Gpl ensures freedom.

You can always sell your code to someone with a different license if he wants to use it. If you just give it to him for free, he uses your code, time, effort, money and makes real money. You don't see a cent. If you sell your code to him for a small fee under a differene licence than gpl, he can make real money and compensate you for your time. With MIT you just give it away for free, valuing your time at zero.

Gpl just ensures that someone else is working with you and not abusing you.

If you start a project with MIT, someone can fork it, create a billion dollar company with it and you are stuck with nothing.

In the real world, people copy your code anyway even if it is licensed under gpl and since they don't have to publish it, it's difficult to detect and sue them.

Gpl was, is and will be free.

As a professional researcher, you are already compensated by someone, e.g. the gov, and you could argue that it is the goal to publish your code for free such that anyone can use it to make a billion dollar company. Ultimately, you are free to choose the license.

With gpl, everyone profits and with MIT someone profits and probably not the author.

I've settled on publishing all researxh-related code under AGPL. My reasoning is that academic research is funded by the public, so naturally the code should be public. Taken a step further, by licensing it under AGPL, I am ensuring that companies who may use the code for profit have to keep that code public and cant just privatize it and lock the public out.

Its unlikely anyone would ever even use my code though so 🤷

How would you feel if Microsoft (for example) took your research code and turned it into a proprietary software product that made them billions of dollars a year, which you could buy and run, but could not modify, inspect its operation, or make use of the modifications that Microsoft made to your work outside of their proprietary product?

Your answer to that question is what determines which license you should use.

I personally do cc0 (though i do not have much published). i can not really be bothered to do proper licensing, and my prefered license (wtfpl) is not considered good (by admins, because it has the f word , boo hoo).

I see it this way: by choosing BSD or MIT over GPL, you're just protecting someone's freedom to take your work and prevent people from using a continuation of it. They can still use it and change it, they just need to give their version to everyone, making even more tools available to everyone to use however they want. When it comes to science, too much is paywalled and restricted already. I wouldn't want to let someone paywall any more stuff if I can help it.

Ultimately it's your work so it's up to you how you want to release it. BSD/MIT aren't really any more or less free than the GPL because they still guarantee the four freedoms. The GPL just prevents downstream projects from denying those four freedoms further down stream, which is seen as important in the free software movement, but it doesn't have to be to you.

One thing to keep in mind with these permissive or "pushover" type licenses although they are free software licenses, normalizing them means that the proprietary software industry ultimately gets to choose what is allowed to be released as free software. There is a warning that "business friendly" free software licensing does not ultimately mean business will be friendly back, especially in an age where there is increasing concern over proprietary software companies taking advantage without either giving back nor funding upstream projects.

It's your choice; there is no "must". Yes, licensing your work with a fully free license like BSD means other people can profit off your work, but so does publishing your work in an academic journal. Copyleft licenses like GPL do nothing to prevent others from using your work to draw conclusions or without distribution to make a profit, either (though CC Share-Alike licenses do). You should also recognise that the likelihood of someone taking academic code and actually using it in something else is often (it depends on the field) very slim.

I was surprised by the fediverse obsession with copyleft and insisting that all projects ought to use it otherwise they're traitors to open source and themselves.

I use MIT for all my scientific work. Very happy with it.