RE: [udig-devel] Re: CRS - "Bursa wolf parameters required."

["Dean C. Mikkelsen" <dcmikkelsen@xxxxxxx>]

Hi Paul,

That is common error, not enough information is provided in .prj file.

Interesting to see it has been bantered around before on maptools.

The EPSG database though is not a catch-all either - that I know - there
are datum's missing and map projections missing. I consider it a
starting point and hope that maybe it pushes the importance of datum's,
transforms and locations to the GIS community. Hopefully it makes people
ask questions about their data and quality. After all we don't want the
GIGO philosophy.

Unfortunately it comes down to guessing sometimes or sending the data
back to the provider and asking for the missing information.

C'est la vie.

Cheers,
Dean


-----Original Message-----
From: udig-devel-bounces@xxxxxxxxxxxxxxxxxxxxx
[mailto:udig-devel-bounces@xxxxxxxxxxxxxxxxxxxxx] On Behalf Of Paul
Ramsey
Sent: April 20, 2005 4:53 PM
To: User-friendly Desktop Internet GIS
Subject: Re: [udig-devel] Re: CRS - "Bursa wolf parameters required."

Dean,
What you describe is essentially the "use the EPSG database" solution 
which has been bandied about geotools for over a year now. It is more of

an implementation effort thing: noone has gone and done it yet. The 
expedient solutions have been "good enough".

That said, the EPSG database alone will not actually solve the problem 
that gave rise to this thread. Jesse was instantiating a projection from

a .prj file associated with a shape file, and that .prj file did not 
include datum transformation information. In order to fix the problem, 
at CRS instantiation time the datum name needs to be matched against a 
"known" datum somehow, so the transformation information can be added 
post facto.

Paul

Dean C. Mikkelsen wrote:

> Just an idea...I've been following this Bursa-wolf thread since the
> start.
> 
> The EPSG has implemented the code into a SQL for MySQL. With some
> modifications to the SQL, could it not be implemented into PostGIS and
> become part of the data model. Hence the codes could be looked up in
the
> DB, rather than maintaining an external file. 
> 
> I'm based in North Vancouver, a geodesist by training, and was
involved
> in the APSG and EPSG over my past 11 years. I was in oil and gas
> exploration data management and GIS.
> 
> I am relatively new to uDig and MapServer, but learning the ropes
pretty
> quickly.
> 
> I may be totally off the boat here, but then again, it's just an idea.
> 
> Kind regards,
> Dean
> 
> Dean C. Mikkelsen, B.Sc., P.Eng.
> Terra ETL Ltd.
> Phone: 604-929-8630
> E-mail: dcmikkelsen@xxxxxxx
> 
> 
> 
> -----Original Message-----
> From: udig-devel-bounces@xxxxxxxxxxxxxxxxxxxxx
> [mailto:udig-devel-bounces@xxxxxxxxxxxxxxxxxxxxx] On Behalf Of Paul
> Ramsey
> Sent: April 20, 2005 3:37 PM
> To: User-friendly Desktop Internet GIS
> Subject: Re: [udig-devel] Re: CRS - "Bursa wolf parameters required."
> 
> Direct GeoTools access to the EPSG database is a "great idea" that
noone
> 
> has actually implemented. It was once implemented, but is not working 
> now. The only "epsg" data we have now is a WKT properties files I
hacked
> 
> up many moons ago, precisely because the EPSG database was not working
/
> 
> ready.
> 
> Yes, if we had an ESPG database, the DatumAliases could also reference

> an EPSG datum ID, which in turn would allow us to look up transforms 
> between those actual datums, which would be very nice. But we would 
> still need to do the DatumAliases trick.
> 
> P
> 
> Matthias Basler wrote:
> 
> 
>>Paul Ramsey wrote
>>
>>
>>>One thing we can do (prepare for the pain) is to add yet more
>>>information to our DatumAliases -- for each DatumAlias, add the BW
>>>transform parameters to WGS84. That way, if we come across a
> 
> particular
> 
>>>datum name, but lack TOWGS[] information, we can fill it in
ourselves.
>>>Otherwise it'll be death and destruction quite frequently: most apps
>>>generating WKT do *not* generate the TOWGS[] portions of the WKT.
>>
>>
>>Maybe this is a very stupid question: Doesn't support GeoTools the use
> 
> of the
> 
>>EPSG database for getting the transformation parameters? AFAIK most
> 
> CRS are
> 
>>exactly defined in the EPSG database (including Bursa wolf parameters
> 
> and other
> 
>>means of reprojection), so all missing information can be retrieved
> 
> from it,
> 
>>cannot it?
>>
>>Since I'm interested in the reprojection functionality myself, maybe
> 
> someone can
> 
>>shortly outline to me how this currently handled in uDig/GeoTools2.1.
>>
>>Matthias Basler
>>c9bama@xxxxxxxxxxx
>>
>>----------------------------------------------------------------
>>This mail was sent through http://webmail.uni-jena.de
>>_______________________________________________
>>User-friendly Desktop Internet GIS (uDig)
>>http://udig.refractions.net
>>http://lists.refractions.net/mailman/listinfo/udig-devel
> 
> 
> _______________________________________________
> User-friendly Desktop Internet GIS (uDig)
> http://udig.refractions.net
> http://lists.refractions.net/mailman/listinfo/udig-devel
> 
> _______________________________________________
> User-friendly Desktop Internet GIS (uDig)
> http://udig.refractions.net
> http://lists.refractions.net/mailman/listinfo/udig-devel

_______________________________________________
User-friendly Desktop Internet GIS (uDig)
http://udig.refractions.net
http://lists.refractions.net/mailman/listinfo/udig-devel