Re: defining "distribution" (Re: A few more LPPL concerns)
On Mon, 2002-07-22 at 09:49, Boris Veytsman wrote: > > From: Jeff Licquia <[EMAIL PROTECTED]> > > > My /usr/local/bin can > > > be NFS-exported to hundreds of computers. Even my box can have > > > hundreds logins there. > > > > Yes, but in the former case, you are distributing the program to > > hundreds of computers. In the latter, hundreds of users are running the > > program. > > I think this is bizarre. On my home network there are several x86 and > several Sparcs. The x86s share /usr while the Sparcs have their > individual /usr each. You are saying that something is legal to do on > my Sparc but not on x86s? And if I take the decision to take a hard > drive from the Sparc and put on the fileserver, the legality changes? Sure. The actual situation changes, so the legal situation could change. What you don't seem to be getting is that *licenses can restrict almost anything*. When talking about hypothetical licenses, then anything could possibly be true. There is no way of talking about the legality of an issue or not outside of particular licenses, unless you're talking about fair use rights. In the case of the GPL and your home network, nothing really changes. You own all of the machines on your home network, so you don't have any obligation you didn't have before concerning source distribution. But I don't know what other software you have installed on those systems, or what their licenses require. (This is a motivation for the DFSG, by the way. When you install Debian main on a system, you know that you have certain rights and responsibilities that are common to all of the software you just installed.) > > And, the GPL says: > > > > > The act of running the Program is not restricted... > > > > Therefore, the act of running the program from /usr/local/bin is not > > restricted. > > I agree that *running* a program is not restricted -- both by GPL and > LPPL. However, I think that when *I* run a program from > /usr/local/bin, where *you* put it with the explicit intent to make it > available for me to run, there are two acts involved. Namely, *you* > distributed the program, and *I* run it. My actions are not > restricted, but *yours* are restricted by GPL and LPPL. By the LPPL, yes. By the GPL, no, that's not how it's interpreted. "make install" makes a copy from one location to another. Since I'm the person making the copy on my own machine, then I'm not distributing to anyone. The fact that I add a user to my box and let my neighbor log in does not create any new copies of anything. And when my neighbor runs a GPLed program, "the act of running the Program is not restricted", so again I have no new restrictions. > Again, imagine a company that put a closed-source derivative of a > GPL'ed program on a company machine and refused to provide sources, > but encouraged employees to run it for company business. Would you > consider this legal? If not, you must recognize the act ot placing the > executables in /usr/local/bin on a machine accessible by others to be > what is it -- a distribution. Personally, I don't have any problem with this. If the company licensed the program under the GPL, and the company isn't distributing the program, then the company doesn't have to provide source. This is a principle that the FSF supports, by the way, although perhaps not to the same extent that I do. > > The key is that the license has to spell it out. The GPL talks about > > "copying, distribution, and modification", but sees fit to not define > > those terms; thus, we can accept a "common sense" definition of the > > terms. The only clarification offered is the statement that "The act of > > running the Program is not restricted"; this means that "copying, > > distribution, and modification" is not considered to be taking place > > when merely running the Program for the purposes of the license. > > I think that LPPL might have this clause as well. The poiunt is, my > "common sense" definition of distribution includes placing the program > in publicly accessed places. Good for you. But your common sense definition isn't written into the GPL, so it doesn't have legal force. If you license any software under the GPL, and you try to get someone for violating the license because that person installed the program without providing source, don't be surprised if you lose. The issue would be whether the user's "common sense" definition was a reasonable one, not whatever unwritten interpretation you happen to put on the GPL but refused to tell anyone. > > The LPPL, by contrast, has a different definition of "distribution": > > > > > Note that in the above, `distribution' of a file means making the > > > file available to others by any means. This includes, for instance, > > > installing the file on any machine in such a way that the file is > > > accessible by users other than yourself. > > > > These two definitions are incompatible, and the GPL's is manifestly less > > restricti
