@t0k@mangeurdenuage@rysiek@jwildeboer free/libre/open do not have technical differences in this context, the definitions for a free or open license are effectively the same.The difference is to which greater context you are signaling belonging, the corporate-oriented (open) or the user-oriented (free/libre).
@t0k @Steinar @rysiek @jwildeboer
>I think LGPL qualifies as forever-open too
>as forever-open too
>open
I have to interject on this, what you are calling open may refer to open-source, which is completely different from what rms/fsf transmits with software freedom, or alternatively named free/libre software.https://www.gnu.org/philosophy/open-source-misses-the-point.en.html
>It just allows that a closed-source ecosystem grows around.
LGPLv* licensed code permits the usage of it in proprietary code BUT the company/developers cannot restrain people from asking/getting the source code/modifications made in the said LGPLv* licensed code which is distributed with the proprietary code.
@t0k@rysiek@jwildeboer We need a word for the LGPL category too. LGPL can't be "swallowed" like BSD, but can be used as an integral part of a closed system.
@rysiek@Steinar When someone says GPL and viral together, I immediately suspect that person opposes Software Freedom. And I tell them that. Has worked quite well ;)
@benjaminpaikjones@rysiek@Steinar@jwildeboer That's also the terms that the CERN OHL (Open Hardware License) is using for its variants OHL-S (strongly-reciprocal) and OHL-W (weakly-reciprocal).
Even though as I understand OHL is not based on copyright law in the same way as GPL/LGPL is.
@t0k@benjaminpaikjones@rysiek@Steinar yeah. Every few years some folks come along saying we need new words. But I never really understand why. We have all the terms we need (permissive, non-permissive, weak/strong copyleft) since years. They are well defined and mostly understood in good ways. But I guess some people just WANT new terms to create confusion. Divide et impera ;)
@rysiek@Steinar@jwildeboer is viral a drop-in synonym for copyleft? I thought it was used to differentiate licenses like GPL from LGPL, which are both copyleft (but I wouldn't consider LGPL "viral")
@whvholst@rysiek@jwildeboer@t0k@Steinar@benjaminpaikjones Still, in the end you're either allowed to create derivatives or you're not, and it's conditional on you paying that privilege forward, or it's not.It's possible that "reciprocal" is more immediately-obvious to newcomers than "copyleft", and it's a term used by some other licenses and entities too. But that's not really related to how copyrightable, pattern-protectable or other neighboring rights-able hardware is.The surface terminology still applies even though the licenses may need to be implemented using different legal language.
@jwildeboer I don't recommend licensing things as AGPL willy nilly, since it can restrict its use even in other Free Software. The AGPL should be used for code expected to be used as a service, otherwise it doesn't really make sense.
@cstanhope@robby Yeah, it makes sense to license software as AGPL if it might at some point end up in a service.And generally, it very well might.The argument when choosing between AGPL and GPL goes something like the argument of GPL vs LGPL. The GPL is analogue to the LGPL on a network service level.www.gnu.org/licenses/why-not-lβ¦
@mangeurdenuage I was replying to a comment that *specifically* mentioned Open Source: "... is anathema to open source ...". I am well aware of the difference between Software Freedom as defined by fsf and open source as defined by the Open Source Initiative.
The intention behind the #AGPL, or any other #FOSS licence for that matter, is to set conditions as to *how* the #IP may be used, not by *whom*, which is what you appear to very disingenuously advocate and what my earlier comment addresses.
@0@mangeurdenuage You are trying to interpret too much. It is as I said. Google will NOT tolerate AGPL licensed code on their systems. That is their decision, which they publicly state. By me merely linking to their statement and a real-world discussion caused by that decision, I am not advocating anything beyond "such a decision to ban the AGPL completely is weird".
@0@mangeurdenuage It is not disingenuous to logically conclude that licensing your code under AGPL, with this public statement in mind, means that Google will not be able to use your code. The only thing it shows is that blanket license bans like this are limiting the freedom of Googlers. By their own decision.
Placing restrictions on, or releasing code with the intention to restrict, who may use it is anathema to open source and a remarkably daft idea for reasons that have been extensively discussed.
A different matter altogether is *under what circumstances* #FOSS code may be reused.
Generally speaking, when I expect that work of mine may end up being used by say #Google, or #RedHat, I release under #GPL to give them an opportunity to contribute back their improvements.
>s listed on the "approved" list of the Open Source Initiative.
OSI and FSF have different definitions of what is free/libre and/or Open Source.
I stick with the FSF's definition. Some can stick with the OSI.
