[GRASS5] Re: Testing GRASS PLUGIN...

[Paul Kelly]

Hello Radim/Luca
On Thu, 3 Nov 2005, Radim Blazek wrote:

> I am using  GPJ_osr_to_grass() and G_make_location().
>
> Paul, can it be that GRASS does not understand a projection created
> by GPJ_osr_to_grass()?

See comments at bottom.

> BTW, it seems that that GPJ_osr_to_grass() returns always 1 even if it fails
> so I test if pointer is set or 0.

Well technically it doesn't really fail---if it can't create a projected 
co-ordinate system, it creates an XY co-ordinate system instead. But this 
is still considered a valid co-ordinate system so it isn't a failure. 
Perhaps it could return a different positive value to indicate a different 
type of success though?

> Radim
>
> On 11/2/05, Luca Casagrande <luca.casagrande at gmail.com> wrote:
>> Hi to all!
>> I was testing the great plugin Radim has made, and i discovered that if i
>> create a NEW LOCATION from qgis and then print out the region info using
>> g.region -p inside GRASS, the datum is not recognized:
>>
>> GRASS 6.0.1 (provadatum):~ > g.region -p
>> projection: 3 (Latitude-Longitude)
>> zone:       0
>> datum:      ** unknown (default: WGS84) **
>> ellipsoid:  a=6378388 es=0.00672267
>> north:      90N
>> south:      90S
>> west:       180W
>> east:       180E
>>
>> The proj string used in qgis was: +proj=longlat +ellps=intl +pm=rome +no_defs
>> (Monte Mario Rome)
>> Am I missing something?

Well that proj string does not contain datum information. BUT... Radim says 
GPJ_osr_to_grass() is used and it takes a WKT string to describe the 
co-ordinate system, not a PROJ string. So is there any way you can show me 
the WKT string that was extracted from QGIS and passed to 
GPJ_osr_to_grass()? Probably there is something not right about the datum 
information there.

Paul