Greetings all, Some time ago the topic of changing the version/flavor of GPL license Koha is released under was discussed. In light of the momentum the community has built up in the area of change in recent months, now would be a good time to resume licensing discussions and bring them to a final resolve. Delaying them only means another round of disruptive changes at some point in the not too distant future. It would be nice to go ahead and have all the laundry cleaned and folded while we are about the task. As I see it, the advantage and rational for moving to GPLv3 are primarily that GPLv3 is compatible with AGPLv3. This allows us to accept work licensed under either of these two licenses. AGPLv3 carries the added advantage of the *additional* requirement that changes, etc. to the code be made available (at least) to those who access the changed code. This gives us quite a measure of protection more against Koha related code becoming locked up on a saas platform somewhere in the nebulous "cloud." Let's hash through this and come to a decision. Kind Regards, Chris
Christopher Nighswonger wrote:
In light of the momentum the community has built up in the area of change in recent months, now would be a good time to resume licensing discussions and bring them to a final resolve.
I cannot contribute fully to this discussion at this time. Please can we try to limit the number of simultaneous disruptive changes? As you know, we have 3.2 releasing, new websites and urgent trademark challenges which I'd like to contribute to and my work is already spread a bit too thin. Please can we leave this until after 3.2.0? Is there code under *GPLv3 which we want to integrate imminently? [...]
AGPLv3 carries the added advantage of the *additional* requirement that changes, etc. to the code be made available (at least) to those who access the changed code. This gives us quite a measure of protection more against Koha related code becoming locked up on a saas platform somewhere in the nebulous "cloud."
No, that additional requirement is one of the disadvantages. AGPLv3 is still a fairly new licence and as far as I know, it's not yet entirely clear what is meant by "access", particularly for a modular system like Koha. It might mean that everyone who sees a page is entitled to the 227Mb source tarball. Who wants to pay for those downloads? We asked questions about this and other vagueness to FSF during the AGPL drafting and the questions were never answered, as far as I know. The drafting process was hampered by use of a buggy, inaccessible and undocumented application called "stet" instead of a nice easy wiki (yes, mediawiki would have been better than stet). Fundamentally, AGPLv3 is based on an absurd idea that one can "ensure cooperation with the community" (source: AGPLv3 preamble). However, cooperation by definition must be voluntary (source: ICA.coop/coop/principles.html ) so legal compulsion is not cooperation. To use a more common phrase, you can lead a horse to water but you cannot make it drink. The GPL already leads us to the water. AGPL is an attempt to make all drink and is doomed to fail. If we force code publication, bad vendors will simply lock the database, keep passwords secret, move some code off to the other side of APIs and so on. Maybe even do other tricks to hamper use of the published code. Meanwhile, we'd be making Koha more difficult and more costly for friendly people to use, by using a less understood license with more requirements to fulfil. So, we hurt our friends without stopping our enemies. Why do it? In the past, I've advocated the share-alike type provisions of the GPL, as well as using BSD-style share-enabled terms, but I feel AGPLv3 share-forced terms go against other vital principles of freedom. So may we postpone the rest of this discussion to post-3.2.0? Regards, -- MJ Ray (slef) Webmaster and LMS developer at | software www.software.coop http://mjr.towers.org.uk | .... co IMO only: see http://mjr.towers.org.uk/email.html | .... op
Hi, On Mon, May 10, 2010 at 9:47 AM, MJ Ray <mjr@phonecoop.coop> wrote:
Christopher Nighswonger wrote:
In light of the momentum the community has built up in the area of change in recent months, now would be a good time to resume licensing discussions and bring them to a final resolve.
I cannot contribute fully to this discussion at this time. Please can we try to limit the number of simultaneous disruptive changes? As you know, we have 3.2 releasing, new websites and urgent trademark challenges which I'd like to contribute to and my work is already spread a bit too thin. Please can we leave this until after 3.2.0?