@jwildeboer yes, it is very important to not look at this backwards. Nothing in AGPL prohibits Google from using software licensed under it. Google is preventing Google from using software licensed under AGPL.If Google decides to change their policy they are free to use AGPL software.@0@mangeurdenuage
@astraluma Yep. That happens frequently. Often means they love to take the work from others for free but donβt think they need to give back. π€·βοΈ
@jwildeboer nice trick! :) A similar thing is true for most of the automobile industry, who wants to make sure their devices are locked down so users cannot flash their own software. This is in contrast to GPLv3, which demands that users of GPLv3 code must be able to change the code running on the device at will.
@feld Yes, I think so, too. If somebody uses a GPL-licenced source file in a released program then I understand that the whole of that program has to be GPL. If that's right then GPL can reasonably be described as viral in a way that BSD, MIT or even LGPL wouldn't.
Personally, I'd never use GPL. The odds and ends I have released have generally been MIT/public domain. I could be persuaded to use LGPL, though.
@edavies@feld@rysiek@Steinar No. Because when you *know* how the GPL works, it is not a surprise or unavoidable thing. The term "viral" was specifically coined to insinuate it's accidental. It isn't.
@feld@rysiek@edavies@Steinar It's a license. Nothing more, nothing less. Some hate it, some like it. These over-the-top rhetorics are plain stupid IMHO.
@jwildeboer@rysiek@edavies@Steinar it restricts freedom. The lies and twisting of language is straight out of politics.Freedom means freedom. If you have any strings attached, it's not free.
@feld@wizzwizz4 The four freedoms are privileges in the sense that it's awesome to be able to do these things, and they are privileges relative to all-rights-reserved, as the latter is legally considered the default and software freedom a beneficial exception, but they are not privileges in the political theory sense.The four freedoms do not entitle you to someone else's effort to provide anything to you, they are freedom from copyright enforcement, freedom from someone preventing you from doing these things. Absent copyright, they require no effort from anyone.Free software licensing is an effort today, whether copyleft or non-reciprocal, because we need to document clearly that anyone in the position to remove our freedoms has given up the privilege (the entitlement to efforts from the state and its legal system) to do so.
@feld@rysiek@edavies@Steinar@jwildeboer No to the first paragraph, no to the second paragraph, no to the first part of the third paragraph, yes to the final six words.
There are free software licenses that let you turn that software into non-free software. However, many copyleft licenses are still free software licenses.
@wizzwizz4@rysiek@edavies@Steinar@jwildeboer there is no debate over the definition of "freedom". it's set in stone.GPL restricts freedom, as do many other licenses.GPL *preserves* access to the code. It does not provide freedom for the person who consumes the code.
0. The freedom to run the program for any purpose.1. The freedom to study and change the program.2. The freedom to share the program with others, freely.3. The freedom to share your modified versions with others, freely.
@wizzwizz4 0. the PRIVILEGE to run the program for any purpose1. the PRIVILEGE to study and change the program2. the PRIVILEGE to share the program with others, freely3. the PRIVILEGE to share your modified version with othersprivileges, not freedoms.The ability for me to do anything I want with the code without anyone being able to tell me what I can and cannot do is the only true freedom.You're just listing privileges masquerading as freedoms because they're Stallman's propaganda
@wizzwizz4@rysiek@edavies@Steinar@jwildeboer if it's free I should be able to use it in a closed source context. no strings attached. free.Free is not the opposite of closed/proprietary. Things can be closed/proprietary and still be free.So many in the open source community are brainwashed into thinking "free" means "perpetually open source" which is NOT the same thing.
@edavies Let me just put it this way. Working at Red Hat since 15 years and being in political lobbying for Software Freedom has made me very aware of the power of words. The endless discussions with legal departments that insist on "viral" and how GPL is somehow evil has hardened me. Sorry for that :) @wizzwizz4@feld@rysiek@Steinar
@wizzwizz4@feld@rysiek@edavies@Steinar It's the age old discussion that will never end. Is freedom baseed on reciprocity (GPL style) or does freedom mean no restrictions (BSD style). It will never be solved. Pick your side.
@jwildeboer Agree, there are good reasons for both GPL and BSD/MIT. I wish I hadn't mentioned my preferences on that as my main point is that I don't think the use of the word βviralβ here is anything like as pejorative as has been made out.
@clacke@rysiek@jwildeboer@t0k@Steinar@benjaminpaikjones The notion of derivation is very much a copyright thing (and very fuzzy in copyright law). Like I said, the relationship between hardware and copyright is complicated and putting requirements on derivation may be moot if copyright is not in play to begin with. So it's nowhere near as binary as you present it. In a lot of cases the honest answer is "I don't know".
