Open BenPortner opened 4 years ago
my opinion:
we probably want our own database. it should be possible to save sources like "exiobase manipulated using mrio_common_metadata". such things are not in any database i know of.
i'm strongly in favor of using existing tools. it saves us a lot of time. i haven't looked into zotero yet but it probably provides nice import / output functions that allow conversion of bib/ris files
The fields in datasource
are pretty parsimonious. Not sure we could cut more.
@cmutel
Adding a field for bibtex/etc. is reasonable
My idea was to change the schema such that datasource
itself has all the fields that a typical bibtex/etc. has. Not sure I understand your suggestion right. You want an extra field where the bibtex/etc. is stored as a string?
Having a data source column is essential, we need to track this, even if most metadata are stored elsewhere. We still also must track data licensing.
I agree! But what should we store in the datasource
table and what should we store "elsewhere"?
the datasources table is a list of sources. our audience is scientists so these sources will be mostly scientific. this raises two questions:
1) do we really want to implement our own database of scientific sources? or is it maybe smarter to use existing databases like google scholar? in the latter case, it would be enough to store google scholar ids and/or links in the table
2) if we want to have our own database: do we really want to build our own tools for managing this database? or can we use existing, open-source tools like e.g. zotero? if yes, then we would probably have to adapt the scheme accordingly (e.g. datasource would have to allow all fields which a bib/ris file can have)