5 ms·

> While I oppose all forms of intellectual property, this opposition is rarely practiced. I regularly publish software that is copyrighted by me. Trying to rea

> While I oppose all forms of intellectual property, this opposition is rarely practiced. I regularly publish software that is copyrighted by me. Trying to read the article in the most positive way, I agree with it. Copyleft would not be needed if copyright laws were reasonable. But, being a practical person, I understand that we live in a different situation. Copyright exists and is designed to serve a set of interests that usually does not include the general public. Copyleft is a great way of assuring freedom. MySQL is an excellent example on how non-copyleft open licenses may give you freedom in practice but end meaning the end of a product. QT is going the same road. Many "open source" products are freemium products were the "open source" part is good enough to make difficult to justify the creation of competitors but bad enough that anyone with a serious use have to pay for the professional version. Linux is great and free because its copyleft license. It makes it really hard to embrace and extinguish the project. > one of my main goals is for as many people as possible to use it, regardless of whether it's other individuals or giant corporations. This is a false dichotomy. The bigger corporations use Linux and other copyleft products. They are just forced to give back to the community, and only if they offer that product externally. There is limitation on the abuse not the use.

burntsushi 6y ago
> This is a false dichotomy. There is no false dichotomy because I never set up any kind of dichotomy. I'm just listing examples of folks that I want to use my code. And you left off the rest of my quote that explains in a bit more detail: > Given the current intellectual property regime, corporations are incentivized against the use of copyleft software in a large variety of circumstances. Thus, copyleft software, particularly libraries, is regularly avoided by corporations. Since some of these corporations also play a huge role in the FOSS world itself, this leads to systemic network effects that push folks strongly towards the use of permissive licenses. Thus, putting aside ideology and activism, it is simply a reality that if I publish a library with a copyleft license, it's less likely to be used by folks, which is inhibits my goal. The fact that corporations do use copyleft in some places doesn't negate this point at all. It's pretty clear that there are a wide variety of circumstances in which copyleft is avoided. This isn't theoretical or hand wavy. I have direct experience with this. One of the criteria Visual Studio Code used to evaluate ripgrep was its permissive license.[1] [1] - https://github.com/microsoft/vscode/issues/19983#issuecomment-282581996 https://github.com/microsoft/vscode/issues/19983#issuecommen...
- hakre 6y ago
  The licensing requirements of third parties then was the reason you switched to MIT license? Was there a key-moment and would you be okay to share more about it?
  - burntsushi 6y ago
    Hmmm, maybe there is some confusion here. At least for ripgrep in particular, it has always been dual licensed under the MIT and the UNLICENSE[1]. This was itself a continuation of a decision I made a bit earlier with regards to another project[2]. Namely, I had been using only the UNLICENSE. But dual licensing with the MIT allowed folks with a more conservative risk profile to use the project. But the point of linking to VS Code's decision was to say that I have direct experience with situations in which a copyleft license might have inhibited use of my software. That is, it's an example that runs counter to my goal of as many people using my code as possible. This doesn't mean copyleft is never used in popular projects or that copyleft somehow perfectly inhibits popularity. I mean, obviously that's not true. But the existence of popular copyleft projects doesn't mean that copyleft isn't itself a barrier to use in other circumstances. [1] - https://github.com/BurntSushi/ripgrep/blob/master/COPYING https://github.com/BurntSushi/ripgrep/blob/master/COPYING [2] - https://github.com/BurntSushi/byteorder/issues/26 https://github.com/BurntSushi/byteorder/issues/26
- Hokusai 6y ago
  > This isn't theoretical or hand wavy. I have direct experience with this. Yes. I never questioned that part. > corporations are incentivized against the use of copyleft software And what is that incentive? Permissive licenses. Remove all MIT licenses and replace them with copyleft ones and you will see drastic increase on Copyleft usage. Opposing copyleft because corporations is harmful for everybody. I understand why corporations choose non-copyleft and that exact reasons is why copyleft is so important.
  - burntsushi 6y ago
    You have it backwards. The incentive is not permissive licenses. The incentive structure confess from the IP regime itself. People treat patents, trade secrets and so on as things of value in their own right. Copyleft could force the publication of code that companies want to otherwise keep private. So they are incentivized to use permissive licenses to avoid that scenario.
yakubin 6y ago
> Copyleft is a great way of assuring freedom. MySQL is an excellent example on how non-copyleft open licenses may give you freedom in practice but end meaning the end of a product. QT is going the same road. Both MySQL and Qt are licensed under GPL/LGPL, unless you pay for a commercial license. So, if anything, they are examples of copyleft licensing.
- rhn_mk1 6y ago
  On the surface of it, yes, but at least in the Qt case, the Qt Company is exempt from obeying the GPL by requiring contributors to assign them copyright, effectively making the codebase permissive for themselves and only themselves.
red_trumpet 6y ago
> Copyleft is a great way of assuring freedom. MySQL is an excellent example on how non-copyleft open licenses may give you freedom in practice but end meaning the end of a product. QT is going the same road. Can you elaborate on this? According to Wikipedia[1], MySQL is licensed under GPLv2, and iirc that is a copyleft license? [1] https://en.wikipedia.org/wiki/MySQL https://en.wikipedia.org/wiki/MySQL