I, for one, do not want to attempt relicensing prior to release of 3.2.0 or as a condition for release of that version. Personally, while I am currently in favor of relicensing to GPL3+ (assuming that we can resolve the matter of dependencies on GPL2 code such as OpenNCIP) and think that the AGPL3 should be considered as well, it would be too disruptive to even think of doing this for 3.2. However, since the proposal is relevant for 3.4, I do think that now is as a good a time as any to start a discussion of this issue. Regards, Galen -- Galen Charlton gmcharlt@gmail.com
On Mon, May 10, 2010 at 10:06 AM, Galen Charlton <gmcharlt@gmail.com> wrote:
Hi,
On Mon, May 10, 2010 at 9:47 AM, MJ Ray <mjr@phonecoop.coop> wrote:
Christopher Nighswonger wrote:
In light of the momentum the community has built up in the area of change in recent months, now would be a good time to resume licensing discussions and bring them to a final resolve.
I cannot contribute fully to this discussion at this time. Please can we try to limit the number of simultaneous disruptive changes? As you know, we have 3.2 releasing, new websites and urgent trademark challenges which I'd like to contribute to and my work is already spread a bit too thin. Please can we leave this until after 3.2.0?
I, for one, do not want to attempt relicensing prior to release of 3.2.0 or as a condition for release of that version. Personally, while I am currently in favor of relicensing to GPL3+ (assuming that we can resolve the matter of dependencies on GPL2 code such as OpenNCIP) and think that the AGPL3 should be considered as well, it would be too disruptive to even think of doing this for 3.2.
I also concur that this is not a change to be made in 3.2.
However, since the proposal is relevant for 3.4, I do think that now is as a good a time as any to start a discussion of this issue.
Ditto. Kind Regards, Chris
On Mon, May 10, 2010 at 9:47 AM, MJ Ray <mjr@phonecoop.coop> wrote:
No, that additional requirement is one of the disadvantages. AGPLv3 is still a fairly new licence and as far as I know, it's not yet entirely clear what is meant by "access", particularly for a modular system like Koha. It might mean that everyone who sees a page is entitled to the 227Mb source tarball. Who wants to pay for those downloads?
Github and the like provide very simple solutions to this problem both with ease of administration and ease of cost. AGPLv3 does not specify (see my previous response to Lars' post.) Besides, there are hosting plans available that provide unlimited bandwidth for very, very few $$$ per month if the FTP route was a necessity. <snip>
Fundamentally, AGPLv3 is based on an absurd idea that one can "ensure cooperation with the community" (source: AGPLv3 preamble). However, cooperation by definition must be voluntary (source: ICA.coop/coop/principles.html ) so legal compulsion is not cooperation.
A careful reading of the second paragraph of the Preamble of GPLv2 (the current Koha license) will reveal the fact that the entire purpose of the license is to ensure cooperation with the goal of ensuring at a minimum three things: "...the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs..." It appears that any form of licensing is an attempt to ensure cooperation of some sort among some people. *All* licensing is, in fact, some form of coercion, period. The unfortunate fact of life is that there is somebody, somewhere who will do wrong even if you will not. It would be wonderful if it were otherwise. The we would not need licenses... or laws for that matter. By adopting *any* sort of license, then, we are "ensuring cooperation" at some level. The entire purpose of any GNU license is to provide "legal compulsion" of those who would use FOSS to abide by the wishes of the copyright holder regarding that code. FOSS would patently fail if there were no way to enforce the desire for the code to be free and open source. So the argument against the use of legal compulsion is really self-defeating in this instance. We cannot license *and* avoid the use of "legal compulsion."
So may we postpone the rest of this discussion to post-3.2.0?
As I stated in my original proposal: We are already very active atm, and now is the time to at least begin discussing this change. Kind Regards, Chris
Christopher Nighswonger wrote:
On Mon, May 10, 2010 at 9:47 AM, MJ Ray <mjr@phonecoop.coop> wrote:
[...] It might mean that everyone who sees a page is entitled to the 227Mb source tarball. Who wants to pay for those downloads?
Github and the like provide very simple solutions to this problem both with ease of administration and ease of cost. AGPLv3 does not specify (see my previous response to Lars' post.) Besides, there are hosting plans available that provide unlimited bandwidth for very, very few $$$ per month if the FTP route was a necessity.
