Rust's Redox OS could show Linux a few new tricks (InfoWorld)

GPL _is_ viral. You might not like the connotations of it, but it IS viral.

Real-life viruses don't require premeditated positive action on the part of their hosts in order to become active. While it is perfectly possible for you to catch a cold simply by being in the wrong place at the wrong time, your code can't be “infected” with the GPL unless you deliberately add GPL'ed stuff to it and then deliberately decide to distribute the result, which is a huge difference. Even then the GPL only applies to the combination of your work with the GPL'ed stuff – your stuff on its own can be licensed differently as long as the terms of its license don't restrict the rights that recipients of the combined result enjoy through the GPL. That is not how actual viruses work.

Calling the GPL “viral” is essentially propaganda.

AFAICT, there's two arguments being made here, one implicit and one explicit. The implicit one is that libreadline is trivial and that it would be unfair for the authors of the proprietary code, which is decidedly nontrivial (a fact which I guess is intended to be reinforced by the presence of patents?), to "lose" "everything".

Libreadline is not exactly trivial but neither is it very complicated. It's not _fair_ that it should infect the whole code base.

(Your comments persist on simply repeating terms that have been pointed out as knowingly (and grossly) misrepresentative ('infect'). To me, this seems like a propaganda campaign and definitely not an attempt at having an intellectually honest discussion.)

AIUI, the 'not fair' comment quoted above concerns the concept of 'derivative' in copyright law. IIUC, you think that it should be 'fair use' to call out to or embed a library and should not be subject to copyright limitations.

(meta: If you think the above is a misrepresentation of your thesis, please consider that it's hard to make out an argument among slogans that are clearly /not/ concerned with making a logical argument, rather with repeating a narrative. This is tiring work. I suspect that's not lost on you, but pointing it out all the same. Personally, I'm thinking that if LWN had some cap on the number or rate of comments per article by any single commenter, that might help with having a more concise discussion and make it easier for more people to be heard. Well, for well-intentioned people at least.)

If this is your point, then is not your frustration misdirected? Of course people who (a) do not want to help companies/individuals produce proprietary software and (b) want to incentivize people to knowingly participate in the production of free software will prevent appropriation of their code to the extent allowed by law.

Not that one can easily sympathize with your PoV -- AFAIUI, your argument is mostly concerned with enabling people to produce proprietary software. Which, it seems to me, is easy enough already (not to go into the larger discussion on whether free work for vendors of proprietary software is 'fair' or indeed whether further skewing the legal framework in favor of the production of proprietary software is socially useful). I mean, it's not like the set of laws concerning copyright and, especially, patents WRT software has been lobbied into existence by software freedom activists. Some might even say the current system started out and has largely evolved to cater to the needs and wants of proprietary software production (though, of course, different classes of proprietary software vendors have different requirements).

Note, that it's entirely within the libreadline's authors right to impose such restrictions. I'm just saying that they are not even close to fair and the rational choice for libreadline users is to walk away.

If you add GPLv3 to the mix then it becomes not only unfair, but also political and deceptive.

'political' isn't a slur. Everything we do has political connotations. When 'political' is used pejoratively, that seems like a way of telling people that they should not be thinking about politics, which seems hypocritical given the, I'm sure you'll agree, very much political narrative your comments have been reproducing. Also, given what we've read so far ('cancer', 'viral', 'infect', 'propagates'), I take issue with your calling anything deceptive. However, if you have an argument to that effect re: the GPLv3, please share it in a clear manner so that we can all consider its merits.

Your wording "libreadline users" seems ambiguous in this context

/Users/ of software using libreadline are probably happily reaping the benefits of using software that they can read the source of and, importantly, collaborate on, receive and distribute modified versions of.

/Developers/ who want to integrate readline in their product seem to be well aware what the licensing conditions are. People who want to develop proprietary software know that this code is not for them. People who want to develop free software may use this code, but they cannot permissively relicense or they would enable others to create proprietary derivatives. If they have reasons to require that their derivative be permissively licensed, they can consider their options and decide appropriately. I honestly fail to see where the (I hope you agree that's accurate) intense hatred of the GPL in your posts here stems from, other than the fact that it may disqualify some code from inclusion in proprietary programs.

Oh, and if you're going to talk about 'community splitting' again, we've already had that discussion here.

Would one expect sympathy if they somehow incorporated a competitor's proprietary code in their own proprietary project?

This is again a wrong analogy. The correct analogy would be a commercial library that prohibits its users from competing with library author's products.

That took quite a few re-reads to (maybe) figure out what you're talking about. IIUC, you're conflating code and people and making the argument that _code_ is a user of that library and hence the analogy should be with _people_ who are bound by an overreaching EULA? Again, IIUC (and your phrasing is really not helping), this seems quite confused. (a) As has been pointed out repeatedly, code does not have agency, its authors do (b) AFAIK, an EULA is /not/ a copyright license, but a contract. If there's an argument to be made here, kindly unpack it cause I, at least, cannot make sense of the above, and I tried to.

Having cleared the copyright considerations, we're left with nomenclature. IIUC, your argument, phrased in a non-propagandory fashion, is that 'hereditary' should apply to the weakest form of copyleft only (i.e. MPL or LGPLv2 with a static linking exception)

Correct. These licenses do not propagate outside of their "genetic line".

whereas strong copyleft should be deemed 'viral'. IMHO, the 'viral' part has been thoroughly taken down

Incorrect. GPL _is_ viral. You might not like the connotations of it, but it IS viral.

Umm, no-one can argue with 'incorrect' :-) Whatever arguments you've presented though...

And OF COURSE it's political. Everything is, even including the GPL itself. Should I quote the section 0 of it?

I think you misunderstood there. As explained above, 'political' is not a slur. However:

The word 'viral', when it comes to agency is so wrong that it seems clear to me that it's use is motivated by political views rather than any desire for accuracy.

i.e. your comments lead me to question your willingness to conduct a discussion (which might result in disagreement, sure) instead of engaging in posturing and staying on message re: cancer, viral, infectiousness and so on.