Linux developers threaten to pull “kill switch” over COC

applecorped

Banned
applecorped
Joined
Mar 8, 2008
Messages
20,145
https://lulz.com/linux-devs-threaten-killswitch-coc-controversy-1252/

A controversy over politics is now seeing some of its developers threatening to withdraw the license to all of their code, potentially destroying or making the whole Linux kernel unusable for a very long time.

Most of the internet could be affected as some Linux devs threaten to rescind code in response to CoC controversy.


Activists from the feminist and LGBTQIA+ communities have been trying to force the Linux project to join the Contributor Covenant since at least 2015. The Contributor Covenant is an agreement to implement a special Code of Conduct (frequently CoC from now on) aimed at changing the predominantly white, straight, and male face of programming
 
I heard about the CoC but haven't bothered looking into if it's as insane as I heard it was. (these things are often exaggerated)

Interesting development, though
 
Here is the CoC.

Basically no harassment, no trolling, no doxxing. The kinds of things that would get you kicked out of a conference, in other words.

It doesn't say anything about race or gender-related quotas or any of the kinds of things these people are scaremongering about; it says absolutely nothing against "meritocracy", and you don't have to be a "feminist" or a queer activist to support a code of conduct like this.
 
Last edited:
The code of conduct doesn't look very different from the kinds of things imposed on employees at most modern workplaces. I certainly don't see anything in there attacking the concept of a meritocracy. Also, is there any indication at to how many significant contributors are actually threatening to pull their code? The article doesn't seem to make that clear so I can't tell if this is a serious risk or just a few people making noise.
 
From what I've read of this, code contributed under GPLv2 would seem to be immune from being rescinded, though, until and unless it's tested in court, I don't think we can be 100% sure.
 
There are lots of big companies with big teams of lawyers that would no doubt fight the "code remains under the license it's released under" fight. Companies that come to mind: Apple, IBM, Sun Microsystems, even Microsoft and especially Google.

Of course, copyright holders can change the license on future editions of the code. But my understanding is that the free software community operates under the assumption that when you release code under a certain license, that code retains the license it was released under.
 
applecorped said:
https://lulz.com/linux-devs-threaten-killswitch-coc-controversy-1252/

A controversy over politics is now seeing some of its developers threatening to withdraw the license to all of their code, potentially destroying or making the whole Linux kernel unusable for a very long time.

Most of the internet could be affected as some Linux devs threaten to rescind code in response to CoC controversy.


Activists from the feminist and LGBTQIA+ communities have been trying to force the Linux project to join the Contributor Covenant since at least 2015. The Contributor Covenant is an agreement to implement a special Code of Conduct (frequently CoC from now on) aimed at changing the predominantly white, straight, and male face of programming
Click to expand...
Shouldn't non white, non hetero and non males submit contributions then? Out of most professions programming is not one where you need to make obvious your various identities and preferences, so long as you can write and submit relevant and good code. You dont even have to meet in person to collaborate. Have programers been including anti LGBTQIABCDEFG... in their coding?
 
Last edited:
I don't use Linux and this link might be wrong but it does appear this whole thing is a myth:

ZDNet - Linus Torvalds and Linux Code of Conduct: 7 myths debunked
No, protesting programmers are not removing code from Linux; there are no purges of politically incorrect Linux kernel developers. And Linus Torvalds is coming back.
Click to expand...

​Linus Torvalds answers 5 questions in BBC letter
In an e-mail to the BBC, Torvalds comes over as a voice of reason in the heated Linux Code of Conduct fights.
Click to expand...
 
Hungry81 said:
Shouldn't non white, non hetero and non males submit contributions then? Out of most professions programming is not one where you need to make obvious your various identities and preferences, so long as you can write and submit relevant and good code. You dont even have to meet in person to collaborate. Have programers been including anti LGBTQIABCDEFG... in their coding?
Click to expand...

The nature of the Code of Conduct seems to have been wildly distorted by the author of the article so I wouldn't place too much stock in the author's characterization of the motives of the people promoting its adoption.
 
Hungry81 said:
Shouldn't non white, non hetero and non males submit contributions then? Out of most professions programming is not one where you need to make obvious your various identities and preferences, so long as you can write and submit relevant and good code. You dont even have to meet in person to collaborate. Have programers been including anti LGBTQIABCDEFG... in their coding?
Click to expand...

Right. I doubt the managers of the Linux kernel code have any idea of the race or sexual identity of the majority of people they accept code from, unless those people bring it up.
 
Main controversy is over following section:
or to ban temporarily or permanently any contributor for other behaviors that they deem inappropriate, threatening, offensive, or harmful
Click to expand...

Too open-ended and thus too abusable.
 
