[GRASS5] Re: g.region -l

Markus Neteler wrote:

BTW: Maybe we should modify '-l' of g.region to work on WGS84 (like d.where)?
Or just advertise that the current datum is used? I am not sure what's
preferred.

Well if you don't know I don't think I'm very likely to know either!
Presumably the feature is there for a reason but as I don't know what
people use it for I wouldn't want to be the one to change it. Better to
keep things consistent unless there's a good reason for changing (there
was for d.where; it used the old coorcnv library and part of the
functionality worked only in UTM locations).

Paul

On Thu, Mar 04, 2004 at 05:10:33PM +0000, Paul Kelly wrote:

Markus Neteler wrote:
>
> BTW: Maybe we should modify '-l' of g.region to work on WGS84 (like d.where)?
> Or just advertise that the current datum is used? I am not sure what's
> preferred.

Well if you don't know I don't think I'm very likely to know either!
Presumably the feature is there for a reason but as I don't know what
people use it for I wouldn't want to be the one to change it.

I think that I implemented it to support myself in downloading satellite
data from the NASA data repository. They always want LatLong for project
areas. So I added -l, and g.region tells me immediately the current
coordinates in LatLong. Copy-paste and I'm done.

Better to
keep things consistent unless there's a good reason for changing (there
was for d.where; it used the old coorcnv library and part of the
functionality worked only in UTM locations).

So that's a vote to change g.region. I tend to agree, because a
reliable WGS84 output is better than a sort of undefined LatLong
output with current datum/ellipsoid.
Other opinions?

Markus

> > BTW: Maybe we should modify '-l' of g.region to work on WGS84
> > (like d.where)? Or just advertise that the current datum is used?
> > I am not sure what's preferred.

[...]

> Better to keep things consistent unless there's a good reason for
> changing

Yes.

> (there was for d.where; it used the old coorcnv library and part of
> the functionality worked only in UTM locations).

So that's a vote to change g.region. I tend to agree, because a
reliable WGS84 output is better than a sort of undefined LatLong
output with current datum/ellipsoid.
Other opinions?

d.where has

Flags:
  -d Output lat/long in decimal degree
  -l Output lat/long referenced to current ellipsoid
  -w Output lat/long referenced to WGS84 ellipsoid using datum
        transformation parameters defined in current location if available
[
  note descriptions aren't very clear and could probably be improved..
    -d is independent of datum/ellps choice;
    -l should change to ellps/datum;
    -w wording is awkward

   the program will prompt you if you try and mix them improperly though.
]

g.region should match, I think all that means is adding a -w flag to
get a WGS84 version of -l. (I think the default action should be to use
the current ellps/datum)

?
Hamish

ps -
GRASS53> d.where -w
[click]

                                                   WGS84 Co-ordinates
             EAST: NORTH: LON: LAT:
pj_transform() failed
cause: failed to load NAD27-83 correction file
ERROR: Error in pj_do_proj()

GRASS53> g.projinfo
----...
PROJ_INFO file:
name: New Zealand Map Grid
datum: nzgd49
nadgrids: nzgd2kgrid0005.gsb
proj: nzmg
ellps: international
a: 6378388.0000000000
es: 0.0067226700
f: 297.0000000000
lat_0: -41.0000000000
lon_0: 173.0000000000
x_0: 2510000.0000000000
y_0: 6023150.0000000000
----...

I thought that used to work.. hmph. same old NTv2 PROJ bug?
If I change back to the 7-param transform it works.