Surely the answer is to make the car's _owner_ responsible for any fines incurred while driving it. You couldn't do this for more serious driving offences, and it wouldn't be right to cancel or endorse the owner's driving licence if you couldn't prove they were the driver, but it would be easy to just send the owner a bill and let them claim the money off their friend if necessary.
I would think that Brussels could also provide a law that it is illegal to sell a crap product to people while telling them that its great:)
Try reading a typical software licence - 'this product is of absolutely no use whatsoever and there is no warranty of any kind'. What should be illegal is claiming one thing in advertisements (eg 'XXX is a reliable platform for e-commerce') and then comprehensively denying it in the licence.
I heard that older Id Software games such as Doom had been released under the GPL. But I can't seem to download them from Id's website, only the old shareware versions. There are plenty of free Doom clones out there, some based on Id's source, but they all require you to have the original.wad files. Maybe Id hasn't released the wads, just the code to the game itself.
So does this Heretic-derivative actually include playable levels? Or does it require you to purchase a copy of Heretic to get the files needed?
There always seems to be some sort of merger or acquisition going on in the software industry. But have there been any important demergers? (Apart from the mother of all demergers which might happen soon, depending on what the DOJ and courts decide.)
The ten-user limit comes from the proprietary license. I could legally put a 100-user or unlimited user limit on my own.
But that would be effectively useless if the library your program includes (whether by static linking or dynamic linking) has a stricter restriction. The program is useless without the library, therefore any restrictions on the library restrict the program as a whole.
If I program using the MFC, my application doesn't suddenly become part of Windows, or owned by Microsoft or anything. It is completely separate entity. It is I that chooses whether to distribute it or not, I who decides whether my users can distribute it or not, and Microsoft has nothing to say in the matter.
Oh really? Try writing an MFC application and distributing it with full source code under a BSD licence. You can't because Microsoft won't allow their MFC code to be distributed in that way. Yet MFC is an essential part of your program, just as essential as any part you wrote yourself.
In some cases library copyright owners restrict you from distributing your program at all, for example if you bought a 'student' licence for the library.
My program is solely mine, and the library belongs solely to the author.
So logically, something which is the joining-together of program + library will have shared ownership. I don't think it's been proven in court (IANAL) whether linking your code with a library makes your code a derived work of that library. It seems this should certainly be true for static linking, dynamic linking is a bit murkier.
If I build an automobile using Craftsman tools, Sears has no legal claim over my car.
That's because your car does not contain the tools as a part of it (and you probably owned the tools to start with). If you photocopied sections of a book and stuck them into your car's manual, the book's copyright owner would be able to stop you distributing the manual unless you removed those sections. This is a silly analogy though.
Slashdot has always had a noticeable bias, both in the stories posted and in the comments. Being bought by Andover/VA is unlikely to make any difference. Slashdot tends to report on stuff that I tend to find interesting, and that's why I read it. I've never considered it to be the ultimate arbiter of truth.
GPL libraries tell me what restrictions I have to put on my own users. They are dictating terms to a third party. No other license that I am aware of does this.
Every proprietary library does this. For example, suppose you are not allowed to modify the library's source code. This restriction applies to you, and you are also forced to restrict your customers in the same way.
And if the library has a ten-user limit, then any software you distribute using the library will have a ten-user limit. So the library is affecting the terms that your users get for your software.
If you consider the program+library as a whole, taking the terms for the whole work to be the conjunction of the terms for both parts, then the library's licence is restricting the licence of the whole work. If the library doesn't come with permission to distribute it, then your program as a whole would be illegal to distribute. Of course you might have wanted such a restriction anyway, but the library has forced it on you.
But by demanding, even indirectly, the terms that *my* users get for *my* software, the GPL is stepping beyond the bounds of copyright. They are attempting to place their own copyright ownership over software that is not theirs.
Not really - the final program's copyright will be shared between you and the FSF (or whoever). To distribute it, you must find terms on which you both agree. That's how copyright law works (IANAL). If rms wanted to use some of your code in Emacs, he would require your agreement to distribute the resulting work, and you could insist on any copying terms you wanted. Of course, he might not agree and would probably decide to do without your code:-)
Developer resources are certainly finite, but remember that projects often don't scale too well beyond a certain number of developers, unless they are modularized into sub-projects.
So we wouldn't necessarily get faster progress if everyone piled into one JFS, particularly not when version 1.0 hadn't been released. This way we get several to choose from and they can borrow features from each other.
