Choose a license and make repo public

[BrianJKoopman]

I think the discussion point of making the code public has come up in the past, the sticking point being the need to choose a license before doing so.

I successfully got the sisock/grafana stack running on NERSC's Spin system yesterday, and it's clear from working on that system that getting the stack running on other computers (including the test institutions) would be simplified by having our built containers available on Docker Hub. This would avoid users needing to build their own containers, and would simplify upgrades when needed.

If we don't want to open things up, I'd look into making a private registry (which is how I'm deploying on Spin currently, using the one NERSC provides).

Assuming that's not the case though, thoughts on licensing?

[nwhitehorn]

I would vote for 2-clause BSD (the citation license) as a universal standard.

[newburgh]

I definitely think we should make it public, I don't have as clear an opinion on the licensing (but whatever is most common is probably fine)

…

On Tue, Dec 18, 2018 at 11:14 AM Brian Koopman ***@***.***> wrote: I think the discussion point of making the code public has come up in the past, the sticking point being the need to choose a license before doing so. I successfully got the sisock/grafana stack running on NERSC's Spin system yesterday, and it's clear from working on that system that getting the stack running on other computers (including the test institutions) would be simplified by having our built containers available on Docker Hub <https://hub.docker.com/>. This would avoid users needing to build their own containers, and would simplify upgrades when needed. If we don't want to open things up, I'd look into making a private registry (which is how I'm deploying on Spin currently, using the one NERSC provides). Assuming that's not the case though, thoughts on licensing? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#18>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AG7iZf8Z55on7JC3aALyAldZ_QdnNNcWks5u6RROgaJpZM4ZYndf> .

-- Laura Newburgh *Assistant Professor of Physics* Wright Laboratory Yale University Department of Physics WL - 210 56 Hill House - 202 phone: 203-432-9168 fax: 203-432-3522

[BrianJKoopman]

I'm fine with the 2-clause BSD. Some of the crossbar templates we've used are MIT licensed, which seems nearly identical, and my understanding is it is fine for MIT licensed code to be sublicensed under BSD licenses. I'm curious what @ahincks thinks, since he's the original author.

[ahincks]

Sorry, have been away from the internet for the past few days. I don't have strong opinion on the licensing: maybe double check that the existing MIT is compatible with the BSD?

Given @nwhitehorn's comment, is this a broader discussion: i.e., do we have/need/want a common standard for Simons?

[mhasself]

Yes, we need a common standard for SO. I've opened an issue in the so_dev_guide to track this.

[BrianJKoopman]

[M]aybe double check that the existing MIT is compatible with the BSD?

I get the sense from reading a few posts that these two are compatible (which makes sense to me, they read almost identically.)

https://opensource.stackexchange.com/questions/217/what-are-the-essential-differences-between-bsd-and-mit-licences

[BrianJKoopman]

This is done. The sisock repo is 'public' as of last week.