Earlier quoted context omitted.
>I want to know what stuff you guys are putting in public GitHub FOSS repos that you don't want replicated in any way... Nobody has claimed that they want this. People just want derived work to adhere to the license they chose for their project. >I also want to know why people think their code is so special that no one else could have ever come up with it independently. Each and every opponent of Copilot is the best…
At one point does AI recreating patterns it has seen from reading source code count as a derived work? What if a human learns to code by reading only GPLed code, does all the code they write fall under GPL as a derived work now?
Give up GitHub: The time has come
441–450 of 585 posts
Re: Give up GitHub: The time has come
#442As a leader of a FOSS project that is on github, and migrated off of sourceforge because SVN and email patches were not scaling - I'm a bit confused by this article. Co-pilot has issues, ergo github is going the way of sourceforge, and so now we must abandon github? Do I have that reasoning correct? We need to: - migrate the bug queue - have all links in commit history break - application integration with githib for…
> As a leader of a FOSS project that is on github...Do I have that reasoning correct? The way I read the article was that by being on GitHub, you are implicitly agreeing to no longer be a FOSS project as regards licensing. GitHub customers can use Copilot to generate proprietary code that's identical to your project's code (several articles I have read call this overall idea "laundering through Copilot", which sounds…
Anyone can already generate proprietary code that is identical to your GitHub project's code just using copy and paste.
For people that care to respect copyright, there's a copilot setting to block exact copies of code in the training set (which only happens a tiny percentage of the time, unless you're actually trying to make it happen).
For people that don't care to respect copyright, git clone is a way more efficient way to violate your license.
Re: Give up GitHub: The time has come
#443Earlier quoted context omitted.
> As a leader of a FOSS project that is on github...Do I have that reasoning correct? The way I read the article was that by being on GitHub, you are implicitly agreeing to no longer be a FOSS project as regards licensing. GitHub customers can use Copilot to generate proprietary code that's identical to your project's code (several articles I have read call this overall idea "laundering through Copilot", which sounds…
> The way I read the article was that by being on GitHub, you are implicitly agreeing to no longer be a FOSS project as regards licensing. In this case, they have violated our license. The license states how it cannot be changed, and implicit changes by the hosting company is not one of them. Isn't there a legal case here? If no, then the license is not violated. This makes me wonder why not take Github to court vs a…
> This makes me wonder why not take Github to court vs a tamper tantrum (we're taking our marbles and going home!)
Mainly because it might be more in line with traditional FOSS ideals to take our marbles and go home? It's hardly a temper tantrum. After the Linux kernel got into one too many arguments with Bitkeeper, there was an inflection point where Torvalds got fed up and just wrote git instead. I think this is an inflection point like that. And "taking Github to court" is much easier said than done for a FOSS project, although I assume that someone will be doing that at some point.
> So, we have to move all that, to another host: who is to say that other host will always be a better FOSS steward? who is to say that other host has anyone near the level of features? An integrated CI/CD and hosting of release artifacts is huge.
In addition to GitLab, I think Sourcehut offers similar features and is also FOSS (AGPLv3).
> So, on the premise that our license has been violated, instead of sue'ing, we should take our FOSS somewhere else? Again, the list is large, we need to migrate all of that.
I don't think anyone is telling you what to do, just offering suggestions. My suggestion is that switching to a forge that respects FOSS is a reasonable alternative to legal action against GitHub or just ignoring the potential license violations via Copilot. You can of course choose any of the alternatives, or the status quo; which is what you seem to have largely convinced yourself is fine. Good luck, and thanks for working on FOSS anyway!
Re: Give up GitHub: The time has come
#444> If it is, as you claim, permissible to train the model (and allow users to generate code based on that model) on any code whatsoever and not be bound by any licensing terms, why did you choose to only train Copilot's model on FOSS? For example, why are your Microsoft Windows and Office codebases not in your training set? This is my favorite question about Copilot ever.
I want to know what stuff you guys are putting in public GitHub FOSS repos that you don't want replicated in any way... I also want to know why people think their code is so special that no one else could have ever come up with it independently. Each and every opponent of Copilot is the best developer ever, I guess? That said, I don't understand the choice to use GPL for any reason, so maybe I'm not equipped to under…
"Replicating" it is fine. Violating the license is not.
Re: Give up GitHub: The time has come
#445Would love to see Fossil get wider adoption: https://www.fossil-scm.org/ Single 6mb executable with a version control system, web server, bug database, forums, import/export/sync with git, repo browser, much saner CLI than git, etc. Been using it for 2+ years for all my projects and love it.