It has always annoyed me how game designers seem to think they know all there is to know about UI design, and insist on creating their own strange point-and-click interface instead of using a Windows-type GUI (or Windows itself) which everyone is familiar with. It's the modern day equivalent of having to enter your name in the high-score table using the joystick, rather than being allowed to actually type the letters on the keyboard.
Inventing a special-purpose UI could be useful for a real-time god game, where everything has to be quickly accessible. But even in these cases the UI could often be made less bizarre and more intuitive without spoiling the game.
Maybe some kind of meta-GUI is needed where you can say what buttons you want in a dialogue box, but not where to place them. The GridBagLayout type stuff is a step in this direction, but you still have to specify the positions of things, even though they are relative rather than absolute.
Things like Glade allow you to describe your GUI in XML. It's possible that you could write a high-level description of what you want in XML, and then use a 'Windows L&F style sheet' or a 'Motif L&F style sheet' together with a set of icons to automatically produce the desired effect. Or maybe this would be better done by actual code rather than getting carried away with style sheets. It would be very cool to simply say what you want the user to input, and have the layout taken care of by platform-specific rules.
When writing CGI programs with Perl and CGI.pm I use a wrapper for getting URL parameters which will check them to ensure they contain only the characters [a-zA-Z0-9_]. It's not necessary most of the time, but it does help protect against new attacks that might be discovered, like this one.
In my opinion, strings passed as URL parameters should be human-readable so that the URL makes some sense; if you're passing long chunks of data such as HTML, it's better to do that with a POST request, which you should also validate thoroughly. So it's no great hardship to insist that HTML tags, hidden null bytes and so on aren't allowed in URLs.
I've played a little tetris program somebody made that was practically immune to analysis other than to the compiler:-)
Maybe that's a way to de-obfuscate programs. Just tell the compiler to assemble them, and the resulting assembler might be easier to understand than the source code. It would probably help to compile for a fairly simple RISC chip (eg ARM, MIPS) that has an easy-to-understand instruction set.
Better still, create a new target processor for gcc that is a simple stack machine or something. Then you could compile your code to this. I think the JVM fits this description, so you could use something which compiles C to Java bytecode, and then use a Java decompiler to spit out Java source. I have no idea if this would be more readable, but it could hardly be any worse...
Some proprietary libraries are available for a fee. But in many cases, the code isn't available at all - for example, if you wanted to use the display code from MS Word as your own rendering library. Of course it is Microsoft's choice whether they wish to make the code available as a library, just as it is the copyright owner's choice whether to license a program under GPL, LGPL, BSD or whatever.
You might complain that if something is a library, it is spiteful to deny it being used in proprietary software. But remember that the library didn't have to be released at all - the copyright owner might have kept it for his own use, or never bothered to write it. That would be just as bad.
Perhaps it will help to look at it like this. If you want to use a proprietary library such as Motif, you have to pay for it. Some libraries are more expensive than others; some are very expensive indeed. If you don't want to pay, you have to find some other code instead. But it is the copyright owner's choice how much to charge, and if they wish to charge a ridiculously high amount, that's their right. Obviously somebody is prepared to pay that, and of course you're not forced to use their library if you don't want to.
Now consider a GPLed library. In this case there is also a price - but it's not money. The payment asked for using the library is that you make your program GPLed. If you think that price is unreasonable, you are free to not use the library. But there will probably be others who think it's a fair price and are happy to pay.
Re:Probably has great applications for walking rob
on
AI Monkey Robot
·
· Score: 2
No, just fit little wheels to the books, and they will come to you;-)
Depends on how you interpret 'as if they were humans'. I'd say that for a bot to fit that definition properly, it should still be convincing even to a person who is suspicious. After all, a real human would still be convincing under the same circumstances (or most of them, anyway).
Perhaps I'm reading too much into the rules of a competition which is just for fun. Probably what the organizers have in mind is a conversation between two bots which looks plausibly human from the outside, not a rigorous test of intelligence.
It seems to me that if you can accuse a GPLed program of 'destroying livelihoods' by not allowing proprietary derived works, you can level exactly the same accusation at MS Office. In fact Office is slightly worse - you can't use their code at all.
(And please let's argue about what the GPL and other licences actually allow, rather than what you think their 'purpose' might be.)
Being able to talk with each other 'as if they were human' would be equivalent to passing the Turing Test. And nobody has managed that, despite fifty years of trying.
Once Linux gets to be more and more mainstream, can we expect to see a block on.iso files as the traffic just gets to be too huge?
