I do understand the law, a patent holder has the right to restrict your use of a GPL'd program that implements his/her patented process. GPL only protects you from copyright claims, and other attempts to deny you the copyright rights of modification, distribution, and copying; it doesn't protect you from a third party's claim against use. --
What's needed is a giant defensive organization (maybe the FSF...) that holds hundreds, or even thousands, of patents
Only if there is some kind of check on the awesome power that such an organization would hold. We already have big organizations with lots of patents and no responsibility to free software, thank you.
I would only let a "defensive organization" hold my patents (if/when I have any) if I had some kind of ironclad promise of defense. I don't want to be in a situation where my big-daddy white knight organization decides that my patent problem is too small or too messy to defend...
Even the FSF, when asking for GNU contributors to sign over copyright, will give a return agreement pledging to only license the work under terms allowing free redistribution. I would expect nothing less from a group that wants me to sign over a patent. --
i.e. instead of "hitting skip on a crappy song, going back to work, getting another topics crappy song and needing to quit work (mentally) to skip it" you can just skip all the shit at once. Well, WADR, the problem is between the keyboard and the chair. "Hitting skip" is not the way to solve the problem of unwanted songs in your playlist. "Deleting the song from the playlist" is the solution. c.f. Larry Wall on "true laziness."
Quite frankly, I hope the CMU network admins get sued under every computer trespassing law available.
Who exactly gets to sue them? The students are the only interested party here, I think (IANAL). And from what I hear, most of those students are guilty of copyright violations.
Can you say countersuit? These guys probably have every reason to just shut up and take their lumps. Even if their lawyers manage to get all of the evidence thrown out because of the breach of privacy, those students still have to go through the trial and bear the legal risk.
I may be an incurable pessimist, but I don't think many undergraduates would be up for it. Maybe if they had a lot of EFF/ACLU backing, or rich parents, but probably not. There are some deep pockets at RIAA (and CMU) -- it would be one hell of a costly battle.
(i.e. I release foo-0.9 GPL'd, foo-1.0 is then non GPL'd commercial software, but is foo-1.0 infected by foo-0.9 because it's a derivative work? Is retroactive license changing allowed? Probably not)
If foo-0.9 is all your work, you can release foo-1.0 under any license you want. You can even release foo-0.9 under any license you want. Copyright is tied to particular manuscript instances, not the abstract string-of-words: you can hand Joe a foo-0.9 under GPL, and Jim a foo-0.9 under MS EULA, and Jim can't redistribute (unless he got Joe to give him a GPL'd copy).
If foo-0.9 is not all yours, then you have to ask all the people who contributed more than N lines of code (IIRC, FSF says N=16, but I could be wrong) to agree to change their licensing terms. Once you have every code author releasing their stuff to you under Some Random License (SRL), you can then go ahead and treat the whole thing as SRL, no matter what previous license terms have been arranged.
This is all as far as I understand it; I've never seen the inside of a law school. (I've seen the inside of a lawyer, but that was on "ER.")
Nope. It says "GPL version 2." Under GPL, you have the option to say "N or later" when you release your own new code. In this case, Linus apparently chose not to.
Sun may have its faults, but one of them is not giving us crap to work with, and I don't think they plan to start any time soon
Historically, companies decide that they can suck and get away with it only *after* they lock the market (via monopoly, or technical hooks a la the ever-changing Word spec). MS has its market locked in. Sun does not have its market locked in... yet.
Scott Myers' "Effectice C++" and "More Effective C++" may be of some use to more advanced C++ programmers.
I agree with most of what you said. I have to say, though, that Effective C++ is a great book even for newbie C++ coders. It covers lots of basic C++ principles, on the order of "Use cout and cin instead of stdio," that IMHO every user of g++ should know.
assuming that their employees could be trained to the OSS vision. Why does this sound frightenly like a Soviet-style "Reducation Camp?"
Well, except that they both involve "education," I have no idea.
Nobody forces you to work for any particular company. If you don't like the company's basic goals and philosophy, find another job. Or convince the company that you're right and they're wrong.
Actually, the Nielsens don't do the "family" thing anymore. They send a booklet to homes at random, and sample only a week of their viewing time. I have no idea how faithfully they process your one booklet after you send it back (presumably they tally everybody's, since they tell you to send the thing back even if you didn't watch any TV at all).
One method of quantum encryption (the one I've studied) is not really encryption at all. Alice sends Bob messages in plaintext, as individual quantum states. Now, if Eve wants to intercept that signal, she must measure the quantum states as they go by; that measurement will necessarily corrupt the signal, and Bob gets a meCsaTe with a f3w bits scraBbled. That's a signal that Eve has been mucking around again -- but better yet, if the message is NOT scrambled, Bob knows that NOBODY could have been snooping; and if the message is scrambled, Bob knows by how much is scrambled how precisely Eve has snooped.
