Relationship between _publ_author.id and _audit_author.id

[Horst Puschmann]

Hello,

I think that this is an important question and I am with James on this one:
I would go with number (4) -- these lists are independent (*).
Conceptually, I don't think that (3) is correct.

In fact, I would go a step further and say that the _publ_author loop
should not be permitted in a data block.

At the point of structure origination, nothing is known (for certain) about
if and where the work might get published.

Assembling the _journal_author list is the very last step and can only be
done after a paper has been accepted for publication, at which point
nothing in the data block(s) should be touched.

Horst

(*)  There are many cases where *none* of the audit_authors end up on the
papers!




On Fri, 7 Aug 2020 at 03:28, James Hester via coreDMG <coredmg at iucr.org>
wrote:

> Dear Core DMG,
>
> Related to my previous email today on author roles is the question of the
> relationship between "publ_author" and "audit_author".  Recall that
> publ_author lists authors on a publication that is based on the CIF file,
> and is used by the journals. audit_author lists authors associated with the
> production of the data. There are 4 possibilities that I see:
>
> (1) Audit author is the root source of author identifiers: all authors in
> publ_author also appear in audit_author
>
> Objection: some authors on the publication may be associated with work
> that is outside the crystallographic experiment and so not properly
> included in audit_author
>
> (2) Publ_author is the root source of author identifiers: all authors in
> audit_author also appear in publ_author
>
> Objection: Some contributions recorded in audit_author may not rise to the
> levels required for ethical inclusion in an author list
>
> (3) A third "author" list is created that becomes the canonical author
> list, from which authors in audit_author and publ_author are drawn
>
> Objection: duplication of information
>
> (4) audit_author and publ_author are formally independent lists, but can
> link authors if required either via a new dataname pointing into the other
> list, or indirectly via e.g. OrCID identifiers.
>
> My preference is for (4) as it recognises both the different underlying
> scopes of the two author categories but also allows common authors to be
> identified, and is minimally disruptive for existing software.
>
> If we had our time again, (3) would be the correct approach and we would
> not duplicate address information in publ_author and audit_author. However,
> making that change now would be very disruptive.
>
> Please comment.
>
> James.
> --
> T +61 (02) 9717 9907
> F +61 (02) 9717 3145
> M +61 (04) 0249 4148
> _______________________________________________
> coreDMG mailing list
> coreDMG at iucr.org
> http://mailman.iucr.org/cgi-bin/mailman/listinfo/coredmg
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.iucr.org/pipermail/coredmg/attachments/20200807/cfb46e23/attachment.html>