Re: defining 'distribution' (Re: A few more LPPL concerns)
On Mon, 2002-07-22 at 10:22, Joe Moore wrote: > Jeff Licquia wrote > > On Sun, 2002-07-21 at 23:10, Boris Veytsman wrote: > >> My /usr/local/bin can > >> be NFS-exported to hundreds of computers. Even my box can have > >> hundreds logins there. > > > > Yes, but in the former case, you are distributing the program to > > hundreds of computers. In the latter, hundreds of users are running > > the program. > > What about a clustered environment, such as Mosix? One virtual machine (the > cluster) consisting of many individual nodes. /usr/local/bin/latex can be > running on any of the nodes. > > Is that distribution? In regards to what? In the case of the LPPL, then yes, because it's explicitly defined that way. For licenses that don't explicitly define distribution (such as the GPL), then it would depend on a lot of factors I don't know. For example, to what extent do all of the machines on the cluster have a separate identity? I doubt such questions will be resolved, because it's just so much easier to provide the source than fight it out. > If I'm the only user on the cluster? > If there are 1000 students logging into the cluster? I don't think either of those facts are relevant to the GPL. Again, I quote: "The act of running the Program is not restricted..." > If the data is not shared over a network, is it distribution? For example, > a SAN-attached shared disk array that provides the same physical partition > to several systems? How is that different than AFS, NFS, or rsync? I don't know. Neither am I quite clear as to how such hair-splitting is relevant to my main point: the GPL is much clearer and less restrictive than the LPPL. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
> From: Jeff Licquia <[EMAIL PROTECTED]> > Date: 22 Jul 2002 00:47:39 -0500 > > On Sun, 2002-07-21 at 23:10, Boris Veytsman wrote: > > Exactly. I really do not see the difference between running a program > > from /usr/local/bin or /afs/whatever/bin/. What is the difference > > between AFS and NFS besides the technical one? > > There is none. In case you aren't aware, I feel NFS exporting > constitutes distribution as well. > Fine > > My /usr/local/bin can > > be NFS-exported to hundreds of computers. Even my box can have > > hundreds logins there. > > Yes, but in the former case, you are distributing the program to > hundreds of computers. In the latter, hundreds of users are running the > program. > I think this is bizarre. On my home network there are several x86 and several Sparcs. The x86s share /usr while the Sparcs have their individual /usr each. You are saying that something is legal to do on my Sparc but not on x86s? And if I take the decision to take a hard drive from the Sparc and put on the fileserver, the legality changes? > And, the GPL says: > > > The act of running the Program is not restricted... > > Therefore, the act of running the program from /usr/local/bin is not > restricted. I agree that *running* a program is not restricted -- both by GPL and LPPL. However, I think that when *I* run a program from /usr/local/bin, where *you* put it with the explicit intent to make it available for me to run, there are two acts involved. Namely, *you* distributed the program, and *I* run it. My actions are not restricted, but *yours* are restricted by GPL and LPPL. Again, imagine a company that put a closed-source derivative of a GPL'ed program on a company machine and refused to provide sources, but encouraged employees to run it for company business. Would you consider this legal? If not, you must recognize the act ot placing the executables in /usr/local/bin on a machine accessible by others to be what is it -- a distribution. > > The key is that the license has to spell it out. The GPL talks about > "copying, distribution, and modification", but sees fit to not define > those terms; thus, we can accept a "common sense" definition of the > terms. The only clarification offered is the statement that "The act of > running the Program is not restricted"; this means that "copying, > distribution, and modification" is not considered to be taking place > when merely running the Program for the purposes of the license. I think that LPPL might have this clause as well. The poiunt is, my "common sense" definition of distribution includes placing the program in publicly accessed places. > > The LPPL, by contrast, has a different definition of "distribution": > > > Note that in the above, `distribution' of a file means making the > > file available to others by any means. This includes, for instance, > > installing the file on any machine in such a way that the file is > > accessible by users other than yourself. > > These two definitions are incompatible, and the GPL's is manifestly less > restrictive than the LPPL's. > Again, I think thay are compatible. However, it is up to David and Frank to clarify this issue. -- Good luck -Boris Either I'm dead or my watch has stopped. -- Groucho Marx's last words -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining 'distribution' (Re: A few more LPPL concerns)