Firstly, it seems unethical to impose even "very, very few $$$ per month" of extra cost on charitable libraries. Secondly, this leads to another of what I think is still one of the Great Unknowns of AGPLv3: if you don't host the source alongside, must the app go offline if it thinks the source has gone offline? There are so many of these lawyerbombs around AGPLv3 that I feel the whole thing is best avoided by sticking with GPLv2, at least until others have trod on some of the big ones, in the absence of any pressing need to switch.
Fundamentally, AGPLv3 is based on an absurd idea that one can "ensure cooperation with the community" (source: AGPLv3 preamble). However, cooperation by definition must be voluntary (source: ICA.coop/coop/principles.html ) so legal compulsion is not cooperation. [...] It appears that any form of licensing is an attempt to ensure cooperation of some sort among some people. *All* licensing is, in fact, some form of coercion, period. The unfortunate fact of life is
<snip> that there is somebody, somewhere who will do wrong even if you will not. It would be wonderful if it were otherwise. The we would not need licenses... or laws for that matter. [...]
I don't see how that's true, unless "cooperation of some sort" means something other than cooperation, such as mere trading. Free software licensing is usually just setting out the terms of trade, but AGPLv3's clod-handed attempts to force public sharing go beyond it. Laws and licences do not prevent anyone doing wrong. I'm sure all of us are quite capable of breaching a copyright licence without much work. If anyone is hoping that adopting AGPLv3 will prevent bad people refusing to share progress with the community, you are doomed to fail. So what is the burning desire for AGPLv3?
So may we postpone the rest of this discussion to post-3.2.0?
As I stated in my original proposal: We are already very active atm, and now is the time to at least begin discussing this change.
I am disappointed by this desire to press ahead with holding a discussion of such a complex topic at such a busy time. It will limit participation and likely leave the discussion incomplete. Regards, -- MJ Ray (slef) Webmaster and LMS developer at | software www.software.coop http://mjr.towers.org.uk | .... co IMO only: see http://mjr.towers.org.uk/email.html | .... op
On Mon, May 10, 2010 at 9:54 PM, MJ Ray <mjr@phonecoop.coop> wrote:
Christopher Nighswonger wrote:
On Mon, May 10, 2010 at 9:47 AM, MJ Ray <mjr@phonecoop.coop> wrote:
[...] It might mean that everyone who sees a page is entitled to the 227Mb source tarball. Who wants to pay for those downloads?
Github and the like provide very simple solutions to this problem both with ease of administration and ease of cost. AGPLv3 does not specify (see my previous response to Lars' post.) Besides, there are hosting plans available that provide unlimited bandwidth for very, very few $$$ per month if the FTP route was a necessity.
Firstly, it seems unethical to impose even "very, very few $$$ per month" of extra cost on charitable libraries.
In that case, github has a free offering... no $$$ per any time increment.
Secondly, this leads to another of what I think is still one of the Great Unknowns of AGPLv3: if you don't host the source alongside, must the app go offline if it thinks the source has gone offline?
Where in the license does it say anything about hosting source alongside the application?
There are so many of these lawyerbombs around AGPLv3 that I feel the whole thing is best avoided by sticking with GPLv2, at least until others have trod on some of the big ones, in the absence of any pressing need to switch.
GPLvX was virtually untested in court until Progress Software Corp. v. MySQL began back in 2002. Apparently that line of reasoning did not stop the originators of Koha from selecting GPLv2+ when they released it in 2000.
Fundamentally, AGPLv3 is based on an absurd idea that one can "ensure cooperation with the community" (source: AGPLv3 preamble). However, cooperation by definition must be voluntary (source: ICA.coop/coop/principles.html ) so legal compulsion is not cooperation. [...] It appears that any form of licensing is an attempt to ensure cooperation of some sort among some people. *All* licensing is, in fact, some form of coercion, period. The unfortunate fact of life is
<snip> that there is somebody, somewhere who will do wrong even if you will not. It would be wonderful if it were otherwise. The we would not need licenses... or laws for that matter. [...]
I don't see how that's true, unless "cooperation of some sort" means something other than cooperation, such as mere trading. Free software licensing is usually just setting out the terms of trade, but AGPLv3's clod-handed attempts to force public sharing go beyond it.
