ogres has quit [Quit: Connection closed for inactivity]
<kentonv>
oh hai, apparently there was a license flamewar while I wasn't paying attention
<kentonv>
I don't think I've ever licensed anything I've written under GPL, whether or not I was trying to build a business around it. I completely respect people's right to specify any restrictions they want on their works, but I think that, practically speaking, such restrictions are ultimately detrimental
<kentonv>
to me, the purpose of open source is to enable innovation by advancing the starting point -- for everyone. It's not productive for people to have to constantly rewrite the basic infrastructure they build on top of.
<kentonv>
in the old days, when everything was proprietary, you had to pay a lot of money, or you had to write things from scratch
<simpson>
kentonv: I mean, as a library author, you might be biased in that direction. But I guess that I have that bias, too.
<kentonv>
now we have a pile of open source frameworks you can build on to make fast progress. And it's great! Whether you're a hobbyist, a community project, or a business, you can do something actually innovative much more quickly.
<kentonv>
but if a framework is GPL, then, realistically, that restricts who can build on it. Which is fine -- the author of the code has the right to place such restrictions on their work. But the really popular infrastructure is, usually, not GPL.
<kentonv>
and GPL projects that are sufficiently popular often end up being rewritten for the sole purpose of having a liberal license, which is kind of sad
<kentonv>
simpson, I agree there's a huge difference between libraries and applications here
<kentonv>
for application code, yeah, I think there's little difference between GPL vs. Apache vs. MIT
<kentonv>
I think the threat of someone stealing your work and extending it is overblown. It just doesn't happen that often in practice, as far as I can see.
<kentonv>
I mean, trying to create a business that immediately has an open source product as a direct competitor (because it is forked from said product) is probably not going to go well. It's much better for the business, usually, to contribute back to said product, even if they don't have to.
<kentonv>
What you do see plenty of is, say, some device, like a DVD player or whatever, that happens to use Linux and Busybox in its implementation (invisible to users), but forgot that they needed to distribute the source code...
<kentonv>
I'm not sure those device makers are making any relevant contributions to the tools they use anyway, so I'm not sure GPL is really winning a lot there. I could be wrong.
<kentonv>
I think that raising the baseline from which people build enables more ideas to be tried out, because it's less expensive to try something. I'm comfortable with startups trying out ideas -- starting from the common baseline, and keeping their code closed. If their idea works, someone will make an open source clone of it soon enough. That's a much better situation than if the idea had never been tried in the first place,
<kentonv>
in my opinion.
<kentonv>
FWIW, business-wise, Sandstorm could probably have operated exactly the same with a GPL license, and probably even AGPL. The significant parts of blackrock that we initially kept closed were not "derivative works" of Sandstorm but rather add-on technology that would not have been "infected" by Sandstorm itself being AGPL. We may have had to release the Stripe billing integration, but that's minor compared to the clusteri
<kentonv>
ng code.
<kentonv>
but even if we had to release the clustering code... I think Sandstorm's real long-term business would have been around running the app store, which would not have been harmed by Sandstorm and Blackrock being 100% open source.
<kentonv>
though, that business model does require that apps on Sandstorm are allowed to be proprietary
<kentonv>
which I am 100% comfortable with
<kentonv>
(A few times, people have asserted that closed-source apps shouldn't be allowed on Sandstorm. I actually take offense to this assertion. Only Sandstorm's authors have any right to say how Sandstorm should be allowed to be used.)
pie__ has quit [Ping timeout: 240 seconds]
<kentonv>
Also, FWIW, I have a hard time believing that license has any significant effect on volume of contributions.
<kentonv>
The main thing that governs volume of contributions, I think, is how easy it is to figure out how to contribute. Projects which are cloning an existing product (e.g. Rocket.Chat cloning Slack) attract lots of contributions because it's very very easy for anyone to decide what would make a good contribution, and then do it.
<kentonv>
Sandstorm is so experimental and unique that it makes it a lot harder for people to jump in.
<kentonv>
Except by writing / packaging apps, which is conceptually straightforward -- so that's where most of the external contribution has been
neynah has joined #sandstorm
<ocdtrekkie>
I finally pushed the "upgrade grains" buttons on Etherpad and Wekan updates from like... a long time ago.