[hoaw]

The whole point of the license is presumably that they disagree with that definition. As far as I know a license can be OSI approved without containing patent grants, effectively making the code unusable for commercial purposes. There is also, again as far as I know, nothing stopping companies from using contracts to restrict the use of code in at least some OSI approved licenses.

That said, I do think they should state that it is "non-commercial open source".

[icebraining]

A definition is not an opinion, it's an arbitrary mapping from a term to a description; you can't really disagree with it. You're free to use another definition for the same term, but if you don't make it extremely clear to you're departing from the most commonly used, don't be surprised if people treat you as any other sleazy salesperson.

The patent grant is a red herring; they're copyright licenses, and judged as so. Unlike software copyright, software patents are not even valid in many countries.

[hoaw]

> A definition is not an opinion, it's an arbitrary mapping from a term to a description; you can't really disagree with it.

How open something has to be, to be called open source is a matter of opinion, so is to some degree what open means in the context of open source. Otherwise they would have to call it "unconditional open source" but then a number of, if not most, licenses wouldn't qualify.

> The patent grant is a red herring; they're copyright licenses, and judged as so.

It isn't a red herring. I use it as an example to show that the openness in OSI approved licenses aren't absolute. Some licenses have other terms. They might for instance retain the moral rights of the author, try to avoid any liability or condition the distribution of software. If you can have those exceptions and still be considered open source I don't see an objective reason why you can't call software where the economic rights are retained open source as well (even though I can understand why people wouldn't want that).

[icebraining]

You're asking why should it mean X or Y, but my argument is that regardless of why it came to have the current definition, with its specific idiosyncrasies, changing it (to anything else) - and especially expanding it - is itself bad, because it makes discourse more confusing and therefore the term less relevant and useful.

There's nothing special about it, they're just two English words. What makes the term special is its origin and history - specifically, how it was coined and spread by the OSI and its members.

So, they should make their own history with a new term. In time it might be more relevant than open source, and that might be great. But don't mix them up.

[hoaw]

I think you are missing my point. An organization like the OSI might define what open source is. But that doesn't make their definition the meaning of open source. The meaning becomes what open source is. Since you have many different licenses under the term open source, some of which contain the non-absoluteness I described earlier, the meaning of open source becomes "software that you can modify etc. under certain conditions". That is why you have open source under a permissive license, under a copyleft license, under a patent granting license etc. That is the historic president. There is therefor no reason why you couldn't have open source under a non-commercial license, other than that this doesn't meet the OSI definition, but these other exceptions do.

But as I said I would still prefer "non-commercial open source". I grants you the right to modify etc. but only if you don't exploit it commercially. Just like copyleft open source grants you those rights, but only if you also distribute the source code.

Apparently though I guess if there was something that was a red herring it is this whole discussion as someone pointed out that common clause doesn't even call the license open source.