Jumping the licensing shark
Jumping the licensing shark
Posted Mar 23, 2023 8:04 UTC (Thu) by LtWorf (subscriber, #124958)Parent article: Jumping the licensing shark
What advantage does adoption with no freedom bring me? I've worked on the thing for free and published it.
I want to help other libre software, I don't want to help companies hire fewer people (the very same people that in their spare time make cool libre software) to make more money. I want more libre software to be available.
Posted Mar 23, 2023 9:36 UTC (Thu)
by jmalcolm (subscriber, #8876)
[Link] (4 responses)
If your goal is to have more libre software available, I am happy that you have copyleft licenses to choose from so that you can create such software ( or even just promote its creation ).
For myself, I am very happy that not all open source licenses are copyleft because I want more open source software available and believe very strongly that, if it all had to be copyleft, we would have a lot less of it.
As for what good it does me to have my software widely adopted, there are many reasons but of course one of the main benefits is that some of those that adopt the software will contribute back. Some copyleft advocates like to pretend that does not happen but of course it does.
Both Xorg and Wayland are governed by the MIT license. Do you benefit from those?
Do you game on Linux? Mesa is MIT licensed as well which allows lots of corporate participation which means better drivers. It is hard to overstate the value for Linux gamers of Proton which of course is a result of Valve adopting Wine to port Windows games as part of their Steamdeck and Linux strategy. Wine is LGPL which is only "weak" copyleft.
Perhaps the most important software created by the FSF is GCC. I prefer Clang and am thankful that Apple for contributed so heavily to this BSD licensed compiler. Despite the availability of GCC, Clang / LLVM has attracted usage and contributions from Microsoft, IBM, Qualcomm, Intel, Meta, and others. I benefit tremendously from the fact that these companies use this software. Mozilla was able to use LLVM as the back-end when they created Rust which is itself released under Apache 2.0. Apache includes a patent grant and is perhaps my favourite license. Amazon, Google, and Microsoft all contribute to Rust. We will all benefit from that I believe.
Anyway, my core point is that users, even commercial users, of software give back to software governed by permissive licenses. There is also a lot of incentive to contribute to the main projects instead of just forking them or taking them entirely internal ( which is still fine by me ). The GPL does not have some kind of exclusive on expanding the universe of open source software that is available for me to use.
In fact, I argue at this point with Open Source being so mainstream that GPL projects are at a disadvantage. A lot of popular GPL software gets "used" by companies ( eg. on the web ) but they do not "distribute" it and do not contribute back even their changes any more frequently than they would if the license was more permissive.
Perhaps the most successful GPL software out there is the Linux kernel. Linux has dominated everywhere. That said, while I see lots and lots and lots of Linux usage in embedded, I rarely see source code. The vast majority of the companies using Linux, especially embedded, are violating the GPL. If the license had been strictly enforced, I wonder if a permissively licensed alternative might have dominated instead ( eg. a BSD perhaps ). Who knows.
Posted Mar 23, 2023 12:52 UTC (Thu)
by pizza (subscriber, #46)
[Link] (1 responses)
There are some technical reasons why GCC was[/is/remains] inferior, but the overwhelming majority of the corporate interest in LLVM has been because it's not copyleft -- that is, one can provide binaries without supplying source code. So in effect it's a way for folks to drastically cut their costs wrt developing/maintaining their own proprietary toolchains.
I'm not saying that's not a worthy goal in of itself (from their perspective definitely, but also in the sense that end-users get a higher-quality proprietary toolchain then would have otherwise happened), but at the end of the day the results are still largely proprietary.
At one point I had five LLVMs on my system, four were proprietary from one vendor or another, and of those, two required a connection to a license server to do anything. (The other was the distro-supplied LLVM package. Actually make that six; Mesa's embedded copy was different). The fact that they were built on "open source" LLVM was irrelevant to me, as the only ones with the ability to pull in a bug fix or add a new feature was the vendor.
Posted Mar 23, 2023 18:16 UTC (Thu)
by jhoblitt (subscriber, #77733)
[Link]
This sounds like a bug, not a feature.
Posted Mar 23, 2023 16:19 UTC (Thu)
by bkuhn (subscriber, #58642)
[Link] (1 responses)
I think folks who make this argument tend to count proprietarized deployments of code that was previously FOSS *as* having more FOSS. I disagree with that assumption; if it's proprietary *now*, it doesn't matter that it came from some FOSS upstream IMO. Proprietary software is proprietary, no matter its earlier provenance.
Posted Mar 23, 2023 16:59 UTC (Thu)
by NYKevin (subscriber, #129325)
[Link]
In other words, the people making this argument do not care at all about whether the amount of proprietary software goes up. They only care about whether the amount of FOSS goes up.
Jumping the licensing shark
Jumping the licensing shark
Jumping the licensing shark
Jumping the licensing shark
Jumping the licensing shark