burntsushi 6y ago

> This is a false dichotomy. There is no false dichotomy because I never set up any kind of dichotomy. I'm just listing examples of folks that I want to use my code. And you left off the rest of my quote that explains in a bit more detail: > Given the current intellectual property regime, corporations are incentivized against the use of copyleft software in a large variety of circumstances. Thus, copyleft software, particularly libraries, is regularly avoided by corporations. Since some of these corporations also play a huge role in the FOSS world itself, this leads to systemic network effects that push folks strongly towards the use of permissive licenses. Thus, putting aside ideology and activism, it is simply a reality that if I publish a library with a copyleft license, it's less likely to be used by folks, which is inhibits my goal. The fact that corporations do use copyleft in some places doesn't negate this point at all. It's pretty clear that there are a wide variety of circumstances in which copyleft is avoided. This isn't theoretical or hand wavy. I have direct experience with this. One of the criteria Visual Studio Code used to evaluate ripgrep was its permissive license.[1] [1] - https://github.com/microsoft/vscode/issues/19983#issuecomment-282581996 https://github.com/microsoft/vscode/issues/19983#issuecommen...

hakre 6y ago
The licensing requirements of third parties then was the reason you switched to MIT license? Was there a key-moment and would you be okay to share more about it?
- burntsushi 6y ago
  Hmmm, maybe there is some confusion here. At least for ripgrep in particular, it has always been dual licensed under the MIT and the UNLICENSE[1]. This was itself a continuation of a decision I made a bit earlier with regards to another project[2]. Namely, I had been using only the UNLICENSE. But dual licensing with the MIT allowed folks with a more conservative risk profile to use the project. But the point of linking to VS Code's decision was to say that I have direct experience with situations in which a copyleft license might have inhibited use of my software. That is, it's an example that runs counter to my goal of as many people using my code as possible. This doesn't mean copyleft is never used in popular projects or that copyleft somehow perfectly inhibits popularity. I mean, obviously that's not true. But the existence of popular copyleft projects doesn't mean that copyleft isn't itself a barrier to use in other circumstances. [1] - https://github.com/BurntSushi/ripgrep/blob/master/COPYING https://github.com/BurntSushi/ripgrep/blob/master/COPYING [2] - https://github.com/BurntSushi/byteorder/issues/26 https://github.com/BurntSushi/byteorder/issues/26
Hokusai 6y ago
> This isn't theoretical or hand wavy. I have direct experience with this. Yes. I never questioned that part. > corporations are incentivized against the use of copyleft software And what is that incentive? Permissive licenses. Remove all MIT licenses and replace them with copyleft ones and you will see drastic increase on Copyleft usage. Opposing copyleft because corporations is harmful for everybody. I understand why corporations choose non-copyleft and that exact reasons is why copyleft is so important.
- burntsushi 6y ago
  You have it backwards. The incentive is not permissive licenses. The incentive structure confess from the IP regime itself. People treat patents, trade secrets and so on as things of value in their own right. Copyleft could force the publication of code that companies want to otherwise keep private. So they are incentivized to use permissive licenses to avoid that scenario.

yakubin 6y ago

> Copyleft is a great way of assuring freedom. MySQL is an excellent example on how non-copyleft open licenses may give you freedom in practice but end meaning the end of a product. QT is going the same road. Both MySQL and Qt are licensed under GPL/LGPL, unless you pay for a commercial license. So, if anything, they are examples of copyleft licensing.

rhn_mk1 6y ago
On the surface of it, yes, but at least in the Qt case, the Qt Company is exempt from obeying the GPL by requiring contributors to assign them copyright, effectively making the codebase permissive for themselves and only themselves.

red_trumpet 6y ago

> Copyleft is a great way of assuring freedom. MySQL is an excellent example on how non-copyleft open licenses may give you freedom in practice but end meaning the end of a product. QT is going the same road. Can you elaborate on this? According to Wikipedia[1], MySQL is licensed under GPLv2, and iirc that is a copyleft license? [1] https://en.wikipedia.org/wiki/MySQL https://en.wikipedia.org/wiki/MySQL