Christian Mayer writes: > > - is readable? (already implemented) > > - is writable? (already implemented) > > - is archivable? (already implemented) > > - data type (already implemented) > > - list of allowed values > > Wouldn't a check for that slow the properties down? Or would it only be > used when you enter data externally?
It depends on the constraint. We'd want to design it to minimize overhead where a constraint is not in use (i.e. a single boolean or null-pointer check); where a constraint is necessary, we probably currently enforce it in C++-code somewhere, so it won't make much of a difference. Note that four of the five above are already implemented. > An other idea could be to add a description to every property. So the > user can figure out what each of those are good for when browsing the > tree. And when we would change our command line options to be property > entries we cuold automatically create the --help option which will > allways stay up to date. I think that's a great idea. There are two reasonable choices: 1. Allow a "description" attribute for each property in the XML markup. It could contain only plain text in a single language, but would be trivially easy to access. 2. Allow a "url" attribute for each property in the XML, pointing to a Web page (or part of a Web page) describing the property. The HTTP server could handle L10N through content negotiation, and we could make the documentation as elaborate as we want; however, it would be hard to stick the docs into the property picker. I'm going to recommend #1, even though it supports only English documentation. We can always add something more elaborate later, but a navel-gazing, chauvenistic, English-only system that everyone uses is better than an idealized, multilingual system that no one does. If anyone complains, I'll consider doing the whole thing in 7th-century Insular Latin instead. All the best, David -- David Megginson [EMAIL PROTECTED] _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
