On Wednesday 28 January 2009, Jon S. Berndt wrote: > > From: LeeE [mailto:l...@spatial.plus.com] > > > > > > How difficult would it be to make available multiple versions > > of FDM, which could then be specified in the aircraft > > configuration? The appropriate FDM (UIUC/JSBSim/YASim/Balloon) > > has to be loaded after the aircraft config has been read, so > > why can't an appropriate version of the FDM be loaded as well? > > > > Personally, I'd like to see this sort thing thing applied to > > all the FG subsystems that are configurable in the aircraft to > > provide backwards compatibility and avoid the need for mass > > updates of the aircraft every time something changes. > > > > LeeE > > I think this would be a nightmare of epic proportions. > > And, actually, the aircraft config is read from the -set file, > no? The right FDM has to be called for the aircraft configuration > file to be loaded. They're all resident, anyhow. > > Jon
Well, ok, I haven't done any serious coding for twenty years, and C++ isn't one of the languages I've worked with, apart from tracking down a few bugs in FGFS, but I can't see why it would be "a nightmare of epic proportions". Is C++ intrinsically incapable of including multiple modules of similar but slightly different functions? The issue regarding all of the FDMs being loaded anyway seems more like an architecture problem than an intrinsic problem inherent in the coding language used, and even if C++ can only produce monolithic s/w where everything must be loaded (which I would find very surprising), even if they are not used, what is the problem with allowing unused modules to be swapped out? Ok, Windows swap handling is, quite frankly, atrocious, but if the swapped modules aren't referenced why should this lead to an overhead while the sim is running? Atm, there are modules for running UIUC, JSBSim & Yasim; what would be the problem with including, for example, JSBSimVx, JSBSimVy & JSBSimVz? Sure, there will be more lines of source to be compiled, but is compile time really an issue? By definition, old versions of the modules would not need maintaining; the whole idea being that they didn't change, and after all, recompiling is something that should only need doing infrequently for users. This really does seem more like an architecture problem, rather than a logic problem, to me. LeeE ------------------------------------------------------------------------------ This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
