You must log in or # to comment.
I’ve always thought of it this way: permissive licenses ensure freedom for developers. Copyleft licenses ensure freedom for end-users.
Open source licenses are only as useful as your ability to enforce them.
Sorta why the FSF was formed.
Because thses lucences working effects folks other then yourself.
So best for all OS developers and users to work together.
I don’t disagree, but in practice, in my experience they and other groups are not helpful when your license actually needs defending.
I have been with multiple different communities that had GPL and other licensed code stolen for profit in proprietary programs. In all instances, the FSF, SFC and EFF were all contacted and nobody cared.
These offending companies do what they do precisely because they know they can get away with it. And most FOSS developers I would argue do not have the resources or desire to go up against a big company taking or misusing their work.
I have been with multiple different communities that had GPL and other licensed code stolen for profit in proprietary programs. In all instances, the FSF, SFC and EFF were all contacted and nobody cared.
at least the SFC did some enforcing that worked, but i got the feeling these organisations are too “nice” , If the case is a slam dunk maybe it is possible to get a lawyer who will work by getting a large percentage of the earnings.
Is it anymore the case with other licences though.
Obscurity is no security at all. If you have no ability to fight to keep tour copy right or patient. People will copy it open or closed.
Even direct machine code can be copied a reverse engineered fairly simply.
So non of this is purely a open source permissive licence issue. Its a big corperations acting like fudal lords issue.
IMO Obscurity is at least as effective as the attacker’s inability to locate the resource, but I don’t recommend that being your only defense for everything of course.
That being said, you’re absolutely right when you look at it that way. If reverse engineering or copying ASM isn’t out of the question, then IMO all bets are off. Even closed source proprietary programs are not immune from that.
But in the general sense of people casually copy/pasting source code, I think the only defense is not having source available in the first place.
You can copy binary code. Just as easy as source code.
It is only when running on a different architecture it gets a bit more complex.
And give the binary is directly translatable by software. Not hugely more complex for any company of the size you are unwilling to fight in court over open source code.
Sorry but no you are wrong. Hading the source in no way makes code harder to copy. Its how most of us hacked into games in the 1990s.
After all binary code is just simpler instruction set that takes very very minimal effort to convert into assembly language. And can be read by many even without that effort.
Its hardly a secret encrypted format. (Unless you are also designing your own hardware and not letting anyone see that. )
I think we majorly disagree on the definition of “harder” and “just as easy” here. I don’t consider that making me “wrong”, I consider it a difference of opinion. One could argue that it is indeed harder to copy assembly code especially when you do not understand it, or like you contradictingly already stated, when the architecture differs. I was speaking in the context of “the general sense of people casually copy/pasting source code” which I was also implying that meant that those people also did not easily understand assembly already. Sorry for the confusion.
Common DeVault W
DreW
I still kinda like the restrictive takes of copyfarleft that prevents for-profit entities to use anything without contributions. Workers, co-ops, nonprofits do not have any of those restrictions.
My biggest issues with those licenses are specifically that they are & never could be GPL-compatible which would encourage permissive licenses for libraries which is part of what both license types want to avoid.
Has anyone gone so far as to dual-license under copyleft & copyfarleft?
Man, people do love arguing about words without providing (or looking up) their definitions.
Does the GPL being non “restrictive” mean I can use GPL code in my proprietary software? What word that doesn’t offend you should I use to describe this fact?
This is as useless as the git main/master branch debate a while ago.
What word that doesn’t offend you should I use to describe this fact?
Conditional, as in you are free to incorporate the GPL code into your work on the condition that you preserve the freedoms downstream. By default you have no rights to distribute; the GPL does not take anything away, but it grants conditional rights. A restrictive license, like your proprietary EULA, would take away rights that the user would normally have.
Whether or not proprietary EULA’s are enforceable is a matter of dispute; see Software user’s rights (D. J. Bernstein). Regardless, as Bernstein notes, this default set of rights does not include distribution, which is governed by copyright law. Free software licenses such as the GPL are copyright licenses and grant limited rights of distribution. Note that even most “permissive” licenses are by definition conditional as they include at least the requirement for attribution.
Note that even most “permissive” licenses are by definition conditional
You do realize the whole discussion is about what terms to use for differentiating between GPL-like “restrictive” licenses and BSD-like “permissive” ones? Saying that both are “conditional” really doesn’t help anyone.
(also “by definition” the license’s grants may be “conditional”, not the license itself - it’s not as if it looses validity under some condition)
You do realize the whole discussion is about what terms to use for differentiating between GPL-like “restrictive” licenses and BSD-like “permissive” ones? Saying that both are “conditional” really doesn’t help anyone.
That’s fair, I suppose - but going back to the word restrictive, you could also frame the requirement for attribution as a “restriction.”
DeVault suggests the term “reciprocal” at the end of his post. Another term often used is “share-alike.” Both of these terms, I think, more accurately hint at the exact conditions of the grant without negatively framing it as a “restriction.”
The problem with a copyleft license is it’s hard to make a commercial software open source because a competitor can simply copy your work and sell it for cheaper.
That isn’t a problem, but a feature, see: https://opensource.net/why-single-vendor-is-the-new-proprietary/
I know. It’s obviously better for the consumer, but it makes it harder to base your business around it, as noted in that article.
So if I want to build a business, I have to look for libraries that are not copy left, and if I want businesses to use my software, I should not license my software as copy left.
No. You should think in terms of offsetting development cost. When you choose non-copyleft you do it to keep code private, which means you will support all dev costs. It limits how the software can grow because it’s basically vertical scalability — not to mention being culturally limited inside the company.
When you choose copyleft you commit to open source and so does everybody who wants a piece of that software, which makes it much easier for everybody interested in it to offset their development through everybody’s efforts.
With open source there are documented positive feedback effects. Companies who grow to depend on specific software find it cheaper and more efficient for their own interests and benefit to maintain fewer permanent developers as high upstream as possible — as opposed to having many occasional developers downstream, dealing with stuff as it trickles down.
FOSS creates reliable, diverse and ultimately healthy software ecosystems because everybody competes to improve the software first and foremost.
These days selling the software itself is rarely successful nor a particularly good business model. Basically only computer games still work like that, and the commercially really successful ones not any more either.
And this is how we got everything must be online/subscription or everything is a web app. And people complain about that too.
Hard agree on this. Sell software and services to companies, only sell services to end users. I believe both selling your service as a dev and selling a service behind a free app are compatible with copyleft.
deleted by creator
@WolfLink @poVoq open source is a development model, not a business model, if you don’t provide value beyond the bits then you’re going to have a hard time but a copyleft license ensures the playing field is level and no vendor can take your code and extend it with added functionality without releasing that code too
That’s true of any free software license, and as far as I know most formerly-free fauxpen source projects were licensed under a permissive license, because the intent was to be “business friendly” open source projects.
In fact, copyleft actually has an advantage here; other members of the community can sell your work, but they are also required to respect the copyleft. Stronger copyleft licenses like the Affero GPL even protect against proprietarizing free software as a “cloud” service, but “business friendly” projects don’t want it.
Makes sense because if you want to make freely available code but want to allow commercial projects to use it you want to use a liberal license because if your code is copy left licensed businesses won’t want to use it.
I’ve seen this in action: I’ve seen a business reject working with one research group because their code was copyleft licensed, so instead they turned to another group offering a liberally licensed competitor.
