Live data from Hacker News

Getting forked by Microsoft

philiplaine.com

651–660 of 1001 posts

Re: Getting forked by Microsoft

#651
post #630

I want to make a point that might be misinterpreted, so I want to make clear I am not at all defending Microsoft. That said, Microsoft isn't a person and has no agency by itself. It is specific persons/developers/managers violating the licenses and stringing along open source developers in bad faith. Who are these people? Why is the blame not falling on them, specifically?

ehm, it doesn't work this way, fortunately ... Microsoft, the corporation, is definitively responsible if there's a copy right violation.

Who exactly did what it's a Microsoft internal thing, unless Microsoft demonstrates that this has been done in bad faith and Microsoft did everything what is "reasonable" to avoid this happening ...

Re: Getting forked by Microsoft

#652
post #426
post #396

Earlier quoted context omitted.

It's in the license file: https://github.com/spegel-org/spegel/commits/main/LICENSE

Right. So Microsoft should just have a copy of this LICENSE file somewhere? Can't we just open a PR to add it to the repo? Did the author do that and did Microsoft decline the PR? Feels like Microsoft was not necessarily trying to steal work (they link the original project in their README).

I think as a bare minimum, they should have kept the original LICENSE, and add theirs on top or something.

Re: Getting forked by Microsoft

#653

There's a lot of blame being assigned to Microsoft, the entire corporation. But I doubt this was a heavily contemplated decision by a room full of executives, or voted on by the shareholders. More likely, this is a way for someone to get ahead in their career at Microsoft by passing off a successful open source project as their own accomplishment. They can steal users from the original project and justify using Micro…

Just because the shareholders didn't vote on it, or an exec didn't explicitly say "hey steal this" does not absolve the company. Leadership doesn't get to throw up their hands and say "not my fault" when something bad happens. It is ultimately the responsibility of the company and its people to create a system where things like this are discouraged or prohibited. Not doing so is tacit approval, especially in this cas…

It's fine that you think corporations are supposed to work that way, and I don't necessarily disagree. But they don't in practice. They don't feel the consequences of bad actions because of legal economies of scale. They also don't backpropagate consequences from the company's bottom line to the individuals responsible. If you were to rectify this so that it works exactly as you envision, you would have made incredible advances in the Principal-Agent problem as it pertains to corporate compensation.

Most corporate actions that 3rd parties consider "bad" are the result of someone inside the corporation having an asymmetric payoff from directing the corporation to do the bad thing. They get the upside from a success, but not the downside from failure.

If you want to stop a certain bad behavior, your best bet is to change individual incentives.

Re: Getting forked by Microsoft

#654
post #19

> I default to using the MIT license as it is simple and permissive What's good about being "permissive"? I keep hearing this argument, but I still don't understand, what's the incentive for authors of one-man projects to choose anything "permissive". Do you enjoy your project getting forked, walled off and exploited for profit by someone who has never done you any good? AGPLv3 still allows forking, still allows maki…

Permissive licenses are about contributing to the trade as a whole, rather than individualism.

Some of us don't believe that the code we write is "ours" in any meaningful way, and don't think strangers using it have any obligation to us just because we typed it once long ago.

Personally, I am happy if my code is of use. If people are using it for evil I'll fight the evil, not try to withhold good things from the world to avoid that possible case. It is an approach that is rooted in sufficiency mindset, rather than capitalistic notions of false scarcity.

My project being forked doesn't cost me anything at all, but caring about it being forked or enforcing a license would cost me time and energy I have no desire to spend. Permissive licenses accurately communicate the levels of fucks I give, while keeping assholes from trying to sue me over having used my contributions to the collective wealth of the profession.

If I make the world better for everyone, of course a bunch of people who never did anything for me are going to be a part of "everyone", basically by definition. What is wild here is that Microsoft didn't follow the extremely minimal requirements of the permissive license.

Re: Getting forked by Microsoft

#655

Earlier quoted context omitted.

.NET is open source and people working on it go into great lengths to ensure it is a good citizen to open-source projects and communities. It has been open source for almost 10 years damn it. All in all what other divisions or teams do is greatly unfortunate because it will get associated with the aforementioned. Personally, this annoys me because other languages like Go or Swift do not receive the same criticism for…

Do these links have much if any merit? I would have to re-check their claims though. As I said, I have misconceptions of .NET, so it is always useful to get to the bottom of it.

The hot reload drama was real, and the decision was backtracked. The rest? I don't think it has any relevance as of today. Many other languages have worse situation when it comes to tooling. Right now, in .NET you can use Visual Studio, Visual Studio Code or any of its forks, Rider, which is now free for non-commercial use, and also Neovim/Emacs/anything which supports LSP and DAP.