Jeff Licquia wrote > On Sun, 2002-07-21 at 23:10, Boris Veytsman wrote: >> > From: Jeff Licquia <[EMAIL PROTECTED]> >> > Date: 21 Jul 2002 22:59:26 -0500 >> > >> > It's crucial to your point, therefore, that there not be a >> > distinction between running the program from /usr/local/bin or >> > /afs/whatever/bin. I think we've shown that this isn't the case, >> > since a sysadmin does not need to give source to every person with a >> > login on his box, but does if he exports the program via AFS. >> >> Exactly. I really do not see the difference between running a program >> from /usr/local/bin or /afs/whatever/bin/. What is the difference >> between AFS and NFS besides the technical one? > > There is none. In case you aren't aware, I feel NFS exporting > constitutes distribution as well. > >> My /usr/local/bin can >> be NFS-exported to hundreds of computers. Even my box can have >> hundreds logins there. > > Yes, but in the former case, you are distributing the program to > hundreds of computers. In the latter, hundreds of users are running > the program. What about a clustered environment, such as Mosix? One virtual machine (the cluster) consisting of many individual nodes. /usr/local/bin/latex can be running on any of the nodes. Is that distribution? If I'm the only user on the cluster? If there are 1000 students logging into the cluster? If the data is not shared over a network, is it distribution? For example, a SAN-attached shared disk array that provides the same physical partition to several systems? How is that different than AFS, NFS, or rsync? --Joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
On Sunday 21 July 2002 22:59, Jeff Licquia wrote: > On Sun, 2002-07-21 at 22:40, Boris Veytsman wrote: > > I think that a sysadmin that put > > a changed copy of latex.fmt in the $TEXFORMATS directory to be used by > > his users, *distributes* a changed LaTeX. You think he does not; the > > problem with your theory is that it undermines both the intentions of > > LPPL AND GPL. > > The problem with this is that the GPL disagrees: > > > 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, and the output from the Program > > is covered only if its contents constitute a work based on the > > Program (independent of having been made by running the Program). > > It's crucial to your point, therefore, that there not be a distinction > between running the program from /usr/local/bin or /afs/whatever/bin. I > think we've shown that this isn't the case, since a sysadmin does not > need to give source to every person with a login on his box, but does if > he exports the program via AFS. It might be instructive to think about Bison here. Anyone who uses bison gets a substantial chunk of Bison's code. Is that distribution (when the user is not the same as or part of the owner of the system)? Yes, I know the doesn't have to be GPL'ed, that's not the question. The only question is whether a copyrighted work (part of the bison executable that happens to be a large chunk of text) is distributed when that new work is made. Given the macro-expansion nature of tex, the same consideration might apply to the latex files. Then again, it might just correspond to the case if bison's output were immediately compiled without ever being written to disk. Lynn -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
Jeff Licquia <[EMAIL PROTECTED]>: > > When I execute a program, this is not a distribution. When I allow > > others to execute it, I distribute it -- even if there is no actual > > copying of bits between magnetic media. > > Actually, it's not clear that this is true. For example, technically a > CD player "makes a copy" of a CD into its internal memory in order to > play the music encoded upon it. This is not considered to be "copying" > because this copy is required in order to gain access to the copyrighted > work you supposedly paid for when you bought the CD. In other words, > playing a CD is a "fair use" copying right, and I can execute that right > whether I'm the person who plunked down $20 for the particular CD or > not. "Fair use" normally means the copying of an excerpt from a work for the purposes of criticism, teaching, etc. I think the "copy" made by a CD player is not copying at all as far as most copyright law is concerned, but I don't have a definition of copying to hand. You should probably compare a CD player's transient copy with home video recording, or copying a CD onto tape so that you can listen to it in a car, or making a back-up copy of a tape, all of which are are allowed in most countries, as far as I know. Edmund -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