I use Fossil. It has some advantages: 1. Hashes are computed without adding any additional headers, so it will be the same as computing the hash normally. 2. The /raw capability is a good thing to have. 3. The deck format is not bad. 4. The command-line interface is less confusing than git. 5. It is written in C. It isn't perfect, but it is more than good enough. (If I do have to change it, I would do "Generalized Fo…
I've read this comment a few times and even unsuccessfully googled /raw and deck format, and this still sounds like GPT-3 to me.
Re: Give up GitHub: The time has come
#446Some of these questions are unfair.
What case law, if any, did you rely on in Microsoft & GitHub's public claim, stated by GitHub's (then) CEO?
Doesn't Google crawl the entire web? How is crawling code different? Copilot is essentially a more intelligent search engine. The only difference is that people want their websites to be searchable and rise to the top of Google they benefit from the increased traffic. Legally I don't see the difference. As Copilot gets more developed, it may become desirable to have your API at the top of Copilot just like search, and this can help drive traffic to FOSS projects. After all, why would a creator not want their FOSS project to be easier to find/use?
why did you choose to only train Copilot's model on FOSS and not on Windows/Office?
Code from Azure and tons of other Microsoft projects is in the training set. Windows and Office are not FOSS and not on GitHub. Obviously it would be a huge security risk to train on OS code.
Can you provide a list of licenses, including names of copyright holders and/or names of Git repositories, that were in the training set used for Copilot?
Again unfair; this is a gotcha type question. There's all kind of code on GitHub with so many different types of licenses. there's bound to be some gray areas and code that inadvertently made it's way into the model. No lawyer would ever expose themselves to that kind of liability.
The bigger issues I see here are copilot is not free but uses free software and that understandably make FOSS community uncomfortable. However, the Copilot models are incredibly expensive to run interact and Microsoft has to cover the bill. Would it be unethical if Google charged you a monthly fee? Arguable not, because it does already in the form of ads. What Microsoft needs to do is have an opt-out standard like the robots.txt file or noindex meta tag. The problem with GitHub is, unlike websites on the web, not everyone uses Github public repos with the express purpose of having them be easily accessible to the public. Another issues is attribution of snippets is a nightmare, but one could argue that devs do with stackoverflow all the time.
Re: Give up GitHub: The time has come
#447Would love to see Fossil get wider adoption: https://www.fossil-scm.org/ Single 6mb executable with a version control system, web server, bug database, forums, import/export/sync with git, repo browser, much saner CLI than git, etc. Been using it for 2+ years for all my projects and love it.
Personally I'd like to see BitKeeper get more adoption also. It has many of the same advantages.
Re: Give up GitHub: The time has come
#448Earlier quoted context omitted.
That's true. I use MIT myself (although I don't care if people cite me). I guess this just all seems rather activist to me but people aren't seeing the big picture; our jobs as programmers are about to change in a very big way. It won't be long before more competition enters the space (e.g. Salesforce and Amazon) and ultimately it won't matter if this model saw some GPL/MIT code because the next one will work twice a…
If you really don't care, please make it known by choosing a license like MIT-0 (MIT No Attribution) or 0BSD (BSD Zero Clause License). You will at least save some people some effort they go to in order to collect and provide attributions to every piece of code used.
Thanks that's very helpful, I'll check them out.
Re: Give up GitHub: The time has come
#449Earlier quoted context omitted.
The reasons that Windows is awful have nothing to do with code quality. Windows is awful because of intentional choices Microsoft made (e.g., bloatware that gets reinstalled with every update, mandatory Microsoft accounts, and mandatory telemetry).
Windows 11 is an example of poor code quality. Bugs everywhere, while the same things work on Ubuntu/popos. Past MS engineers have been commenting for a decade on how MS has grown too big, can't manage, and has become a monolith "too big to fail". By nature when engineers are small pieces of a giant machine, they don't do their best work. And those with the experience move on to better things.
The biggest issue with Windows isn't poor code or shitty engineering, it's the support for legacy software. MS engineers are some of the smartest in the world. The devs can fix the code and make a much better OS but that would break boomer software used by big banks that haven't updated since the 80s. When Microsoft write code, it has to promise support for decades, that means having to maintain the same old outdated APIs for many years.
Re: Give up GitHub: The time has come
#450Earlier quoted context omitted.
There's a setting on GitHub that blocks any suggestions that exactly match code in the training set. I doubt you'd ever get in trouble for code that was similar in structure but different variables etc from existing licensed code (especially since most small snippets of code are not terribly unique to begin with).
Then Microsoft should write this as a legal statement on their part that they will take responsibility for. But I doubt they will ever do that.
Microsoft is not the author of the software that copilot helps produce. the person sitting at the keyboard using copilot is the author.