This article is reproduced from: yatennosyo , author : tisonkun
Original link: https://mp.weixin.qq.com/s/FsolP0IPIq0pFolPVDD8Ug
Today's content is to carry an old article written before, Copyleft is not a virus, the following text.
Today, I read the words defending the copyleft open source agreement in the book " Open Source for Business[1] " written by the famous open source agreement lawyer Heather Meeker. I deeply feel that in the past few decades, the open source agreement has been slandered as a virus, and it has been so-called contagion at every turn. The ethos of sexual intimidation of users is unreasonable. Therefore, this article is written to reiterate that the Copyleft open source agreement is not a virus on the basis of the opinions in the book.
Maybe people who grew up in the information age no longer remember the grievances between the free software movement and a company at the end of the last century, but the labels that the two sides put on each other in the debate at that time had a profound impact. The most widely circulated among them is the software licensed under the Copyleft open source agreement represented by GPL, which is characterized as a virus or a source of infection.
The copyleft open source agreement requires that the original agreement must be kept intact when distributing software licensed under the agreement and its derivative works. That is to say, after you get a Linux source code licensed under the GPLv2 agreement, if you want to redistribute the obtained Linux source code, or make modifications on its basis and then distribute its revision or derivative work, then you Must follow the GPLv2 license. Because copyleft agreements such as GPLv2 grant those who receive the source code the right to run, modify, and redistribute the source code freely, free software is guaranteed to remain free as it circulates.
Those who characterize software licensed under copyleft open-source agreements as viruses or sources of infection are attacking the above requirements for derivative works. Taking the GPL as an example, if a piece of software is a combination of GPL-licensed code and other code, then it is a derivative work of the GPL-licensed code. The GPL requires that in this case, all code of the entire software should be licensed under the GPL so that it can be freely run, modified, and redistributed. The argument sounds like GPL-licensed code would "turn" all code that integrates with it into GPL-licensed code, and thus, misinterpreted, would force your proprietary code or trade secrets to be made public. This is the accusation that the copyleft open source protocol is a virus.
However, is this really the case?
Heather Meeker refuted this claim based on her extensive legal experience and criticized the long-standing stigma of open source protocols by lawyers. She mentioned,
In fact, if a company were to combine GPL and proprietary code in a way that violated GPL, the result would be that GPL had been violated - no more, no less.
That is to say, the result of violating the GPL is that the GPL is violated. Does this have the taste of the recently popular "listen to your words, listen to your words"?
Legally, this means that authors of GPL software can be held accountable for their work being used in violation of copyright law, and if the GPL is terminated due to non-compliance by the user, authors can take action against unauthorized use of the GPL software measure. From the actual effect, the former means that the infringer needs to compensate, and the latter means that the infringer needs to stop using the GPL software.
As you can see, there is no accountability for other code, and certainly no statement that proprietary code or trade secrets must be disclosed, nor is it possible to force changes to proprietary code protocols.
In other words, unless you intend to continue to legally use the GPL software, in which case you can comply with the GPL's requirements by releasing the other integrated code under the GPL license, at most you'll be required to pay damages for the infringement, and stop the infringement behavior. This is a perfectly normal requirement. It should not be labeled as a virus or contagion, and there is no mandatory meaning in it.
Heather Meeker mentioned in the original text that this is just a case of incompatibility between the copyleft open source protocol and the proprietary protocol, not a case of one assimilating the other. An apt metaphor is a software conflict, not a software virus.
Sincerely said.
The above text, talk about the recent new insights.
There are still people who ask whether the self-developed software will be infected after the introduction of GPLv3 licensed software. As mentioned above, the term contagion is misleading. For this word, I thought of two better alternatives. One is that the software is affected by the terms of the GPLv3, and the other is that the software is also required to be released under the GPLv3 agreement.
The impact is relatively neutral and will indeed be affected. Software that has introduced APL 2.0 licenses will also be affected, and the LICENSE files and NOTICE of APL 2.0 need to be preserved. The request shows a subjective will, and you may be sued and forced to stop the infringement or compensate for losses. You can choose to release it under the GPLv3 agreement to stop the infringement, or you can choose not to release it and recall it to stop the infringement.
Finally, one more footnote to the release is required. Does GPLv3 require derivative works to be open source to the world? Yes and no. Spiritually so, but in practice the terms govern the act of publishing. That is to say, if you are releasing and using it within the enterprise, and you cannot get the distribution outside the enterprise, then you can open source the "world" within the enterprise. This is also the problem that AGPL tries to solve. When your software is accessed through the network, it is considered to be some form of release, and then the source code needs to be provided to those who can access your software through the network.
The birth of the cloud has led to a new software development landscape, and open source protocols must also evolve with the times. I think that's why Heather Meeker was involved in drafting the SSPL and Elastic protocol to explore. For the time being, we don't know which protocol in the new era can inherit the spirit of Copyleft and deal with the new software development situation, but we will know eventually.
References
[1] Open Source for Business: https://book.douban.com/subject/35309516/