On Sun, 2002-07-21 at 23:10, Boris Veytsman wrote: > > From: Jeff Licquia <[EMAIL PROTECTED]> > > Date: 21 Jul 2002 22:59:26 -0500 > > > > It's crucial to your point, therefore, that there not be a distinction > > between running the program from /usr/local/bin or /afs/whatever/bin. I > > think we've shown that this isn't the case, since a sysadmin does not > > need to give source to every person with a login on his box, but does if > > he exports the program via AFS. > > Exactly. I really do not see the difference between running a program > from /usr/local/bin or /afs/whatever/bin/. What is the difference > between AFS and NFS besides the technical one? There is none. In case you aren't aware, I feel NFS exporting constitutes distribution as well. > My /usr/local/bin can > be NFS-exported to hundreds of computers. Even my box can have > hundreds logins there. Yes, but in the former case, you are distributing the program to hundreds of computers. In the latter, hundreds of users are running the program. And, the GPL says: > The act of running the Program is not restricted... Therefore, the act of running the program from /usr/local/bin is not restricted. If it helps, you can think of a network mount as the act of distribution. Or, perhaps, it would be more correct to say that the act of editing /etc/exports or /etc/samba/smb.conf or their AFS equivalent is the act of distribution, and the mount is the moment of receipt. > You cannot have legal definitions depending on technical details of > network protocols. Why not? "Distribution of this program is only allowed over the network protocol recognized by the Internet Engineering Task Force (IETF) as 'HTTP 1.1'." (Non-free, of course, but there you are.) > If something is illegal to share through AFS, it is > still illegal to share through NFS, FTP or HTTP. Legal definitions > *may* depend only on some high-level details of the protocols, like > whether the protocol allows anonymous exchange etc. Legal definitions may depend on absolutely anything the licensor decides they may depend on, besides those conditions forbidden by law. I can, as mentioned before, only allow distribution over HTTP, or only on CD-ROM, or only allow it within the USA, or only after I've been paid $99. This follows from my ability to forbid distribution entirely. The key is that the license has to spell it out. The GPL talks about "copying, distribution, and modification", but sees fit to not define those terms; thus, we can accept a "common sense" definition of the terms. The only clarification offered is the statement that "The act of running the Program is not restricted"; this means that "copying, distribution, and modification" is not considered to be taking place when merely running the Program for the purposes of the license. The LPPL, by contrast, has a different definition of "distribution": > Note that in the above, `distribution' of a file means making the > file available to others by any means. This includes, for instance, > installing the file on any machine in such a way that the file is > accessible by users other than yourself. These two definitions are incompatible, and the GPL's is manifestly less restrictive than the LPPL's. > If you contend that this sharing is illegal through AFS, then you need > to show me what features of AFS make it illegal and why they do not > apply to NFS or simple logging to a box. Otherwise your conclusion > will hold for these kinds of sharing too. For NFS, I agree. For logging on to a box, I will, again, point you to the clause in the GPL: > The act of running the Program is not restricted... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
