Re: [Gusdev-gusdev] SRes.ReviewStatus

[Jonathan C. <cra...@pc...>]

Joan-

Joan Mazzarelli wrote:
> Hi Jonathan,
> 
> I think we need to discuss this a bit more in terms of how this will impact other code we have
> which recognizes this assignment, and it is not clear what type of entry has this set in GUS30
> now (1 manually created Entry was created by hand; review is not needed.) for it to be taken.
> 

OK, although I'd like to get something settled soon, because I'm on a very
tight schedule.  Can you be more specific about the adverse impact you think
this change will have on our existing code?  In particular, why does having
5 terms in the controlled vocabulary make things any more difficult than
having 3?  It seems that in either case code that previously relied on a
single bit (manually_reviewed) will now have to query a controlled vocabulary
instead.

Your second point is a good one, which is that we may not have an easy way to
determine which entries currently in gusdev should be assigned the ReviewStatus
"manually created."  I'm certainly open to dropping this term if people don't
think it will be useful.  Jonathan, can you tell us how you're using this term
now?  In the short term (i.e. later today), however, I would propose the
following mapping (for entries currently in gusdev with a non-null
manually_reviewed column):

manually_reviewed    ->        ReviewStatus
0                              unreviewed
1                              reviewed, correct

I believe this correctly represents our existing semantics, *except* for entries
that have been manually created (which are simply marked as manually_reviewed = 1).
Doing this we don't lose any information and could go back later and attempt to
identify entries that had been manually created.  Or we could later decide that
we want to drop the "manually created" term completely, and could do so without
affecting any of the migrated gusdev data.

Jonathan