Several drawbacks to this method: you have to be able to measure individual quantum states (which means expensive equipment); you have to have a network that doesn't lose much of the information in the fragile quantum states (or at least loses a definite, known amount of information). That means you probably can only use it over a local network, at least with current (read: "at least 10 years ahead") technology. (You also have to know the expected information content in the message to determine that something's been screwed up, but that's usually easy.)
Another method involves using EPR pairs to "teleport" a message by measuring one of the pair (which causes the other to collapse based on the choice of measurement) -- you need to send a classical message with the signal (so speed-of-light won't be violated with this method), but the classical message is meaningless without the EPR state, and vice-versa. Again, good luck carrying around your very own EPR pairs; hope you have a huge budget!
I suspect they put the 'e' and 't' keys on the horizontal axis because horizontal movements are easier than vertical ones. However, it may be conceptually easier to put space, backspace, and enter on the horizontal axis instead. To move forward, I want to flick the pen forward, not down!
On the other hand, there's a nice symmetry to the keyboard in their current positions. Down could mean "hitting the space bar," up could mean "stretching your pinky up to hit that backspace key," and down-left looks like the symbol on many enter keys.
I just think it's more viscerally appealing to see the cursor move the same way your pen moves; besides, that's the way Grafitti already does it, so it can't be wrong.:) Perhaps there could be an option (or HackMaster thingy) to swap the N/S fields with E/W...
Having a few concentrated, talented developers, is better than having a million semi-witless volunteers. And guess what, boys and girls, is the best way to conetrate a group of engineers? Thats right kids - pay them and put them in the same building. Oops, I forgot, corporations aren't in style with the/. crowd, are they?
Well, if you only give me the XOR there, I'd also prefer a core group of savvy engineers. But if I could have both, I'd jump at the chance. God knows some of the stuff I've written professionally could have used a hundred extra eyeballs, even though it was good enough for a corporate environment. Most of my early ('96) stuff would have brought me acute embarrassment if the source was freed... and yet, frighteningly enough, it's still being used because nobody's bothered to do better.
So, yes, I guess corporations aren't in style on/. Perhaps it's because corporations tend to settle for sloppy code where they can, and the average FS developer is something of a perfectionist?
Nonsense! They're supposed to be mostly believable until you get it -- and then you're supposed to feel foolish for missing the subtle inconsistencies. (For example: no link to the actual announcement article, and no huge fanfare on the site's front page.)
They're also supposed to happen in the AM, because that's when all the reasonable people are sleepy and gullible.
I know I was fooled for a few minutes... but that's because I read/. while waiting for the morning tea water to boil.
Slashdot Mirror
Last I saw, the Junkbuster was free software. As I'm pretty sure you know, its sole purpose is to block sites you don't want to see. It can easily be combined with a firewall prohibiting outbound connections to port 80 to force a machine's users (i.e. your kids) to rely on it for web browsing. Sounds like a net filter to me.
No two censors agree on what should be censored,
Here's the thing: I'm pretty sure that any Scientologist will agree that xenu.net should be blocked. I'm pretty sure that much of PFLAG thinks that godhatesfags.com should be blocked. I'm pretty sure that there are a lot of other small groups who agree about blocking/unblocking decisions, and are willing to spend the time to write their own filter.
They can each have their own filter.
I understand that the Scientologists actually have their own special net filter, to keep their flock focused on acceptable thoughts (or whatever). That's fine. They chose to be Scientologists (dumbasses... but I digress).
I use the Junkbuster to block ad.doubleclick.net. That's also fine. I chose not to be tracked or advertised to. I've also imposed that decision on my girlfriend, who chooses to use my box to access the net, and on anybody else who would (hypothetically) want and get an account on my machine.
Both cases are voluntary censorship, and perfectly valid.
For that matter, as far as the software itself goes, I share many goals with the Scientologists. We both want the software to be an effective blocking agent. We just disagree on what to block. For that reason, we can collaborate on building open-source blocking software, and use different blocklists.
An "Open Source mindset" should be able to encompass a group of interested individuals who choose to block their own access to a bunch of sites. The problem comes in when they block other people's access to the sites. Junkbuster can be just as restrictive as any other Censorware program, if you use it that way. It can be very effective in ensuring that your public library will be safe from http://hottits.com (and http://aclu.org). As long as you can choose your own filter, though, there's no problem -- and since (by definition) you can't control what free software is used for, you'll see open-source filter programs developed for voluntary self-censorship, and used (by jerks) to censor others.
--
I'm not a lawyer, though, so what do I know? At any rate, I think this isn't so much a legal question than a practical question of 'whose cops patrol my neighborhood?'