View it how you will. In the case you propose, we are simply establishing an additional clarification to the terms of trade. The intent of GPL licenses is to preserve the open nature of code released as free and open source. Not only is it the right of the receiver to have access to the code, etc. but it is also the right of the author to have their intent that the code be free and open respected. The terms of the trade are such that in return for the right to use, etc. this code, you agree to release any changes you make to said parties. As I pointed out previously: All licensing agreements "force" some points. Just violate one and see how quickly it is en-forced. (Pun intended.) It actually seems to me that the matter of not forcing cooperation is in contradiction to the expressed desire to force cooperation in the matter of trademarks, etc. by de-listing, etc. Cooperation is, after all, cooperation.
Laws and licences do not prevent anyone doing wrong. I'm sure all of us are quite capable of breaching a copyright licence without much work.
Quite right. If laws and licenses did that, we'd only need a few to fix all problems. What they do do is give recourse for redressing wrongs.
If anyone is hoping that adopting AGPLv3 will prevent bad people refusing to share progress with the community, you are doomed to fail.
I don't think anyone is under that illusion. As I said, "The unfortunate fact of life is that there is somebody, somewhere who will do wrong even if you will not."
So what is the burning desire for AGPLv3?
To put in place a mechanism for redressing violations of terms of license. We can chatter away about how a company reneged on its verbal obligations to contribute code back to the community,but we have nothing to fall back upon with any (even hopeful) force of law. It is not a panacea for all ills, only another hopeful obstacle in the path of misbehavior.
So may we postpone the rest of this discussion to post-3.2.0?
As I stated in my original proposal: We are already very active atm, and now is the time to at least begin discussing this change.
I am disappointed by this desire to press ahead with holding a discussion of such a complex topic at such a busy time. It will limit participation and likely leave the discussion incomplete.
Participation will only be limited if we want it to. No one involved in this project is not busy. If we believe that we have a good thing going in Koha, we must make the time to do the not so nice parts of the project too. Kind Regards, Chris
On Mon, 2010-05-10 at 23:14 -0400, Chris Nighswonger wrote:
So may we postpone the rest of this discussion to post-3.2.0?
As I stated in my original proposal: We are already very active atm, and now is the time to at least begin discussing this change.
I am disappointed by this desire to press ahead with holding a discussion of such a complex topic at such a busy time. It will limit participation and likely leave the discussion incomplete.
Participation will only be limited if we want it to. No one involved in this project is not busy. If we believe that we have a good thing going in Koha, we must make the time to do the not so nice parts of the project too.
This would be true if time were not a finite resource. Unfortunately it is. As much as we would like to have more time in a day it is simply not possible. I agree with Galen; this is an unusually busy time for Koha developers because of the impending release. I think that this is not the best time for this discussion. -- Michael Hafen Systems Analyst and Programmer Washington County School District Utah, USA for Koha checkout http://development.washk12.org/gitweb/ or git://development.washk12.org/koha
Hi, On Tue, May 11, 2010 at 10:36 AM, Michael Hafen <mdhafen@tech.washk12.org> wrote:
This would be true if time were not a finite resource. Unfortunately it is. As much as we would like to have more time in a day it is simply not possible. I agree with Galen; this is an unusually busy time for Koha developers because of the impending release. I think that this is not the best time for this discussion.
