Vegard Ovesen wrote:

> Sent: 18 October 2004 09:26
> Subject: Re: [Flightgear-devel] Submodels
> On Monday 18 October 2004 00:24, Vivian Meazza wrote:
> >
> > The ability to set a serviceability state for each submodel system would
> > seem to be the correct approach, but if I understand your proposal
> > correctly, it will end up in more files overall.
> Actually the systems and instrumentation configuration files are already
> in
> CVS. So my point of view is that integrating the submodels configuration
> into
> the already existing systems configuration file would result in fewer
> files.
> It seems to me that your point of view is that adding systems and
> instrumentation configuration files would result in more config files.
> Which
> of course is true, but as I said systems and instrumentation config is
> already there (and the won't go away ;-)).
> >
> > As a major user and part-author of the submodel system, I have no
> > objections,
> A quick grep through the base package revealed that three aircraft use the
> subsystem: F16, Spitfire and Hunter. I will of course move the subsystem
> config to system config so that they don't get broken.
> > but David Culp was the originator: his view might differ. You
> > may wish to seek his approval before going ahead with this change.
> On Wednesday 08 September 2004 01:11, David Culp wrote:
> > > David Culp, is it ok if I modify the new submodel so that it can be
> > > configured in the systems.xml file along with the rest of the systems?
> Or
> > > do you have another plan for this?
> >
> > Sounds OK to me.
> >
> >
> > Dave
> I'm not that an experienced programmer, so I am wondering if the vector of
> submodels approach has superior preformace compared to my approach.

Having dwelt on this overnight, I think that there are advantages to the
vector approach. It is more or less consistent with the electrical system. I
briefly suggested separating the various sub-types of submodel during
development, but Erik advised against IIRC.

On the other hand, it would be nice to place the serviceability property
where it more properly belongs: with each sub-sub-model.

Unless there is a pressing need to change, or there are definite advantages,
I would suggest that we leave things as they are for now. If we all agree,
I'll look at moving the serviceability property once I have completed the
Seafire (nearly there).


Flightgear-devel mailing list

Reply via email to