Re: [Gnumed-devel] Forms redux, was: Re: Keeping gnumed focussed toward 0.1

[Richard Terry]

Sorry to come in late on this debate, can someone explain simply what it is.

If you are simply talking about a method of producing a common form which 
works across everything (pathology/radiology/physio/cardiovascular/vascular 
bal bla then that is very simple and I did it years ago and use it daily, but 
I suspect that this is not what you are talking about.

Regards

Richard

On Fri, 18 Feb 2005 08:16 am, Ian Haywood wrote:
> Karsten Hilbert wrote:
> > - it is reinventing an object database inside a RDBMS
>
> Indeed, isn't that what the whole business object layer is about? ;-)
>
> > - form definition and form data is not independant of form code
> >   - this is my biggest grief
>
> I suppose I can't see how you can ever get that independence, but (as
> always) I'm happy to be corrected.
>
> > - you are proposing to mix middleware and UI code
>
> At some point UI code needs to exist, and needs to talk to
> the laTeX engine, I don't see inheriting classes, with defined interfaces
> is really any different.
>
> > Personally, I do think with some effort form defs *can* be
> > made quite passive.
>
> I agree this is the ideal. Please come up with a proposal, I've racked my
> brain, I'm unable to come up with something that will print a script or a
> complex referral letter (including lists of meds and scripts), without
> embedding Python code, or extending LaTeX in fairly radical and odd ways.
> Also, we want to produce electroic HL7 forms, passive templates don't make
> much sense there. The only way I can see is to use super-high-level
> placeholders, like "list of meds", "list of diseases", however these will
> be both locale- and engine- (because they are LaTeX tables) specific.
>
> Ian
>
> P.S. to me, a reasonable demostrable version (which isn't exactly the same
> as 0.1) include printing scripts and referral letters. Hopefully once we've
> got this cracked it won't be too hard.