v.in.garmin - detect gpstrans or gardump instead of user flag

[kyngchaos]

v.in.garmin uses a flag for the user to specify whther to use gardump as an alternate to the default gpstrans to get data from the GPS. This requires the user to know which is installed in their system, or bundled with a binary distribution of GRASS. This can also confuse the user if gardump is installed (whether they realize it or not) and they don't use the flag, then v.in.garmin tells them that gpstrans is not installed. Something like:

if [ "`type -t gpstrans`" ] ; then
	USE_GPSTRANS=1
else
    if [ "`type -t gardump`" ] ; then
        USE_GARDUMP=1
    else
	    g.message -e "either gpstrans or gardump is required, install one first\
	    http://gpstrans.sourceforge.net\
	    http://www.snafu.org"
	exit 1
    fi
fi

Then test for USE_GPSTRANS or USE_GARDUMP later in the script for which to use.

[hamish]

On reflection I think the flag is better. Of course we can improve the error message to suggest using the flag if the download program isn't found.

The two programs (and gpsbabel too for that matter) are not direct equivalents. Each one has its own strengths WRT attribute handling on different garmin models (usually WRT waypoint metadata fields).

I have both gardump and gpstrans installed, so hardcoding the script to run the first one it finds removes my ability to choose. Adding some sort of environment variable to override that hardcoding is a regression from the current situation IMO.

I don't consider it such a bad thing that the user know what 3rd party programs they have installed and are running.

Hamish

[hamish]

closing bug as "won't fix" for reasons stated earlier in the bug report.

Hamish