@juliank@robby@jwildeboer If you use the GPL, make sure to also dual-license under AGPL. That makes it easier for people to make AGPL derived works,without having to have a confusing aggregate license.
@wizzwizz4@juliank@jwildeboer This is not necessary. Section 13 of the GPLv3 allows combined works to be licensed under the AGPLv3. When combined the AGPLv3 applies to the whole combined work.
@emacsomancer For the intellectual discussions, sure. For day to day use I describe GPL style as Share-Alike, because that is what it means and that term is also used in the CC world (Creative Commons).
You also can't create a GPL derivative work licensed under AGPL without messing around with what's the original code and what's the AGPL'd code β you can't say βthe whole code is licensed under AGPLβ, but only that βit is combined GPL / AGPL, and can be used under AGPLβ.
@wizzwizz4@jwildeboer@juliank@robby Some parts being offered as this or that license means that code is being offered separately from the combined work, and that's fine, it can continue to do that, but the author of the combined work will "license the entire work, as a whole, under this License to anyone who comes into possession of a copy" to quote the GPLv3.That means when you are running a GPLv3 or AGPLv3 program, in that moment the whole work is licensed to you under that license, and all the terms apply to it.
The terms of this License will continue to apply to the part which is the covered work, but the special requirements of the GNU Affero General Public License, section 13, concerning interaction through a network will apply to the combination as such.
@wizzwizz4@jwildeboer@juliank I see what you mean. It seems like more of a convenience to distribute your code under both AGPLv3 and GPLv3 when the GPL is desired, to allow the whole work to be licensed as cleanly AGPL when combined. It legally doesnβt change usage / compatibility, right?
@robby It has notable impact in that you can then simply take the source code and make all your changes AGPL-only, which is something the language in the GPL is supposed to limit.
@juliank@jwildeboer@wizzwizz4@robby The GPLv3 explicitly allows adding the requirements added by the AGPLv3, that's how you can legally redistribute AGPLv3+GPLv3 software at all.
@jwildeboer@juliank@wizzwizz4@robby I don't know if people do this, but the way I read it, an AGPLv3 program that links with e.g. zlib should offer "an opportunity to receive the Corresponding Source of" the exact version of zlib used to build the program.I'm not 100% clear on this, because the definitions are circular: The covered work is the Program and the Program is whatever is covered by the license. π
It does mention the GPLv3 specifically and specifically says that GPLv3 parts also need to be offered, but I don't think that implies non-GPLv3 parts are exempt, it's just that two copyleft licenses are necessarily incompatible unless made explicitly compatible. For the parts with licenses that allow it, they're simply part of the covered work, no explanations necessary.
@juliank@jwildeboer@wizzwizz4@robby I think it's the explicit text in the AGPLv3 that "the work with which it is combined will remain governed by version 3 of the GNU General Public License" that allows you to do that. That text wouldn't be required otherwise.
@jwildeboer@rysiek@webmink@pettter@Steinar@paoloredaelli The term "viral" doesn't have to be useful in a positive way, and in this space it wasn't originally intended to. For constructive use, the other terms suggested here are better.
@jwildeboer@wizzwizz4@rysiek@edavies@Steinar there is no discussion to be had. Only one side is trying to change the definition of words to fit their agenda.Is it really that hard to say "The GPL preserves openness of the code for everyone" instead of lying and saying "it gives the consumers freedom, for our very niche definition of freedom"
@clacke@rysiek@jwildeboer@t0k@Steinar@benjaminpaikjones All of this is perfectly navigable for players like ARM, much less so for open hardware projects. I do recommend the stuff Andrew Katz has written about this, including talks at FOSDEM.
@whvholst@rysiek@jwildeboer@t0k@Steinar@benjaminpaikjones I don't see why ASICs would be very different from a fixed medium like a CD-ROM, but law has done weirder things.I see why they might be treated different from FPGAs, which I suppose might be treated more like executing software or somewhere in between.Thanks for the suggestion on material, I'll check that out.
@clacke@rysiek@jwildeboer@t0k@Steinar@benjaminpaikjones Yes, unless those designs would involve ASICs. Come to think of it, FPGAs are an interesting edge case here, again. Basically, in hardware you cannot solely rely on copyright because the "maker's mark" is not always there. So you end up with more exotic IPR that typically requires (some form of) registration (patents, semiconductor topologies). Which doesn't work too well with open hardware either. So, it's very, very messy.