[GRASS5] Re: s.in.garmin

[Hamish]

[cc the grass5 list for the status report]

> The only (?) way to go seems to be 
>  http://gpsbabel.sourceforge.net/
> because it supports tons of formats. And or contacting
> Hamish (I cc to both).
...
> > We're still trying to work out (Daniele is busy with this) an easy
> > way of uploading GPS data into grass57. Among the several problems
> > we found, the s.in.garmin program is (obviously) absent from
> > grass57. Would it be possible to reimplement it?


Right now s.in.garmin and v.in.garmin are written and tested for 
GRASS 5.3 (uses WGS84 datum support & m.proj2 transforms but the old
sites/vector ASCII formats). I never really finished updating
v.in.garmin in CVS; I have something here that works but never cleaned
it up. I have been meaning to do that.

Looking at 5.3's s.in.garmin, I don't see anything to stop it working
with 5.7, now that 5.7 has s.in.ascii. It should just work! Just copy
the script and try. Please report success or failure so it can be added
to 5.7 if it works or fixed to do so.

v.in.garmin needs a rewrite to work with 5.7, and probably a total
rewrite is best. It would be great to have attributes for each node
along the line with time, speed, direction, or whatever else the track
data provides. Basically it writes out an old vector-ascii file and
imports that. The new vector-ascii format isn't too different from the
old, its style makes doing real-time logging a pain though[*]. This is
why I asked about changing it a while back[**].

[*] both having to know in advance how many points make up a line and
having to write the dig_att/, dig_cats/ at the same time while listing
categories on different lines after all points are recorded is a real
pain. Maybe write data to a .dbf file and import that at the end? I'm
not sure, haven't really spent much time (zero) on it and don't plan to 
any time soon:
see "Vector ascii Format Specification" in grass51/doc/vector/vector.html
re. attaching categories to nodes, especially [ FIELD CATEGORY ] after
all the y,x[,z] (!#$%^%$@ !!) data points.


[**] http://article.gmane.org/gmane.comp.gis.grass.devel/3102

maybe there should be 'v.in.nmea' which logs output from gpsd[***]?
 (my m.realtime_gps pretty much already does this)


[***] what GPSdrive uses,  http://www.pygps.org/gpsd/gpsd.html


> > Another (more complex??) possibility would be to use GPSMan (a good,
> > lively project) instead of the (dead?) one present in grass5.

grass5 uses 'gpstrans' which works just fine for Garmins, and is alive
and well as far as I can tell. GPSman is a GUI monstrosity which isn't 
any good for automated use by grass.. I never liked it much anyway;
*horrible* UI.


> > I have seen somebody has worked on this subject, for the 5.3 branch:
> > http://op.gfz-potsdam.de/GRASS-List/Archive/msg10661.html
[details m.realtime_gps]


I'll have to look at gpsbabel.


good luck,
Hamish