On Wed, Dec 10, 2008 at 5:07 AM, Sutripta <sutripta@...> wrote:
> Feels great when the chief babelhead himself replies!
Recent travels (and more to come) have made it hard to keep up with the
list, but I try.
> > GPX doesn't actually say that trackpoints require time or that time
> > forward.
> Agreed. But all GPX handling programs (or at least the ones I have seen)
> expect the points to be ordered in space, and would go bonkers if we fed it
> a randomised list. We cannot validate the data w.r.t. spatial location, but
> we can validate w.r.t. time. Since the data is logged in
It may be XML hair-splitting, but you can't. The XSD for GPX doesn't say
that time has to march forward.
It's the difference between "valid" and "sensible".
> Actually, Ive got a serious question to ask:-
> What would your reaction be if someone (me) took your code, stripped away
> everything not related to say gpx, or kml, and extended it for those two
Within the bounds of the GPL, my reaction doesn't actually matter.
The spirit of the project, though, is that if you have enhancements to make
that may be useful to others, sending to the project for consideration for
inclusion is good.
A filter, for example, that did numerical analysis on tracks to smooth them
and remove "zingers" has been discussed from time to time. Our track filter
already has some code to identify non-squenced time events; perhaps that
code is an interesting starting place for you.
P.S. It's a mailing list. Change the subject at any time, but leave the
message ID in place and you're fine.