Re: [Gnumed-devel] Data packs

[Jim Busser]

On 2011-08-31, at 3:26 PM, Karsten Hilbert wrote:

> What is the actual question ?

Looking at the following,

> STIKO           | Juli 2002     | Ständige Impfkommission, Robert-Koch-Ins
> ATC/DDD-GM-2004 | 2004          | ATC/DDD-GM; Systematik\r; Version 2004\r; 
> Aust NIPS       | 2005          | Australian National Immunization Program
> SFCVC           | December 2004 | Starfleet Central Vaccination Council
> ATC             | 2009-01-EN    | Anatomical Therapeutic Chemical Classifi
> ATC             | 2009-01-DE    | Anatomical Therapeutic Chemical Classifi
> ATC             | 2009-01-FR    | Anatomical Therapeutic Chemical Classifi
> FreeDiams       | 0.5.4         | "FreeDiams" Drug Database Frontend
> LOINC           | 2.26          | LOINC® (Logical Observation Identifiers 

it is not clear to me that any should be "mandatory".

It would be better that bootstrappers not be inserting content which is not 
required and whereas one of the sources above (FreeDiams) was a result of my 
using it as a prescribing tool and maybe one or more of the others are a result 
of my running the v14->v15 fixup (which loaded vbanait's drug list) I suspect 
some of the others formed part of a default bootstrap.

Can we establish whether there is any reason (medically) to require that any of 
the above be loaded into GNUmed by default?

Where they are unsuccessfully argued as 'mandatory' I would suggest that the 
bootstrapper for new databases (which begins, I think, at v2) be refactored to 
leave out the sql which is not mandatory and which can simply be left in 
[...]/gnumed/data

??

-- Jim