> Well, although SPDX counts 665 licences, there really just 3 main kinds:
> 1. licences with no restrictions (like MIT)
> 2. licences that require you credit the original author ("attribution" licences, including the Apache Licence)
> 3. licences that require you credit the original author and that derivative works have the same licence ("copyleft"/"share-alike" licences like the GPL)
MIT requires attribution, doesn't it? MIT (permissive) / MPL (non-viral copyleft) / AGPL (viral copyleft) seems like a better grouping to me; I rarely find myself reaching for any other licenses.
I do wish there were a shorter copyleft license though. I appreciate how transparent and readable MIT is.
It's just weird that it has "Microsoft" in the name. Though I suppose not much weirder than MIT license being named after a University or BSD license being named after a University's Unix distribution.
Also even Microsoft doesn't use the MS-RL much anymore having standardized more on MPL where they use copyleft licenses and Apache License most everywhere else.
Essentially licensing your software like this behaves like ASL unless you: modify + distribute (either binaries or by creating a service). Then you owe the changeset back, but it does not have a viral clause like the AGPL.
This solves a large part of the greedy AWS problem (Amazon copying entire open source projects and contributing nothing back), but also strikes a balance and allows API Compatibility.
I really like the EUPL on paper, and I've been told by Joinup's legal support that it should be a valid "change licence" for BUSL, in case I ever want that.
But I'm concerned about the compatibility clauses becoming a loophole for hostile forks. Then again, half the point of the EUPL is admitting that only a court can judge what is or isn't a derivative work (unlike the legal fiction in the GPL's viral clause), so I guess these uncertainties are part of the deal.
This may be a hot take, but I love the un-usable for business licenses and the fun multiple lowest-to-highest paid worker and ethical restrictions (No use in weapons, for example) licenses, but not for their direct restrictions but rather because almost no large business will ever want to touch them, but small players and individuals don't care.
Those licenses let me say "This is open to the individual and small business, but not a mega corp" without actually needing to define a hard cut off.
Besides, it's not like most developers of FOSS software that use these have the time/money/energy to bother to sue over infringement anyway, so practically this is their main purpose.
I'm a big fan of CC0. It's my go-to for any side projects I work on, for all kinds of reasons, but mostly just because I feel it minimizes economic deadweight loss by incurring zero additional transaction costs.
CC0 wasn't intended for code and you can be hurt by liability/warranty issues. Creative Commons themselves suggest pairing it with an Unlicense or 0-Clause BSD.
> Well, although SPDX counts 665 licences, there really just 3 main kinds:
> 1. licences with no restrictions (like MIT)
> 2. licences that require you credit the original author ("attribution" licences, including the Apache Licence)
> 3. licences that require you credit the original author and that derivative works have the same licence ("copyleft"/"share-alike" licences like the GPL)
MIT requires attribution, doesn't it? MIT (permissive) / MPL (non-viral copyleft) / AGPL (viral copyleft) seems like a better grouping to me; I rarely find myself reaching for any other licenses.
I do wish there were a shorter copyleft license though. I appreciate how transparent and readable MIT is.
The author must have meant something like Zero-Clause BSD, equivalent to public domain in the US. https://opensource.org/license/0bsd
The MS-RL is still the shortest, clearest copyleft license I'm aware of it. It is nearly as readable as the MIT. https://opensource.org/license/ms-rl-html
It's just weird that it has "Microsoft" in the name. Though I suppose not much weirder than MIT license being named after a University or BSD license being named after a University's Unix distribution.
Also even Microsoft doesn't use the MS-RL much anymore having standardized more on MPL where they use copyleft licenses and Apache License most everywhere else.
Open license #666: the goatse license, which is absolutely not rubbish https://github.com/153/goatse-license
My favorite is EUPL: https://joinup.ec.europa.eu/collection/eupl/introduction-eup...
Essentially licensing your software like this behaves like ASL unless you: modify + distribute (either binaries or by creating a service). Then you owe the changeset back, but it does not have a viral clause like the AGPL.
This solves a large part of the greedy AWS problem (Amazon copying entire open source projects and contributing nothing back), but also strikes a balance and allows API Compatibility.
I really like the EUPL on paper, and I've been told by Joinup's legal support that it should be a valid "change licence" for BUSL, in case I ever want that.
But I'm concerned about the compatibility clauses becoming a loophole for hostile forks. Then again, half the point of the EUPL is admitting that only a court can judge what is or isn't a derivative work (unlike the legal fiction in the GPL's viral clause), so I guess these uncertainties are part of the deal.
Most of the licenses discussed in the article are demonstrably NOT open source licences at all.
This may be a hot take, but I love the un-usable for business licenses and the fun multiple lowest-to-highest paid worker and ethical restrictions (No use in weapons, for example) licenses, but not for their direct restrictions but rather because almost no large business will ever want to touch them, but small players and individuals don't care.
Those licenses let me say "This is open to the individual and small business, but not a mega corp" without actually needing to define a hard cut off.
Besides, it's not like most developers of FOSS software that use these have the time/money/energy to bother to sue over infringement anyway, so practically this is their main purpose.
I'm a big fan of CC0. It's my go-to for any side projects I work on, for all kinds of reasons, but mostly just because I feel it minimizes economic deadweight loss by incurring zero additional transaction costs.
CC0 wasn't intended for code and you can be hurt by liability/warranty issues. Creative Commons themselves suggest pairing it with an Unlicense or 0-Clause BSD.
https://opensource.org/license/unlicense
https://opensource.org/license/0bsd
Oh, do they? I always thought it was considered basically identical to those.
Brb, dual licensing everything under 0BSD.
The moment that hits 666, it ticks right on over to 667.
People have their beliefs; and not only does no-one want to release The Satanic License, no-one's gonna want it to remain that unlucky for long.
Weird little monkeys we are, for the amazing things we can be.
> no-one wants to release The Satanic License
You hang out with a different crowd than I do then. Perhaps the Satanic Temple should release an open source license to claim the #666 spot.
The classic Smoking Blood-Drenched Apparition With Fangs (SBDAWF) license is a close runner-up. https://duckduckgo.com/?q=%22sbdawf%22&ia=web
I'm getting flashbacks to Intel's 667MHz processors
Need one more.