Re: contacts, was: Re: [Gnumed-devel] DrugBrowser nicer piccie

[Richard Terry]

On Wednesday 01 February 2006 18:56, Karsten Hilbert wrote:
> On Mon, Jan 30, 2006 at 11:20:28AM +1100, Richard wrote:
> > I've switched at the moment to working on a contacts database I"ve done
> > all my fields for both contacts (general - including organisations) and
> > patients demographics, including (Sigh!!!!, after much agonising, full
> > normalisation of communication data - phones/fax/carrier pigeons, etc, so
> > that one can have multiple home phones/mobiles/email addresses etc,
> > without limiting the display.
>
> I wonder what all the agony was for considering we already
> have a demographics schema featuring

You don't 'get it', Karsten. I am trying to learn. My learning method is to do 
something - on my own, from scratch. That way I get to find out what I don't 
know, and what I have to learn. We All learn differently. I am not as young, 
computer coversant and clever as you are. I get totally confused with gnumed/ 
your nomenclature etc. When I'm up to speed I'm sure I will understand it, 
and then hopefully be able to contribute, but until I get something working 
myself, see how postgres works/how it links to python, I can't possibly be of 
use to anyone.

I take in what you say below. Once I have done my exercise I will revisit that 
stuff.


Thanks for your understanding and continued support and assistance to let me 
evolve an understanding in my own way, otherwise, I will give up again, as I 
have ever single time over the last 6 years, hence never got up to speed.

I get totally put off my what seems to be your attitude, of 'why are you 
bothering - we have already done that.....', so I've always got to the point 
where I just say "F*** it, I won't ever bother if that is the attitude, its 
like bashing my head against a brick wall'. The level of knowlege you assume 
often when I ask questions is way way way above my level of understanding.

Ian has been helping me with basic, concrete examples that I can get my head 
around. I'd love that you can field my questions as well.

I made this point 6 years ago on this list. If those of you who really do 
understand python programming, actually invested 6 months in tuition of those 
on the list who want to help, but can't, than you would have a number of 
programmers working in unison by now, and  the crappy code they produced 
could be gradually honed, or they could do small tasks which might be 
monotonous to you.

Regards

Richard



> - people with
> - any number of past and present names per person all
>   searchable concurrently
> - first, middle, last and additional name in each
> - countries -> regions -> places -> streets -> addresses
>   fully hierarchially normalized with a couple of views for
>   convenience
> - any number of IDs per person
> - any number of communications channels per person
> - any number of jobs per person
> - none of this restricting the display of any of it in any way
> - any of the "anys" independant of each other
>
> Two things to consider:
>
> 1) Yes, there may be room for improvement.
> 2) We don't currently support organizations too well.
>
> IOW: I would love to know where our current schema does not
> deliver (regarding people). Show me a use case that cannot
> be supported under it and you'll have the best argument in
> the world to want changes to the demographics schema in
> GNUmed.
>
> > I/ve imported my 6500 patients into postgresql and can pull names
>
> The above schema was tested with 120000 (One Hundred Twenty
> Thousand) patients and verified to deliver named patient
> searches in around 100-150 milliseconds.
>
> There is certainly no question that AU is entitled to have
> its very own demographics *widget* (possibly even
> middleware) optimized for local needs. The need for a
> different schema (apart from, perhaps rather large,
> additions regarding *contacts at organizations*) would need
> to be established unless a rewrite is what is *wanted*. And a
> rewrite better be capable of handling needs other than AU if
> it wants to claim to be GNUmed.
>
> Karsten