> From: Jeff Licquia <[EMAIL PROTECTED]> > Date: 21 Jul 2002 22:59:26 -0500 > > It's crucial to your point, therefore, that there not be a distinction > between running the program from /usr/local/bin or /afs/whatever/bin. I > think we've shown that this isn't the case, since a sysadmin does not > need to give source to every person with a login on his box, but does if > he exports the program via AFS. > Exactly. I really do not see the difference between running a program from /usr/local/bin or /afs/whatever/bin/. What is the difference between AFS and NFS besides the technical one? My /usr/local/bin can be NFS-exported to hundreds of computers. Even my box can have hundreds logins there. You cannot have legal definitions depending on technical details of network protocols. If something is illegal to share through AFS, it is still illegal to share through NFS, FTP or HTTP. Legal definitions *may* depend only on some high-level details of the protocols, like whether the protocol allows anonymous exchange etc. If you contend that this sharing is illegal through AFS, then you need to show me what features of AFS make it illegal and why they do not apply to NFS or simple logging to a box. Otherwise your conclusion will hold for these kinds of sharing too. -- Good luck -Boris When you meet a master swordsman, show him your sword. When you meet a man who is not a poet, do not show him your poem. -- Rinzai, ninth century Zen master -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
On Sun, 2002-07-21 at 22:40, Boris Veytsman wrote: > > From: Jeff Licquia <[EMAIL PROTECTED]> > > Date: 21 Jul 2002 20:34:32 -0500 > > > You're right, and there may be software you can't install on your AFS > > drive in this instance, because you're "distributing" software to those > > thousand computers. This is irrespective of whether any of those > > thousand computers actually execute the software or not. > > > > This could even apply if you have a license to the software; it may be > > OK to install the software in /usr/local/bin, but not in > > /afs/campus/foobar/bin. > > > > For GPL software, you just have to make sure that any of the > > thousand-or-so computer owners can get to the source for those > > programs. The easiest way to do that would be to export the source > > trees via AFS; they could also stick a written offer in their student > > handbook, or whatever. > > > > Exactly. If you are installing GPL'ed software in this manner, you > must satisfy the distribution conditions of GPL. In the same manner > you must obey distribution conditions of TeX and LaTeX. > > My point is, LPPL or Knuth's license is no more restrictive *in this > respect* than a meaningful interpretatation of GPL. Your original point, way back in the thread, was: > I think that a sysadmin that put > a changed copy of latex.fmt in the $TEXFORMATS directory to be used by > his users, *distributes* a changed LaTeX. You think he does not; the > problem with your theory is that it undermines both the intentions of > LPPL AND GPL. The problem with this is that the GPL disagrees: > 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, and the output from the Program > is covered only if its contents constitute a work based on the > Program (independent of having been made by running the Program). It's crucial to your point, therefore, that there not be a distinction between running the program from /usr/local/bin or /afs/whatever/bin. I think we've shown that this isn't the case, since a sysadmin does not need to give source to every person with a login on his box, but does if he exports the program via AFS. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
> From: Jeff Licquia <[EMAIL PROTECTED]> > Date: 21 Jul 2002 20:34:32 -0500 > You're right, and there may be software you can't install on your AFS > drive in this instance, because you're "distributing" software to those > thousand computers. This is irrespective of whether any of those > thousand computers actually execute the software or not. > > This could even apply if you have a license to the software; it may be > OK to install the software in /usr/local/bin, but not in > /afs/campus/foobar/bin. > > For GPL software, you just have to make sure that any of the > thousand-or-so computer owners can get to the source for those > programs. The easiest way to do that would be to export the source > trees via AFS; they could also stick a written offer in their student > handbook, or whatever. > Exactly. If you are installing GPL'ed software in this manner, you must satisfy the distribution conditions of GPL. In the same manner you must obey distribution conditions of TeX and LaTeX. My point is, LPPL or Knuth's license is no more restrictive *in this respect* than a meaningful interpretatation of GPL. -- Good luck -Boris Extremism in the defense of liberty is no vice... moderation in the pursuit of justice is no virtue. -- Barry Goldwater -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