Something like that happened here at Imperial College London when Debian 2.1 was released. So many people were downloading the ISO images from SunSITE that the whole of.ac.uk started to get congested. So the bandwidth was throttled in order to give other network users a chance. Unfortunately this had the side effect of throttling bandwidth for the college as a whole.
Slashdot Mirror
Can somebody explain why a right not to self-incriminate is actually a good idea? I'm sure there's a good reason, just not sure what it is.
Surely the answer is to make the car's _owner_ responsible for any fines incurred while driving it. You couldn't do this for more serious driving offences, and it wouldn't be right to cancel or endorse the owner's driving licence if you couldn't prove they were the driver, but it would be easy to just send the owner a bill and let them claim the money off their friend if necessary.
You just unplug the mouse and presto, text mode installer.
Try reading a typical software licence - 'this product is of absolutely no use whatsoever and there is no warranty of any kind'. What should be illegal is claiming one thing in advertisements (eg 'XXX is a reliable platform for e-commerce') and then comprehensively denying it in the licence.
Yes, and I could easily have grabbed Heretic from a warez site long before the source was released.
I heard that older Id Software games such as Doom had been released under the GPL. But I can't seem to download them from Id's website, only the old shareware versions. There are plenty of free Doom clones out there, some based on Id's source, but they all require you to have the original .wad files. Maybe Id hasn't released the wads, just the code to the game itself.
So does this Heretic-derivative actually include playable levels? Or does it require you to purchase a copy of Heretic to get the files needed?
There always seems to be some sort of merger or acquisition going on in the software industry. But have there been any important demergers? (Apart from the mother of all demergers which might happen soon, depending on what the DOJ and courts decide.)
But that would be effectively useless if the library your program includes (whether by static linking or dynamic linking) has a stricter restriction. The program is useless without the library, therefore any restrictions on the library restrict the program as a whole.
Oh really? Try writing an MFC application and distributing it with full source code under a BSD licence. You can't because Microsoft won't allow their MFC code to be distributed in that way. Yet MFC is an essential part of your program, just as essential as any part you wrote yourself.
In some cases library copyright owners restrict you from distributing your program at all, for example if you bought a 'student' licence for the library.
So logically, something which is the joining-together of program + library will have shared ownership. I don't think it's been proven in court (IANAL) whether linking your code with a library makes your code a derived work of that library. It seems this should certainly be true for static linking, dynamic linking is a bit murkier.
That's because your car does not contain the tools as a part of it (and you probably owned the tools to start with). If you photocopied sections of a book and stuck them into your car's manual, the book's copyright owner would be able to stop you distributing the manual unless you removed those sections. This is a silly analogy though.
Slashdot has always had a noticeable bias, both in the stories posted and in the comments. Being bought by Andover/VA is unlikely to make any difference. Slashdot tends to report on stuff that I tend to find interesting, and that's why I read it. I've never considered it to be the ultimate arbiter of truth.
Every proprietary library does this. For example, suppose you are not allowed to modify the library's source code. This restriction applies to you, and you are also forced to restrict your customers in the same way.
And if the library has a ten-user limit, then any software you distribute using the library will have a ten-user limit. So the library is affecting the terms that your users get for your software.
If you consider the program+library as a whole, taking the terms for the whole work to be the conjunction of the terms for both parts, then the library's licence is restricting the licence of the whole work. If the library doesn't come with permission to distribute it, then your program as a whole would be illegal to distribute. Of course you might have wanted such a restriction anyway, but the library has forced it on you.
Not really - the final program's copyright will be shared between you and the FSF (or whoever). To distribute it, you must find terms on which you both agree. That's how copyright law works (IANAL). If rms wanted to use some of your code in Emacs, he would require your agreement to distribute the resulting work, and you could insist on any copying terms you wanted. Of course, he might not agree and would probably decide to do without your code :-)
Developer resources are certainly finite, but remember that projects often don't scale too well beyond a certain number of developers, unless they are modularized into sub-projects.
So we wouldn't necessarily get faster progress if everyone piled into one JFS, particularly not when version 1.0 hadn't been released. This way we get several to choose from and they can borrow features from each other.
Yeah, but won't they stop being a 'partner' once VA have Freshmeat instead? Oh well, not a big deal I suppose.
How will it affect Freshmeat (owned by Andover) and LinuxApps (a 'partner' of linux.com, owned by VA)?
It has always annoyed me how game designers seem to think they know all there is to know about UI design, and insist on creating their own strange point-and-click interface instead of using a Windows-type GUI (or Windows itself) which everyone is familiar with. It's the modern day equivalent of having to enter your name in the high-score table using the joystick, rather than being allowed to actually type the letters on the keyboard.
