@marinade awww they want to redirect the volunteer mapping efforts towards "open" maps that do not have a share-alike clause so that they can use the data however they want without having to negotiate with the volunteers!
I understand many wants location and other tech details removed for #privacy reasons. But removing credit, copyright and #license info is for me like stealing. And makes me hesitant to post any of my images on Mastodon.
Even Meta/Facebook now preserves this info in images posted to their platform(s) - as shown...
Let's pretend we're proponents of free and open source software, enlist an army of week intentioned FOSS developers to contribute to our project, and once successfully deployed in many enterprises across the industry...
Pull the rug out and convert it into a proprietary product with a bunch of undisclosed, hidden code that we won't ever show you - Muahahaha...
Yeah. I see this happening right now in several prominent and celebrated open source projects that you're probably completely oblivious to those sinister objectives.
This is why the most ubiquitous desktop operating system in the world is Minix.
What's that you say?
Yup, Minix. But that's no secret, the cat was out of the bag on that one a few years back (after being secretly so for many years).
Before you contribute any more code, translations, or documentation to a software project, consider this:
Now, there's another point to be made here, without specifically naming any projects currently abusing user contributions. Let's call this hypothetical project "hammer&anvil", itself a fork of a popular software project - but claims it's all about being free and transparent, wanting to distinguish itself from the project it's forked from by adopting GPL3 instead of a permissive license.
Sure, the project's BDFL (let's call her "Strawberry Daiquiri"), says one day, "were forming a fork of project X because they've formed a company and I'm afraid what they are going to do with X because it's under a permissive license. This girl will be brutally transparent and completely run by the community under the philosophy of anarchy, but we're going to call it a sociocracy so you don't know that it's really just me making a proprietary product for my own ambitions".
Well, Miss Daquiri decides to capture by capitalizing upon the sentiment that folks have for Copyleft - it's supposed to protect free software, right?
Well, this fork (hammer& anvil) is a hosted solution - meaning SaaS, meaning, it runs elsewhere (other than in your computer) in the cloud as a publicly accessible service. Hmmmm.
That means that the most appropriate Copyleft license is likely the AGPL, and not the GPL as one would expect fur a desktop or other local program that you actually download and install in your laptop or server.
The GPL requires that when you distribute (give away or sell) your program, either by letting someone download or handing it to them on a USB stick, Etc., You must also make available ALL of the source code, including any changes you've made to the program.
But if you run a modified GPL program as a service in the cloud you don't have to provide ANY off the changes you've made to the code.
Hmmm.
With AGPL you do have to supply your users with ANY code modifications you've made to the running service to which they have accounts...
So let's just say that you fork Mastodon, and call it Glitch-Soc, modify it, and run it in the cloud for people to create accounts on and use (for free or for monthly subscription fees - it doesn't matter). ANY and ALL changes to the code base that you make MUST be made available anytime a user asks for the source code, because it's an #AGPL licensed product.
And in reality, such is actually the case with this exceedingly popular and capable #fork. It's a fine product in it's own right.
But had you changed the license to all contributions moving forward to #GPL, you wouldn't have to provide any modifications you made (unless you give or sell the software product itself on say, a USB stick or via download).
Why? Because you're just allowing them to access and use your service, your not actually giving them the program to use for themselves elsewhere - so any modifications you made since forking under a different license (GPL instead of AGPL) isn't something you have to show them.
You've essentially created a #proprietary product (if you're so nefarious as to hide your code changes by butt disclosing them), the only code of which you must supply being that which existed under the AGPL before you forked it.
Both #Copyleft and permissive open source #licenses like #BSD and #MIT can be a good thing, or they can be abused beyond the intentions of the #FOSS inclined project contributors. Just make sure that you understand what can and cannot be changed where your intended purpose for the #distribution and #availability of source code is concerned....
There are BIG differences between the ramifications of each #license and how they can affect transparency and distribution of your free gifts to the world.
In our hypothetical scenario with hammer&anvil, the #BDFL, #Strawberry Daiquiri, has decided that she's going to launch a hosted service, and she's going to include things that you don't see and can't be aware of behind the scenes which, if disclosed, you would have nothing to do with - but you'll never know what kinds of scary things she's done with the product that only resembles the original on the surface, because Miss Daquiri will never have to show you the code she has added behind the scenes.
"Beautiful Victor, Beautiful."
-The Monster, speaking to his creator in the film, 'Frankenstein, The True Story'.
@fsf Where can I read about the legal licensing and copyleft issues surrounding generative AI algorithms like LLMs (Large Language Models) like Chat-GPT or Copilot, trained on GPL'd source code?
I wonder if there is a need for a new license that explicitly makes training generative AI on open source code requires the AI model to be open sourced?