On Sun, 2002-07-21 at 20:18, Boris Veytsman wrote: > > From: Jeff Licquia <[EMAIL PROTECTED]> > > Date: 21 Jul 2002 18:07:50 -0500 > > > > On Sun, 2002-07-21 at 16:49, Boris Veytsman wrote: > > > This is the root of our disagreement. I think that a sysadmin that put > > > a changed copy of latex.fmt in the $TEXFORMATS directory to be used by > > > his users, *distributes* a changed LaTeX. You think he does not; the > > > problem with your theory is that it undermines both the intentions of > > > LPPL AND GPL. You see, there is no reasonable difference between a > > > sysadmin who put a closed copy of a GPL'ed program in /usr/bin, and a > > > cunning manager who made this program NFS-accessible "for execution > > > only" by the people paying a fee. > > > > Sure there is. In one case, a single computer can use the resulting > > binary; in the other, multiple computers can, after paying an access > > fee. > > What do you mean by "single computer"? A system administrator might > install a copy of a program on an AFS drive, where it can be executed > by any of thousand computers in a campus-wide network. Note that > people on this network might pay a fee to connect (e.g. tuition paid > by students in a university). > > If you allow this to be done, than anybody can easily circumvent GPL > by setting up a "virtual campus network" with the sole intention of > executing closed programs, effectively stolen from the GPL communinty, > by his paying customers. You're right, and there may be software you can't install on your AFS drive in this instance, because you're "distributing" software to those thousand computers. This is irrespective of whether any of those thousand computers actually execute the software or not. This could even apply if you have a license to the software; it may be OK to install the software in /usr/local/bin, but not in /afs/campus/foobar/bin. For GPL software, you just have to make sure that any of the thousand-or-so computer owners can get to the source for those programs. The easiest way to do that would be to export the source trees via AFS; they could also stick a written offer in their student handbook, or whatever. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
> From: Jeff Licquia <[EMAIL PROTECTED]> > Date: 21 Jul 2002 18:07:50 -0500 > > On Sun, 2002-07-21 at 16:49, Boris Veytsman wrote: > > This is the root of our disagreement. I think that a sysadmin that put > > a changed copy of latex.fmt in the $TEXFORMATS directory to be used by > > his users, *distributes* a changed LaTeX. You think he does not; the > > problem with your theory is that it undermines both the intentions of > > LPPL AND GPL. You see, there is no reasonable difference between a > > sysadmin who put a closed copy of a GPL'ed program in /usr/bin, and a > > cunning manager who made this program NFS-accessible "for execution > > only" by the people paying a fee. > > Sure there is. In one case, a single computer can use the resulting > binary; in the other, multiple computers can, after paying an access > fee. > What do you mean by "single computer"? A system administrator might install a copy of a program on an AFS drive, where it can be executed by any of thousand computers in a campus-wide network. Note that people on this network might pay a fee to connect (e.g. tuition paid by students in a university). If you allow this to be done, than anybody can easily circumvent GPL by setting up a "virtual campus network" with the sole intention of executing closed programs, effectively stolen from the GPL communinty, by his paying customers. -- Good luck -Boris In order to get a loan you must first prove you don't need it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
On Sun, 2002-07-21 at 16:49, Boris Veytsman wrote: > > Date: Sun, 21 Jul 2002 14:32:39 -0700 (PDT) > > From: Mark Rafn <[EMAIL PROTECTED]> > > > > > Suppose I take a GPL'ed program, change it and put the closed version > > > (sans sources) on my own machine. I did not violate GPL yet. Now > > > suppose that I make the drive NFS-exportable and encourage my paying > > > customers to mount it and access the program from there. Would I > > > violate GPL? I think yes. > > > > Absolutely. This is distribution in the classic sense. No additional > > definition of "distribution" is necessary. > > Note that in this case the customers might not copy the program but > just execute it from this location. This doesn't matter. The fact that you have allowed remote access, and the third party has taken advantage of it, makes it distribution. > > It's not distribution for me to install a package on a system I administer > > (or just have an account) and allow others to execute it. They can > > "access" it in terms of execution, but if they copy it, they do so without > > my permission (and without yours). > > This is the root of our disagreement. I think that a sysadmin that put > a changed copy of latex.fmt in the $TEXFORMATS directory to be used by > his users, *distributes* a changed LaTeX. You think he does not; the > problem with your theory is that it undermines both the intentions of > LPPL AND GPL. You see, there is no reasonable