[postgis-users] Vacuum analyse ERROR: Operation on twoGEOMETRIESwith different SRIDs

[Markus Schaber]

Hi, Strk,

strk at refractions.net wrote:

>>The only think I think we could possibly do is reduce the SRID mismatch
>>check from elog(ERROR) to elog(WARNING), but I'm not sure I'd be too keen to
>>do that...
> 
> WARNING was not supported in postgresql 2.1.

Is it supported in 7.2 and newer?

If yes, we can make it a warning at least for a sensible subset of the
functions. If not, is it possible to use ERROR on the older PostgreSQL
servers, and WARNING on the newer ones?

Btw, I think changing it to WARNING does not obsolete the need for the
SRID=-1 exception, but does considerably lower its priority.

Markus