Hungry81 said:
Shouldn't non white, non hetero and non males submit contributions then? Out of most professions programming is not one where you need to make obvious your various identities and preferences, so long as you can write and submit relevant and good code. You dont even have to meet in person to collaborate. Have programers been including anti LGBTQIABCDEFG... in their coding?
Click to expand...

The code of conduct does not actually say anything about that. The author of the text linked in the OP is scare mongering.

The CoC was adopted 2 years ago and none of the above concerns have become a reality.
 
Molinaro said:
The code of conduct does not actually say anything about that. The author of the text linked in the OP is scare mongering.

The CoC was adopted 2 years ago and none of the above concerns have become a reality.
Click to expand...

Actually, while the template is old and been in use on many projects for years, it is a new addition to the Linux Kernel.

Linus committed the new code of conduct to the repo, replacing the long standing simple code he wrote long ago, on the 16th of last month.
https://git.kernel.org/pub/scm/linu.../?id=8a104f8b5867c682d994ffa7a74093c54469c11f

For a long explination read I am truly sorry' for my 'unprofessional' rants, I need a break to get help @ https://www.theregister.co.uk/2018/09/17/linus_torvalds_linux_apology_break/
 
Anybody else get a flash back to "The live of Brian", where they got into an equal rights arguement and settled on "males have the right to get pregnant"?

And I do think 'Q' covers it, both definitions of the word.
 
Forget Linux, the entire architecture of traditional computers should be scrapped. It is so transphobic. How offensive must it be to LGBQ+STD!?CIA^7 persons that all computing is predicated on binary theory? Quantum computers can't come soon enough.
 
I don't know ... I'm sitting in my favorite computer chair, touching myself, right?

My Hackintosh is down on the floor doing what it likes to do ...

We catch each others eye and we both start laughing ...*



* with apologies to Dave Attell.
 
Rougarou said:
Of course, copyright holders can change the license on future editions of the code. But my understanding is that the free software community operates under the assumption that when you release code under a certain license, that code retains the license it was released under.
Click to expand...


In my understanding, under the GPL derivative works have to be published under the same license ("copyleft"), hence it's debatable if the authors even have the right to change the terms on future editions of their code.
 
baron said:
Forget Linux, the entire architecture of traditional computers should be scrapped. It is so transphobic. How offensive must it be to LGBQ+STD!?CIA^7 persons that all computing is predicated on binary theory? Quantum computers can't come soon enough.
Click to expand...
'Fuzzy' logic...
 
Childlike Empress said:
In my understanding, under the GPL derivative works have to be published under the same license ("copyleft"), hence it's debatable if the authors even have the right to change the terms on future editions of their code.
Click to expand...

It's my understanding that the copyright holder of a creation holds all rights to do as he or she wishes with their own creation, unless they sell those rights to someone else. If I give you license to use my code in some way or another, I don't think I am subject to the terms of that license myself.
 
Last edited:
Klimax said:
Main controversy is over following section:


Too open-ended and thus too abusable.
Click to expand...

That is the issue. The wording basically says " if we don't like the cut of your jib for any reason we can do whatever we want. " if you read the whole thing the examples at the start are just that, examples, not a list.

There is a difference between accepting people and prostration, this falls firmly in the second catagory.
 
Molinaro said:
The code of conduct does not actually say anything about that. The author of the text linked in the OP is scare mongering.

The CoC was adopted 2 years ago and none of the above concerns have become a reality.
Click to expand...

Sounds suspiciously like the gamergate/comicsgate/"Star Wars is bad now"/BernieBro/etc "white men are under attack!" stuff we keep seeing, then?
 
The problem with claiming a meritocracy is it requires being able to assess merit. Just like memory, it seems the more we learn on the subject the more we learn humans are not great at it.
 
Mumbles said:
Sounds suspiciously like the gamergate/comicsgate/"Star Wars is bad now"/BernieBro/etc "white men are under attack!" stuff we keep seeing, then?
Click to expand...

Poisoning the well fallacy and showing quite an ignorance. Already quoted what is causing such upheaval.
 
Childlike Empress said:
In my understanding, under the GPL derivative works have to be published under the same license ("copyleft"), hence it's debatable if the authors even have the right to change the terms on future editions of their code.
Click to expand...

Linux is licensed under GPL v2, which doesn't have a no-taksie-backsie clause in it. In other words, the license doesn't guarantee that the copyright holder won't unilaterly revoke the license at any time. The GPL v3 does have such a clause. This is the interesting thing brought up in the initial article: the addition of the clause in GPL v3 could be legally argued to mean that GPL v2 doesn't imply non-revocability. After all, why else would they have added the clause? Without the GPL v3, an easier argument might be made in court that the GPL v2 would be non-functional without implied non-revocability.

