Citations for data processed by openEO

[m-mohr]

It would be good to have a possibility to cite (any maybe reproduce) results processes with openEO.

@bgoesswe will be working on this for his thesis, as far as I know. It will probably need integration in the API, that's what this issue is for. ;)

[bgoesswe]

I created a summary of what I want to do with my master thesis in my own github repository.
For the openEO api, I have following suggestions (which are also in more detail in the summary):

• Adding an endpoint to get the back end specific environment that is interesting for the user.
  • This can result in a backend version number, that gets updated on every change of the backend
  • Endpoint Suggestion: add it to /capabilities, or add /version endpoint
• Adding additional information about the job execution environment to the GET /jobs/ endpoint
  • Only for users interesting information about the concrete job execution environment
  • Only information that is not a security risk for he backend provider.
• Adding two endpoints for job comparison
  • One endpoint for a comparison of two different runs on the same job (same job id)
  • One endpoint for a comparison of two (or more) different jobs
  • Endpoint Suggestion: POST /jobs//diff/ with the other job ids in the body.
  • How the comparison result can look like is also in the summary.

I am happy for any feedback on the suggestions and on what you think is interesting for typical openEO users.

[m-mohr]

@bgoesswe Your first bullet point was implemented in #154.
All the other proposals seem well suited in a "reproducibility" extension or so, see the plans in #120.

[bgoesswe]

Thanks , there are some additional things that came up and would be interesting for reproducibility reasons:

• In the capabilities endpoint there could also be, if existing, the url of the Github repository of the Back end driver, it would be even better to add also the currently checked out commit. So a possibility would be to add it to the version field in addition to the back end version in some kind of dictionary.

• If the back end has implemented data citation (like following this recommendations), the citation/identifier of the data can be added to the metadata of the job at /jobs/{job_id} and also at the data description /collections/{name}.

I am working on data citation on the EODC back end, so it would be nice to be able to use it productively.

[m-mohr]

@bgoesswe If you want the new endpoints and fields to be added anytime soon, I suggest that you fork the API repository and work on the 0.4.0 branch to add your endpoints and fields to the OpenAPI document. Based on that we can discuss. Afterwards, we can integrate it or make a extension.