[GRASSLIST:828] Re: What is the distance tolerance for v.what.vect?

[Hamish]

> I'm trying to upload the navigation timestamps of the start and end
> points of a sampling transect to a series of vector points I created
> with v.to.points (see attached image). As you can see, my interpolated
> points from v.to.points (red crosses) only overlap on the start and
> end of line points (blue circles). I can't seem to get the timestamp
> attributes uploaded from the circles to the crosses, even though they
> clearly overlay. How exact do the coordinates from the two vectors
> have to be to be considered coincident? 

I am not 100% clear on the commands you used, but..
'v.to.points -v' to get end nodes as points, then upload values to them?

line nodes can not have attributes, only lines. e.g. in the v.in.garmin
script for each track it labels the line with a start_time, end_time,
start_lat, start_lon, end_lat, end_lon. (lat/lon so you can figure
direction). 

> The manual page makes mention of v.distance being used, but doesn't
> mention the tolerances that must be satisfied for successful attribute
> upload.

v.distance is correct. "upload=" looks for the nearest feature.

Many vector modules require that features have category numbers to work.
(v.category op=add)
?

I take it you know start/end times and will interpolate intermediate
postions and times?

see also Radim's v.lrs.* modules, new in CVS.


Hamish