Search components

[fschiettecatte]

• Each search component needs to be specified (I think Orion Buske has some good information for that)

• There is plenty of good stuff to steal from @Buske's mockup of MME v2..

---

• Need to figure out sources vs. id and CURIEs (I think Chris Mungall was maintaining a list of CURIEs) ? Should we use CURIE's or some way to identify the resource type when using IDs? #29
• Operator needs to be defined, eg. EQ, NEQ, LT, LTE, GT, GTE. Which query boolean operator paradigm should be used? #9
• Version numbering as defined above does not resolve to the search components, see above? Version numbers resolve to components, but not the search items in the search components #11
• Do we want to require support for a minimal list of search components? Should we require default a required set of supported components? #15
• What is the process for adding search components? How do we specify new components? #26
• Do we allow custom search components? Should we allow for private / custom components? #17
• Do we want to support defined sets search components for certain communities (MME)? Do we want to support defined sets of components/records? #28

[Relequestual]

Hope you don't mind my continuing editing of your comments! tidy tidy

Can we get some expansion on "Need to figure out sources vs. id and CURIEs (I think Chris Mungall was maintaining a list of CURIEs) ?" please?

[fschiettecatte]

One of the things we discussed at length in MME 1.0 was to use CURIEs for identifiers (See issue 113 in MME 1.0).

But we have a source included in the examples. It is not clear what should go there and how this is different from a CURIE. Just want to make sure we think this though.

[Relequestual]

Rigth, this sounds sensible, like something we should bake in from the start. I'll open a new issue.