More importantly, Linux isn't just licensed under GPL v2, rather all contributions to Linux are licensed under GPL v2. Contributors to Linux do not hand over their copyright. As mentioned in the article, some OSS projects do demand that of contributors, such as GNU Emacs. If you want to contribute code to Emacs, you have to hand over your copyright on that code. Thus Emacs could not be threatened in this way, since the maintainer of Emacs owns the copyright for Emacs in it's entirety. The copyright for the Linux kernel is a mishmash of ownership from a bunch of individuals.

I really hope this is a tempest in a teapot, and people don't pull the trigger on this. It would be disastrous not just for the Linux kernel itself, but OSS as a whole. There is an entire chain of liability that would exist that would make the use of OSS commercially completely untenable.

edit:
Another article from ZDnet: https://www.zdnet.com/article/what-happens-if-you-try-to-take-your-code-out-of-linux/

It makes me feel a little better, but not entirely. The fact of the matter is that this sort of thing, afaik, has never been settled in a court of law. I worry people deciding to attempt revocation will bring up all the FUD of the 90s, even if a court ultimately agrees with Heather Meekers take on this from the article. If there were a mass revocation by kernel developers, at best we are looking at a repeat of the loooong drawn out nightmare of the early 2000's SCO fiasco all over again.
 
Last edited:
I'm in two minds whether to post again on this thread. I'd like to, but I heard that one of the people who coded Chrome has different views to my own and might have said something I find offensive. With that in mind I should really make a stand and boycott his work. I mean, it's not that I don't give a fat monkey **** who codes my product as long as it ******* works.
 
baron said:
Forget Linux, the entire architecture of traditional computers should be scrapped. It is so transphobic. How offensive must it be to LGBQ+STD!?CIA^7 persons that all computing is predicated on binary theory? Quantum computers can't come soon enough.
Click to expand...

Oh, but we already have the transputer for those who want to do things a different way.
https://en.m.wikipedia.org/wiki/Transputer
 
AIUI, you can not revoke a granted license as long as the original license doesn't explicitly say this is a possibility (GPL2 does not allow revocation of license). What can happen is that a contributor can withdraw their code for future releases so no new releases could be made with that copyright holders' code in it.

If it happens the kernel team will just have to re-write the missing code portion from scratch. So far nobody has withdrawn from the project and a request for volunteers by one of the team (IIRC, a Google employee in the Android division) has yielded over 50 developers willing to step up if needed.

Then there is the commercial side, hundreds, possibly thousands, of large corporations want the kernel to stay alive for use in their products. It seems a virtual certainty that if needed they will pay for developers to contribute, especially now that the CoC more closely conforms to standard corporate requirements.

If anyone does withdraw causing a problem for Linus and the team it may convince Linus to move to GPL3.
 
paulhutch said:
AIUI, you can not revoke a granted license as long as the original license doesn't explicitly say this is a possibility (GPL2 does not allow revocation of license). What can happen is that a contributor can withdraw their code for future releases so no new releases could be made with that copyright holders' code in it.
Click to expand...

Actually I have looked up more on the relevant law, and it turns out you are partly right/partly wrong. You can revoke a license, however U.S. Copyright law puts pretty strict limits on this (specifically 17 USC 203 "Termination of transfers and licenses granted by the author"). This puts my mind more at ease.

Essentially, you can only revoke a license 35 years after granting it, and then you only have a 5 year window to do so. Furthermore you must give 2-10 years notice of the revocation (you can give notice before the 35 years is up).

So bullet dodged, the threats of revocation are baseless. Nothing in the Linux kernel has been licensed between 35-40 years ago, so nothing can be revoked. Even once the point in time is reached where someone could revoke their license, there would be plenty of time to replace the offending code.

edit: Furthermore, the nature of OSS dissemination and use makes the entire revocation process untenable to use anyways. When you redistribute GPL software, you aren't sub-licensing, the license specifically specifies that the licensor is granting the license to the next party directly. So every person using a piece of GPL'd software is a direct licensee of the copyright holder. Here is the kicker: in order to use this revocation process, you must provide in writing, a copy of the revocation notice to both the licensee, and the Copyright Office.

edit 2: Obviously OSS is used worldwide, and the above doesn't apply everywhere. In fact the U.S. appears to be something of an outlier here. Even if the GPL is a bare license rather than a contractual one, I don't see any language indicating that 17 USC 203 shouldn't apply. However elsewhere bare licenses could be revocable at any time, this isn't part of the Berne Convention. So havoc could still be wrought internationally if the courts there treat the GPL as a bare (i.e. gratuitous) license. I think this is where a lot of the claims that it is revocable at any time are coming from. In the U.S. it probably isn't true (though probably still needs to be tested in court), but internationally it could be true.
 
Last edited:
You must log in or register to reply here.

Back
Top Bottom