To clarify, I am not calling for an end to this discussion, nor have I done so. I think that now is an appropriate time to talk about the licensing in preparation for 3.4, although if active developers want to send a bugfix for each email they add to this thread, I won't complain. Regards, Galen -- Galen Charlton gmcharlt@gmail.com
Given that this discussion has been somewhat carried on through the wiki (http://wiki.koha-community.org/wiki/General_Meeting,_June_2_2010), I thought it would be good to keep it on list as well. MJ posts the following argument to the above referenced agenda:
* AGPL does not prevent unfriendly vendor lock-in (still achievable through access control, particularly to the databases) while introducing onerous burdens on friendly hosters. http://lists.katipo.co.nz/pipermail/koha/2010-May/023816.html
This would probably be a good time to point out that 3.4 will have the capability of doing a total db dump from within the staff client. This should greatly mitigate the potential of an hostile vendor holding data hostage. A vendor would have to deliberately remove or disable this functionality to prevent a client from retrieving a full copy of their database. (Of course any client who contracts with a vendor *without* writing data protection assurances into their contract is asking for trouble to start with.) The matter of "onerous burdens on friendly hosters" has been addressed in previous communications and so those arguments are not repeated here. Kind Regards, Chris On Mon, May 10, 2010 at 11:14 PM, Chris Nighswonger <cnighswonger@foundations.edu> wrote:
On Mon, May 10, 2010 at 9:54 PM, MJ Ray <mjr@phonecoop.coop> wrote:
Christopher Nighswonger wrote:
On Mon, May 10, 2010 at 9:47 AM, MJ Ray <mjr@phonecoop.coop> wrote:
[...] It might mean that everyone who sees a page is entitled to the 227Mb source tarball. Who wants to pay for those downloads?
Github and the like provide very simple solutions to this problem both with ease of administration and ease of cost. AGPLv3 does not specify (see my previous response to Lars' post.) Besides, there are hosting plans available that provide unlimited bandwidth for very, very few $$$ per month if the FTP route was a necessity.
Firstly, it seems unethical to impose even "very, very few $$$ per month" of extra cost on charitable libraries.
In that case, github has a free offering... no $$$ per any time increment.
Secondly, this leads to another of what I think is still one of the Great Unknowns of AGPLv3: if you don't host the source alongside, must the app go offline if it thinks the source has gone offline?
Where in the license does it say anything about hosting source alongside the application?
There are so many of these lawyerbombs around AGPLv3 that I feel the whole thing is best avoided by sticking with GPLv2, at least until others have trod on some of the big ones, in the absence of any pressing need to switch.
GPLvX was virtually untested in court until Progress Software Corp. v. MySQL began back in 2002. Apparently that line of reasoning did not stop the originators of Koha from selecting GPLv2+ when they released it in 2000.
Fundamentally, AGPLv3 is based on an absurd idea that one can "ensure cooperation with the community" (source: AGPLv3 preamble). However, cooperation by definition must be voluntary (source: ICA.coop/coop/principles.html ) so legal compulsion is not cooperation. [...] It appears that any form of licensing is an attempt to ensure cooperation of some sort among some people. *All* licensing is, in fact, some form of coercion, period. The unfortunate fact of life is
<snip> that there is somebody, somewhere who will do wrong even if you will not. It would be wonderful if it were otherwise. The we would not need licenses... or laws for that matter. [...]
I don't see how that's true, unless "cooperation of some sort" means something other than cooperation, such as mere trading. Free software licensing is usually just setting out the terms of trade, but AGPLv3's clod-handed attempts to force public sharing go beyond it.
View it how you will. In the case you propose, we are simply establishing an additional clarification to the terms of trade. The intent of GPL licenses is to preserve the open nature of code released as free and open source. Not only is it the right of the receiver to have access to the code, etc. but it is also the right of the author to have their intent that the code be free and open respected. The terms of the trade are such that in return for the right to use, etc. this code, you agree to release any changes you make to said parties.
As I pointed out previously: All licensing agreements "force" some points. Just violate one and see how quickly it is en-forced. (Pun intended.)
It actually seems to me that the matter of not forcing cooperation is in contradiction to the expressed desire to force cooperation in the matter of trademarks, etc. by de-listing, etc. Cooperation is, after all, cooperation.
Laws and licences do not prevent anyone doing wrong. I'm sure all of us are quite capable of breaching a copyright licence without much work.
Quite right. If laws and licenses did that, we'd only need a few to fix all problems. What they do do is give recourse for redressing wrongs.
If anyone is hoping that adopting AGPLv3 will prevent bad people refusing to share progress with the community, you are doomed to fail.
I don't think anyone is under that illusion. As I said, "The unfortunate fact of life is that there is somebody, somewhere who will do wrong even if you will not."
So what is the burning desire for AGPLv3?