Inventing a special-purpose UI could be useful for a real-time god game, where everything has to be quickly accessible. But even in these cases the UI could often be made less bizarre and more intuitive without spoiling the game.
Maybe some kind of meta-GUI is needed where you can say what buttons you want in a dialogue box, but not where to place them. The GridBagLayout type stuff is a step in this direction, but you still have to specify the positions of things, even though they are relative rather than absolute.
Things like Glade allow you to describe your GUI in XML. It's possible that you could write a high-level description of what you want in XML, and then use a 'Windows L&F style sheet' or a 'Motif L&F style sheet' together with a set of icons to automatically produce the desired effect. Or maybe this would be better done by actual code rather than getting carried away with style sheets. It would be very cool to simply say what you want the user to input, and have the layout taken care of by platform-specific rules.
When writing CGI programs with Perl and CGI.pm I use a wrapper for getting URL parameters which will check them to ensure they contain only the characters [a-zA-Z0-9_]. It's not necessary most of the time, but it does help protect against new attacks that might be discovered, like this one.
In my opinion, strings passed as URL parameters should be human-readable so that the URL makes some sense; if you're passing long chunks of data such as HTML, it's better to do that with a POST request, which you should also validate thoroughly. So it's no great hardship to insist that HTML tags, hidden null bytes and so on aren't allowed in URLs.
Maybe that's a way to de-obfuscate programs. Just tell the compiler to assemble them, and the resulting assembler might be easier to understand than the source code. It would probably help to compile for a fairly simple RISC chip (eg ARM, MIPS) that has an easy-to-understand instruction set.
Better still, create a new target processor for gcc that is a simple stack machine or something. Then you could compile your code to this. I think the JVM fits this description, so you could use something which compiles C to Java bytecode, and then use a Java decompiler to spit out Java source. I have no idea if this would be more readable, but it could hardly be any worse...
Some proprietary libraries are available for a fee. But in many cases, the code isn't available at all - for example, if you wanted to use the display code from MS Word as your own rendering library. Of course it is Microsoft's choice whether they wish to make the code available as a library, just as it is the copyright owner's choice whether to license a program under GPL, LGPL, BSD or whatever.
You might complain that if something is a library, it is spiteful to deny it being used in proprietary software. But remember that the library didn't have to be released at all - the copyright owner might have kept it for his own use, or never bothered to write it. That would be just as bad.
Perhaps it will help to look at it like this. If you want to use a proprietary library such as Motif, you have to pay for it. Some libraries are more expensive than others; some are very expensive indeed. If you don't want to pay, you have to find some other code instead. But it is the copyright owner's choice how much to charge, and if they wish to charge a ridiculously high amount, that's their right. Obviously somebody is prepared to pay that, and of course you're not forced to use their library if you don't want to.
Now consider a GPLed library. In this case there is also a price - but it's not money. The payment asked for using the library is that you make your program GPLed. If you think that price is unreasonable, you are free to not use the library. But there will probably be others who think it's a fair price and are happy to pay.
No, just fit little wheels to the books, and they will come to you ;-)
Depends on how you interpret 'as if they were humans'. I'd say that for a bot to fit that definition properly, it should still be convincing even to a person who is suspicious. After all, a real human would still be convincing under the same circumstances (or most of them, anyway).
Perhaps I'm reading too much into the rules of a competition which is just for fun. Probably what the organizers have in mind is a conversation between two bots which looks plausibly human from the outside, not a rigorous test of intelligence.
A slightly clearer comparison:
GPL:Proprietary:
It seems to me that if you can accuse a GPLed program of 'destroying livelihoods' by not allowing proprietary derived works, you can level exactly the same accusation at MS Office. In fact Office is slightly worse - you can't use their code at all.
(And please let's argue about what the GPL and other licences actually allow, rather than what you think their 'purpose' might be.)
Being able to talk with each other 'as if they were human' would be equivalent to passing the Turing Test. And nobody has managed that, despite fifty years of trying.
Bastard Operator From...
Just a quick reminder:
GPL: you can use this code in your own programs, but they have to be GPLed.
Proprietary: you can't use this code in your own programs at all.
Which one is more 'spiteful', as you put it?
Something like that happened here at Imperial College London when Debian 2.1 was released. So many people were downloading the ISO images from SunSITE that the whole of .ac.uk started to get congested. So the bandwidth was throttled in order to give other network users a chance. Unfortunately this had the side effect of throttling bandwidth for the college as a whole.