--
--
How's this for an excuse: "Average people will use our new, spiffy Windows 3.1 point-n-click interface to select files. Only power users will use the command line now... and power users aren't a significant portion of our target market for DOS-Win3.1."
Yes, it sucks, but it's the logic of the (overrated) free market: if we (think we) won't get an acceptable ROI, we won't do it. It's not a valid excuse, but an excuse nonetheless...
--
If there just happens to be a proprietary product out there that meets your needs, go ahead and buy it. Just go in with your eyes open:
- You're at ProprietaryCorp's mercy if there's a problem, or if you want a new feature implemented. You don't have source; you have to ask them kindly to do it for you -- they may or may not, depending on the negotiation. If you had gone with an open solution, you at least have the option of getting somebody else to do the customization (or at least threaten PropCorp with going elsewhere).
- Your code is only being improved by PropCorp's coders. Will they find and fix as many bugs as the free software community would? Will their programmers write clean code, especially if nobody outside PropCorp will ever read it?
- What if PropCorp goes out of business, or gets bought out by your arch-rival?
- Paranoia mode on: You know what PropSolution says it does. What exactly does it do? Is it secure? Does it give J Random Employee at PropCorp a back door -- or is its security model just broken? If it uses crypto, what kind of crypto does it use, and how reliable is the implementation?
- It's late here. I could go on, but then again, I could sleep -- and I'm preaching to the choir.
:)
Now, if PropSolution[TM] does not exactly fit your needs, what then? If you need a special foobar for your company, you can either pay a PropCorp to code it into their proprietary product (if they'll give you the time of day) -- or you can pay a contractor (or your own coding team, if you have one) to code it into an open product. Either way, if you want something new, you're paying somebody.--
In the Santa Clara trade secret case, yes. They're even trying to claim that the mere existence of a shrinkwrap license somewhere binds everybody to that license, even if the person in question has never seen the license text.
If you include the east coast case, they're claiming more than that... They're claiming that the DMCA + CSS combo effectively puts the software and data in a cheap cardboard box, with a 'video out' jack. If you try to open the box to exercise your fair use rights to the data inside, you violate the DMCA first, before any fair use questions enter into it -- the DMCA prohibits opening the box without permission, and if there's no other way to get fair use, well, life's tough. And they claim that this isn't "interoperability," but "piracy," so that loophole is moot; they are trying to protect movies through code, not the code itself -- so they claim. That's what's at stake in the east coast case.
Pay no attention to the code behind that curtain! You could be committing a copyright violation!
--
Something smells putrid, even... I didn't even mention that the GPL is of no use whatsoever against the claims of a third party with a patent. They never agreed to the GPL, they have the patent covering the GPL'd work, the GPL doesn't apply to them. The only benefit in the GPL in this case is that it encourages prior-use precedents through earlier code release.
A while ago I tried to put together a framework for explicit patent grants alongside GPL redistribution... and then I realized that that kind of agreement won't bind a third party.
This sort of legal crap is why software patents are a Bad Idea[TM].
--
Yes, in the preamble... Unfortunately, in the license itself, it says:
Clause 7 only says that a patent license can't restrict the modification/distribution requirements of the GPL without violating the GPL. It doesn't forbid using a patent to restrict use of the program.I am not a lawyer, but as far as I do understand the law, a patent holder has the right to restrict your use of a GPL'd program that implements his/her patented process. GPL only protects you from copyright claims, and other attempts to deny you the copyright rights of modification, distribution, and copying; it doesn't protect you from a third party's claim against use.
--
Only if there is some kind of check on the awesome power that such an organization would hold. We already have big organizations with lots of patents and no responsibility to free software, thank you.
I would only let a "defensive organization" hold my patents (if/when I have any) if I had some kind of ironclad promise of defense. I don't want to be in a situation where my big-daddy white knight organization decides that my patent problem is too small or too messy to defend...
Even the FSF, when asking for GNU contributors to sign over copyright, will give a return agreement pledging to only license the work under terms allowing free redistribution. I would expect nothing less from a group that wants me to sign over a patent.
--
--
i.e. instead of "hitting skip on a crappy song, going back to work, getting another topics crappy song and needing to quit work (mentally) to skip it" you can just skip all the shit at once. Well, WADR, the problem is between the keyboard and the chair. "Hitting skip" is not the way to solve the problem of unwanted songs in your playlist. "Deleting the song from the playlist" is the solution. c.f. Larry Wall on "true laziness."
--
Who exactly gets to sue them? The students are the only interested party here, I think (IANAL). And from what I hear, most of those students are guilty of copyright violations.
Can you say countersuit? These guys probably have every reason to just shut up and take their lumps. Even if their lawyers manage to get all of the evidence thrown out because of the breach of privacy, those students still have to go through the trial and bear the legal risk.