To put in place a mechanism for redressing violations of terms of license. We can chatter away about how a company reneged on its verbal obligations to contribute code back to the community,but we have nothing to fall back upon with any (even hopeful) force of law. It is not a panacea for all ills, only another hopeful obstacle in the path of misbehavior.
So may we postpone the rest of this discussion to post-3.2.0?
As I stated in my original proposal: We are already very active atm, and now is the time to at least begin discussing this change.
I am disappointed by this desire to press ahead with holding a discussion of such a complex topic at such a busy time. It will limit participation and likely leave the discussion incomplete.
Participation will only be limited if we want it to. No one involved in this project is not busy. If we believe that we have a good thing going in Koha, we must make the time to do the not so nice parts of the project too.
Kind Regards, Chris
On ke, 2010-05-19 at 07:23 -0400, Chris Nighswonger wrote:
Given that this discussion has been somewhat carried on through the wiki (http://wiki.koha-community.org/wiki/General_Meeting,_June_2_2010), I thought it would be good to keep it on list as well.
For what it's worth, I agree: it's better if discussion is kept in one place, so everyone can easily follow it.
This would probably be a good time to point out that 3.4 will have the capability of doing a total db dump from within the staff client.
As it happens, I made a proof-of-concept implementation of this yesterday. It relies on a cron job to call mysqldump; possibly an on-demand dump might be better.
(Koha main list dropped from cc) Chris Nighswonger <cnighswonger@foundations.edu> wrote:
On Mon, May 10, 2010 at 9:54 PM, MJ Ray <mjr@phonecoop.coop> wrote:
Secondly, this leads to another of what I think is still one of the Great Unknowns of AGPLv3: if you don't host the source alongside, must the app go offline if it thinks the source has gone offline?
Where in the license does it say anything about hosting source alongside the application?
Nowhere, but it does say "your modified version must prominently offer all users interacting with it remotely through a computer network (if your version supports such interaction) an opportunity to receive the Corresponding Source". If the source has gone offline, there is no opportunity to receive it, is there?
There are so many of these lawyerbombs around AGPLv3 that I feel the whole thing is best avoided by sticking with GPLv2, at least until others have trod on some of the big ones, in the absence of any pressing need to switch.
GPLvX was virtually untested in court until Progress Software Corp. v. MySQL began back in 2002. Apparently that line of reasoning did not stop the originators of Koha from selecting GPLv2+ when they released it in 2000.
There are very few bits of language in GPLv2 which are anything like as vague and confusing as the AGPLv3 clause. [...]
I don't see how that's true, unless "cooperation of some sort" means something other than cooperation, such as mere trading. Free software licensing is usually just setting out the terms of trade, but AGPLv3's clod-handed attempts to force public sharing go beyond it.
View it how you will. In the case you propose, we are simply establishing an additional clarification to the terms of trade. The intent of GPL licenses is to preserve the open nature of code released as free and open source. Not only is it the right of the receiver to have access to the code, etc. but it is also the right of the author to have their intent that the code be free and open respected. The terms of the trade are such that in return for the right to use, etc. this code, you agree to release any changes you make to said parties.
Forced publication was never part of the free and open nature before. That is not a clarification. It's a fundamental shift between user freedoms and author freedoms, driven by a reactionary approach to unfriendly users who lend their computing resources to others.
As I pointed out previously: All licensing agreements "force" some points. Just violate one and see how quickly it is en-forced. (Pun intended.)
It's actually quite rare. When it does happen, it is painful, as I discovered to my cost in a past job, but it is still rare.
It actually seems to me that the matter of not forcing cooperation is in contradiction to the expressed desire to force cooperation in the matter of trademarks, etc. by de-listing, etc. Cooperation is, after all, cooperation.
Trademarks are not copyright and trying to treat them exactly the same will lead to some very unhealthy conclusions. We can look for the same freedoms (use, adapt, study and share) but in different ways. Also, I'm pretty sure I've never expressed any desire to force cooperation. [...]
So what is the burning desire for AGPLv3? [...] It is not a panacea for all ills, only another hopeful obstacle in the path of misbehavior.
