> I meant that Doctors - and Doctoresses ;-) - program many little tools
> (drawing curves, giving risk evaluation using Framingham equations, and so
> on). In France, a contest was recently organised to elect the best of it !
>
> But as these tools are autonomous, and then launched outside patient
record
> software, they need re-enterering lots of datas to work. The straitforward
> effect of that is that they remain gadgets.
>
> If people took time to developp it, it is because they needed that
function.
> That is the reason why I think a good medical record must be able to act
as
> a Middleware ; it is a genuine *plus* for open source approach.
I am a strong believer in small autonomous *pipeable* tools - that is what
makes UNIXes rock. With XML etc. we have the possibility of formalizing a
piping protocol. That way, all the language discussion is futile, software
becomes more robust (it is much easier to test & debug small tools than
large monolithic applications).
However, this is difficult to achieve in a GUI world.
Horst