Does the FSF have any written opinions or educational materials related to this topic of the relationship between copyleft and generative AI trained on copyleft source code?
I notice that copyright holders, like Sarah Silverman, Christopher Golden, and Richard Kadrey are suing Meta and OpenAI for copyright infringement, so I hope that copyleft will have its legal battles in a similar vein.
Also, I'm betting that existing GPLv3 licenses don't mention being used as training data in other algorithms. I imagine there could be stronger legal wording in future licenses that specifies consequences for using copyleft software as training data for closed source algorithms.
Defending the copyright system seems at the heart and core of the copyleft free software agenda, and currently Sarah Silverman et al seems to be leading the charge.
Given that the latest #twittermigration to the #fediverse is apparently caused by Elmo wanting to be paid for the content being used for training of #LLMs (as was the case for the #redditdark selfown) - I wonder if we need a #creativecommons#license that explicitly denies such use.
If that is wanted by the author, one can still license in such ways in addition, I guess?
Are there software licenses out there that guarantee to the user the ability to receive updates and improvements to a piece of open source software?
I don't know of any license other than the GPL that provides this guarantee to the user.
I know that I'm more motivated to learn a new tool when it is protected by the GPL because of this guarantee that future versions will also be open source.
When I license my own software, I choose the Affero GPLv3 because it provides this extra guarantee to the users, even if the software serving the user is on a webserver.
Your link is an answer to my question. The only other open source license that requires disclosure of modified source code is the Mozilla License. It's still not as strong as the AGPL, which defines distribution as network use.
We're not going to bother getting our #submersible#safety-rated because “[t]he vast majority of marine (and aviation) accidents are a result of operator error, not mechanical failure,” is certainly a #novel take on things.
Not perhaps one I would have run with, but they didn't ask me.
For anyone wondering why safety and design is critical for submersibles, at 3800m depth the pressure is around 5400 psi. A #crack or #buckle converts you into a #pink#mist#instantly.
You go to various remote or isolated or primitive parts of the world, and you can definitely find the local equivalent of Billy-Jo Bob's Real Flying Airline, with a #plane that is held together with hopes & dreams, a #pilot who hasn't seen a #license form since 1978 and smells of #booze, and desperate passengers...
Well, it is probably going make a comeback once #Meta (like #Facebook and #Instagram), and other services with ToS/ToA that many question as far as content licensing… connects to #ActivityPub.
Or, maybe not.
Regardless, All content that I own in the Fediverse, unless otherwise specified, are under a Creative Commons Attribution-ShareAlike 4.0 International License. You should see this note on my respective account bio/description.
If I am using someone else's work, I always opt for #CC0 or #PublicDomain works (which doesn't need attribution, but I try to still provide it); #CCBy; or #CCBySA, as much as possible, with appropriate attribution in Creative Commons licensed works.
No, the re-share and quote features do not violate a CC License (not even the “All Rights Reserved” ‘license’).
It is re-uploading, making derivatives of, re-distributing, that may or may not violate a License. In CC works under a CC By and CC By-SA licenses (to mention two), as long as you fulfill the requirements, it is safe. However, in “All Rights Reserved” content it is a violation. Including translations of an “All Rights Reserved” work you have no permission to translate.
When you install flatpaks from the command line, there is no warning about what permissions will be used by that flatpak, and whether or not this is proprietary software. Again, much worse than what we are used to with regular packages.
The #Apple Wallet app in #IOS 15.4 or later allows you to store a digital copy of your #US#DriversLicense. Supposedly this is to make #TSA checkpoints less cumbersome to get through. Part of the process of enabling this is taking a #selfie to prove the owner of the #license is present during the process. The State of #Georgia just rolled out this capability on May 18. Less than a week later the Georgia Department of Driver Services felt the need to send out this notification on their social media: "PLEASE WEAR CLOTHES when taking your selfies for your Digital Driver's License and IDs" (emphasis added) #UnintendedConsequences#IPityTheReviewers#IOS15#AppleWallet
The paper with that figure got accepted by a journal requiring a CCBy4 license.
What's the license of figures that we produce by using that R package? The code of the package is released under GPL3, but not clear to me if that applies also to the figures produced by me with that package.
No, you can't license my cat picture to Elon, Jack and Mark.
When you post on #Bluesky, #Twitter or #Meta you agree to grant them a very broad perpetual license to the content, including the right to sublicense. On Mastodon, most instances do not take a license. Any bridge that takes content from Mastodon and, without permission, puts that content on one of these platforms is violating that user's #copyright to the content. You can not #license content which you do not own.
@mastodonmigration that last claim is not quite right, is it? You can #license content which you do not own if you have been granted the right to sublicense it (which those sites do, as you note), but I think most mastodon instances don't demand that, or not in a way OK for those services, so most of your post is correct. Bridges must check what they're transporting.