do closed source / non-OSI licensed computational models meet OMF standards?

[jedbrown]

By way of example, NSF's MolSSI Community Code Partners includes Gaussian and VASP, both of which have non-open licenses that forbid publishing benchmarks (cf, Banned by Gaussian). A purist would mandate OSI-approved licenses in hopes of incentivizing transparency and reproducibility, but that will exclude significant research sectors. Alternatively, it could be tiered and best-effort, allowing the status quo to persist, though that undermines some of the objectives of OMF.

[alee]

IMO we should take a strong stance as the Open Modeling Foundation that requires transparency in any blessed digital artifacts - no room for closed source anything that would meet the minimal standards guidelines..

[alee]

related to #90

[cmbarton]

I strongly agree. This will make some people uncomfortable. But as one participant reiterated today, if you are doing science, you need to be transparent. That means that we may not have a lot to offer industry. But maybe more than it seems.

[platipodium]

While I personally agree with @cmbarton, I fell that as on organization we should not exclude non-OSI codes. Some thoughts

1. We might require the metadata to be FAIR and open, this would make many non-OSI codes at least FAIRer
2. we have the problem of open but not OSI codes that cannot be relicensed
3. some closed source is still "fair" and interacting with open source codes. That's one of the reasons we have the LGPL