This community has so far been relatively unwilling to activate the obstacles it already has placed in that path. Another one will only serve to deter and scare its friends further, because the only people who heed the obstacles are those who respect its wishes. Furthermore, this isn't even an obstacle in the path of the bad hosters: "One problem which the GNU Affero GPL does not address is the problem of Software as a Service (SaaS). It is impossible, as far as we know, to address this problem with a software license." -- http://www.gnu.org/licenses/why-affero-gpl.html
So may we postpone the rest of this discussion to post-3.2.0?
As I stated in my original proposal: We are already very active atm, and now is the time to at least begin discussing this change.
I am disappointed by this desire to press ahead with holding a discussion of such a complex topic at such a busy time. It will limit participation and likely leave the discussion incomplete.
Participation will only be limited if we want it to. No one involved in this project is not busy. If we believe that we have a good thing going in Koha, we must make the time to do the not so nice parts of the project too.
There is no point adding not so nice parts that bring no benefit unless you think this is an S+M community or something. Instead, let's find ways to make Koha more fun and more general! Regards, -- MJ Ray (slef) Webmaster and LMS developer at | software www.software.coop http://mjr.towers.org.uk | .... co IMO only: see http://mjr.towers.org.uk/email.html | .... op
On May 9, 2010, at 5:18 PM, Christopher Nighswonger wrote:
As I see it, the advantage and rational for moving to GPLv3 are primarily that GPLv3 is compatible with AGPLv3. This allows us to accept work licensed under either of these two licenses.
I don't know if that is true. From what I understand you can't mix GPLv3 and AGPLv3 code. You can link to libraries that are compiled with GPLv3 in AGPLv3 (or vice versa), but the code can't be mixed in the same code base. And AGPLv3 does not allow linked of GPLv2 libraries. Given these difficulties, I'd much rather stay with GPLv2. -Matt -- Matthew Butch Technology Support Analyst Penn Manor School District 717-872-9500 x 2385 http://www.pennmanor.net
On Tue, May 11, 2010 at 10:21 AM, Matthew Butch <matt@pennmanor.net> wrote:
On May 9, 2010, at 5:18 PM, Christopher Nighswonger wrote:
As I see it, the advantage and rational for moving to GPLv3 are primarily that GPLv3 is compatible with AGPLv3. This allows us to accept work licensed under either of these two licenses.
I don't know if that is true. From what I understand you can't mix GPLv3 and AGPLv3 code. You can link to libraries that are compiled with GPLv3 in AGPLv3 (or vice versa), but the code can't be mixed in the same code base. And AGPLv3 does not allow linked of GPLv2 libraries.
Combining is specifically permitted. Paragraph 13 of AGPLv3 clearly states (emphasis mine): "Notwithstanding any other provision of this License, you have permission to link **or combine** any covered work with a work licensed under version 3 of the GNU General Public License into a single combined work, and to convey the resulting work. The terms of this License will continue to apply to the part which is the covered work, but the work with which it is combined will remain governed by version 3 of the GNU General Public License." http://www.gnu.org/licenses/agpl-3.0.html Kind Regards, Chris
On Tue, 11 May 2010, Matthew Butch wrote:
On May 9, 2010, at 5:18 PM, Christopher Nighswonger wrote:
As I see it, the advantage and rational for moving to GPLv3 are primarily that GPLv3 is compatible with AGPLv3. This allows us to accept work licensed under either of these two licenses.
I don't know if that is true. From what I understand you can't mix GPLv3 and AGPLv3 code. You can link to libraries that are compiled with GPLv3 in AGPLv3 (or vice versa), but the code can't be mixed in the same code base. And AGPLv3 does not allow linked of GPLv2 libraries.
You are correct that you cannot mix GPLv2 only and AGPL or GPLv3 code. You can mix GPLv3 and AGPLv3 code, the result is AGPLv3 As noted in the license, this does not change the code in the original package, but if you make any changes and don't explicitly dual-license them then the resulting changes cannot be merged back into the GPLv3 codebase. David Lang
participants (8)
-
Chris Nighswonger -
Christopher Nighswonger -
david＠lang.hm -
Galen Charlton -
Lars Wirzenius -
Matthew Butch -
Michael Hafen -
MJ Ray