MO_942551040047_* citation

The “source citation” listed for the model does not match the Extended KIM ID or the citation listed in .eam.alloy file, which do agree. I believe the extended ID is correct but the source citation is wrong.

Hi Zach,

This potential was developed by Cai and Ye but was submitted to LAMMPS by
Sang-Pil Kim. We sent Kim several messages as his last known address at Brown
but never received a response.

Thus, the reference that is listed is the one provided by the submitter. Under our current policy/system there is no way to change the officially listed reference (it becomes "locked in" when the KIM item is published). However, I suggest that you add a comment about the original publication to the item's wiki page. This will be a good way to provide this information on the openkim.org site in close proximity to the item itself.

Cheers and thanks!

Ryan

I made a wiki contribution, twice: _000, _001

I have a few concerns:
(1) Current policy/system: Is this the best policy for orphans with
indisputably inaccurate metadata? Can the original developers assume
ownership of their own models that were submitted by a third party?

(2) Two wikis for one "model" with small "implementation" differences. Why
should a new API release require mode developers maintain an additional
wiki page? The underlying EAM model did not change.
(3) Blame: I suspect that KIM repository admins can see that I made a wiki
edit. I the spirit of wiki, shouldn't the public see the complete history
of all edits to the wiki?

Cheers,

-Zach

I recall a discussion of a possible procedure for allowing entries with unresponsive contributors to be adopted by others, at the KCC. This seems like a good candidate for such a thing. Ryan?

As for the wiki, should it be version independent, maybe tagged with what version was the latest when a given change was made?

  Noam

I made a wiki contribution, twice: _000, _001

I have a few concerns:
(1) Current policy/system: Is this the best policy for orphans with
indisputably inaccurate metadata? Can the original developers assume
ownership of their own models that were submitted by a third party?

This certainly highlights an area of policy that I struggle to find the best balance for. I can summarize with two questions: What constitutes "the data" and "the meta-data"? How should history an provenance be maintained for "the data" and "the meta-data"?

So, do we need to keep full history of the meta-data or just the most current version? Does the meta-data "come along" with the data when a KIM Item is downloaded? ...

Our scheme and infrastructure would have to be rather more complicated if we have to keep separate histories for the data and the meta-data. Ultimately, we need a comprehensive set of definitions and procedures for how to handle all situations. A case-by-case approach (or "handle it as it comes up") is, I think, asking for inconsistency and problems down the line.

I think we are very open to a discussion and new ideas for how to do all this. However, we'll want to converge on our best guess at the optimal solution before we make changes to our existing scheme....

I recall a discussion of a possible procedure for allowing entries with
unresponsive contributors to be adopted by others, at the KCC. This seems
like a good candidate for such a thing. Ryan?

This is actually the reverse case. Here the original creator is not available and a 3rd party has submitted the Model to KIM. I think we would be happy to work with all parties with a stake in any KIM Item to come to a solution on ownership/maintenance of items that gives credit and acknowledgment to the appropriate parties.