difference between a > sysadmin who put a closed copy of a GPL'ed program in /usr/bin, and a > cunning manager who made this program NFS-accessible "for execution > only" by the people paying a fee. Sure there is. In one case, a single computer can use the resulting binary; in the other, multiple computers can, after paying an access fee. > When I execute a program, this is not a distribution. When I allow > others to execute it, I distribute it -- even if there is no actual > copying of bits between magnetic media. Actually, it's not clear that this is true. For example, technically a CD player "makes a copy" of a CD into its internal memory in order to play the music encoded upon it. This is not considered to be "copying" because this copy is required in order to gain access to the copyrighted work you supposedly paid for when you bought the CD. In other words, playing a CD is a "fair use" copying right, and I can execute that right whether I'm the person who plunked down $20 for the particular CD or not. (Of course, the whole mess that is music distribution rights is way beyond this discussion.) This is different from the so-called "clickwrap" licenses because you generally have to agree to those at *installation* time, not every time you run the program. Of course, clickwrap licenses have their own problems. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
On Sun, Jul 21, 2002 at 05:49:34PM -0400, Boris Veytsman wrote: > This is the root of our disagreement. I think that a sysadmin that put > a changed copy of latex.fmt in the $TEXFORMATS directory to be used by > his users, *distributes* a changed LaTeX. You think he does not; the > problem with your theory is that it undermines both the intentions of > LPPL AND GPL. You see, there is no reasonable difference between a > sysadmin who put a closed copy of a GPL'ed program in /usr/bin, and a > cunning manager who made this program NFS-accessible "for execution > only" by the people paying a fee. > > When I execute a program, this is not a distribution. When I allow > others to execute it, I distribute it -- even if there is no actual > copying of bits between magnetic media. Um, no. This is commonly referred to as "use", not "distribution". The "cunning manager" is also not distributing, unless the intention of his placing the stuff there was in reality to allow people to download a modified LaTeX. At which point we are into the realms of expensive lawyers arguing the toss over his state of mind at the time he put it up. How the hell do you expect a company who wish to use a modified version of LaTeX on their server to do so if you define the sysadmin's installing it as "distribution to the users", as you do above? Thank goodness that does not appear to be Frank or David's interpretation of things... Cheers, Nick -- Nick Phillips -- [EMAIL PROTECTED] Slow day. Practice crawling. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: defining "distribution" (Re: A few more LPPL concerns)
> Date: Sun, 21 Jul 2002 14:32:39 -0700 (PDT) > From: Mark Rafn <[EMAIL PROTECTED]> > > > Suppose I take a GPL'ed program, change it and put the closed version > > (sans sources) on my own machine. I did not violate GPL yet. Now > > suppose that I make the drive NFS-exportable and encourage my paying > > customers to mount it and access the program from there. Would I > > violate GPL? I think yes. > > Absolutely. This is distribution in the classic sense. No additional > definition of "distribution" is necessary. > Note that in this case the customers might not copy the program but just execute it from this location. > > It's not distribution for me to install a package on a system I administer > (or just have an account) and allow others to execute it. They can > "access" it in terms of execution, but if they copy it, they do so without > my permission (and without yours). This is the root of our disagreement. I think that a sysadmin that put a changed copy of latex.fmt in the $TEXFORMATS directory to be used by his users, *distributes* a changed LaTeX. You think he does not; the problem with your theory is that it undermines both the intentions of LPPL AND GPL. You see, there is no reasonable difference between a sysadmin who put a closed copy of a GPL'ed program in /usr/bin, and a cunning manager who made this program NFS-accessible "for execution only" by the people paying a fee. When I execute a program, this is not a distribution. When I allow others to execute it, I distribute it -- even if there is no actual copying of bits between magnetic media. -- Good luck -Boris The game of life is a game of boomerangs. Our thoughts, deeds and words return to us sooner or later with astounding accuracy. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]