I may be an incurable pessimist, but I don't think many undergraduates would be up for it. Maybe if they had a lot of EFF/ACLU backing, or rich parents, but probably not. There are some deep pockets at RIAA (and CMU) -- it would be one hell of a costly battle.
--
If foo-0.9 is all your work, you can release foo-1.0 under any license you want. You can even release foo-0.9 under any license you want. Copyright is tied to particular manuscript instances, not the abstract string-of-words: you can hand Joe a foo-0.9 under GPL, and Jim a foo-0.9 under MS EULA, and Jim can't redistribute (unless he got Joe to give him a GPL'd copy).
If foo-0.9 is not all yours, then you have to ask all the people who contributed more than N lines of code (IIRC, FSF says N=16, but I could be wrong) to agree to change their licensing terms. Once you have every code author releasing their stuff to you under Some Random License (SRL), you can then go ahead and treat the whole thing as SRL, no matter what previous license terms have been arranged.
This is all as far as I understand it; I've never seen the inside of a law school. (I've seen the inside of a lawyer, but that was on "ER.")
--
Nope. It says "GPL version 2." Under GPL, you have the option to say "N or later" when you release your own new code. In this case, Linus apparently chose not to.
--
Historically, companies decide that they can suck and get away with it only *after* they lock the market (via monopoly, or technical hooks a la the ever-changing Word spec). MS has its market locked in. Sun does not have its market locked in... yet.
--
I agree with most of what you said. I have to say, though, that Effective C++ is a great book even for newbie C++ coders. It covers lots of basic C++ principles, on the order of "Use cout and cin instead of stdio," that IMHO every user of g++ should know.
--
Well, except that they both involve "education," I have no idea.
Nobody forces you to work for any particular company. If you don't like the company's basic goals and philosophy, find another job. Or convince the company that you're right and they're wrong.
--
Actually, the Nielsens don't do the "family" thing anymore. They send a booklet to homes at random, and sample only a week of their viewing time. I have no idea how faithfully they process your one booklet after you send it back (presumably they tally everybody's, since they tell you to send the thing back even if you didn't watch any TV at all).
--
One method of quantum encryption (the one I've studied) is not really encryption at all. Alice sends Bob messages in plaintext, as individual quantum states. Now, if Eve wants to intercept that signal, she must measure the quantum states as they go by; that measurement will necessarily corrupt the signal, and Bob gets a meCsaTe with a f3w bits scraBbled. That's a signal that Eve has been mucking around again -- but better yet, if the message is NOT scrambled, Bob knows that NOBODY could have been snooping; and if the message is scrambled, Bob knows by how much is scrambled how precisely Eve has snooped.
Several drawbacks to this method: you have to be able to measure individual quantum states (which means expensive equipment); you have to have a network that doesn't lose much of the information in the fragile quantum states (or at least loses a definite, known amount of information). That means you probably can only use it over a local network, at least with current (read: "at least 10 years ahead") technology. (You also have to know the expected information content in the message to determine that something's been screwed up, but that's usually easy.)
Another method involves using EPR pairs to "teleport" a message by measuring one of the pair (which causes the other to collapse based on the choice of measurement) -- you need to send a classical message with the signal (so speed-of-light won't be violated with this method), but the classical message is meaningless without the EPR state, and vice-versa. Again, good luck carrying around your very own EPR pairs; hope you have a huge budget!
--
On the other hand, there's a nice symmetry to the keyboard in their current positions. Down could mean "hitting the space bar," up could mean "stretching your pinky up to hit that backspace key," and down-left looks like the symbol on many enter keys.
I just think it's more viscerally appealing to see the cursor move the same way your pen moves; besides, that's the way Grafitti already does it, so it can't be wrong. :) Perhaps there could be an option (or HackMaster thingy) to swap the N/S fields with E/W...
--
Three words: YANKEE GO HOME.
Seen all over Europe in/after WW2.
--
Well, if you only give me the XOR there, I'd also prefer a core group of savvy engineers. But if I could have both, I'd jump at the chance. God knows some of the stuff I've written professionally could have used a hundred extra eyeballs, even though it was good enough for a corporate environment. Most of my early ('96) stuff would have brought me acute embarrassment if the source was freed... and yet, frighteningly enough, it's still being used because nobody's bothered to do better.
So, yes, I guess corporations aren't in style on /. Perhaps it's because corporations tend to settle for sloppy code where they can, and the average FS developer is something of a perfectionist?
Or maybe we're all commies. :)
--
They're also supposed to happen in the AM, because that's when all the reasonable people are sleepy and gullible.
I know I was fooled for a few minutes... but that's because I read /. while waiting for the morning tea water to boil.
--
Well, of course you pronounce the slash! Otherwise, you wouldn't be giving Slashdot the recognition it deserves...
--
Okay, next caller...
--