Hot reload in general is difficult to make work in something that is mainly compiled, for example it does not work with F# right now, but there is someone in community working on making it a possibility. It's regular activities you'd see in other ecosystems.

E.g. I think NetCoreDbg, as an alternative to closed vsdbg that has usage restrictions, works well enough to fully enable the standard workflow when using VSCodium/Cursor/Neovim/etc. I know people use the latter with both C# and F# without sacrificing user experience in comparison to languages like Rust. It's just text editor, language server + debugger integration and CLI. You would hear about "refactorings" and "advanced features" from those who are used to more IDE-like experience provided by VS or Rider but, for example, many refactorings are also available in VSC/VSCodium because they are just a feature of the language server based on Roslyn analyzers and auto-fixers. It works with anything that integrates that and the language server itself ships with SDK to my knowledge.

All in all, the tooling situation is pretty good with multiple IDEs, commercial and community tools offered to be able to program in .NET languages, most languages HN loves to sign praises to do not have this. The same applies to GUI frameworks too - it's funny to read how .NET is "anti-linux" because out of AvaloniaUI, Uno, MAUI and a bunch of smaller libraries MAUI does not happen to target Linux. Some people just like to hate something, and if the reason for that goes away they come up with a new one.

Re: Getting forked by Microsoft

#656

While Microsoft is certainly in the wrong for removing the copyright notice, I think the author has zero basis for complaint otherwise. If you're going to release software with one of the most permissable licenses, you need to accept that for all it entails. Consider what you're comfortable with and pick an appropriate license relative to your values.

Did they complain about anything else?

Yes, he complains in the last few paragraphs that he feels like this form is a competitor. Says that users sometimes come to him asking for help with the Microsoft fork, etc. Those all very much fall into the domain of "what did you think MIT meant exactly", imo at least.

Re: Getting forked by Microsoft

#657
post #579

Earlier quoted context omitted.

Yes, charity. That's exactly what these trillion dollar empires think of those open source maintainers. Microsoft pulled this same stunt multiple times on os maintainers.

Open source has been hijacked by trillion dollar hyperscalers. It's time we switch to "fair source" or "equitable source". Put MAU/DAU/ARR/market cap limits in your license. Open to everyone with a market cap under $1B or revenues under $100M. All others, please see our "business@" email. Place viral terms like the AGPL that requires that all other systems touched by your code to be open - especially the backend/serv…

I don't understand why we don't just lean into the "osi = corporate, copyleft = good faith" model that's worked perfectly well for the last thirty years.

Re: Getting forked by Microsoft

#658

Earlier quoted context omitted.

Open source has been hijacked by trillion dollar hyperscalers. It's time we switch to "fair source" or "equitable source". Put MAU/DAU/ARR/market cap limits in your license. Open to everyone with a market cap under $1B or revenues under $100M. All others, please see our "business@" email. Place viral terms like the AGPL that requires that all other systems touched by your code to be open - especially the backend/serv…

Those companies can produce legal abstraction hacking solutions faster then you can develop shielding ones. You needs something poisonous ,costing money or work with each usage preventing mass adoption without a complete rewrite .

Open source will inevitably succeed, but only in the long run. In the short term VC (or tech giant) cash will dominate any conversation. There's absolutely nothing you can legally do from preventing reimplementation (which is a good thing, because it means over the long term we will reimplement everything as free software).

Re: Getting forked by Microsoft

#659

Earlier quoted context omitted.

Assuming that win.com is able to be stand alone and doesn't require win.bat, then yes, Bill can license both of these components separately, one under the GPL and another proprietary. The Free Software Foundation (FSF) describes this copyleft aspect of the GPL in terms of "derivative works" associated with GPL-licensed software. When two components are related to each other in a derivative way, then the GPL says that…

The GP says: >>> At this point I'd include some of the code as binary blobs and "pay me for the source!". In addition to GPL! So, the proposal is to hide the source code and IIUC if someone does this, the whole complete project can not be released as GPL.

That's incorrect. As the original author of the work, you can release the project under whatever license you choose. Doing so may make it impossible for someone else to meaningfully comply with it, but that's their problem, not yours. It doesn't stop you from choosing the GPL, even if it's a bizarre option for that particular project.

Re: Getting forked by Microsoft

#660
post #19

> I default to using the MIT license as it is simple and permissive What's good about being "permissive"? I keep hearing this argument, but I still don't understand, what's the incentive for authors of one-man projects to choose anything "permissive". Do you enjoy your project getting forked, walled off and exploited for profit by someone who has never done you any good? AGPLv3 still allows forking, still allows maki…

Because when people start an OS project, they want to help people and grow. MIT license is the best license if your goal is to help other people. It's the worst license for building a business, but that's usually not what people think about when starting a project
Post reply on HN