[postgis-devel] Fixing floating point precision issues under gcc

[Nicklas Avén]

wow, great!

Is that the issue that have forced us to round most of the regression
tests? I think x86 has been the problem most of the times.

What about this ticket:
http://trac.osgeo.org/postgis/ticket/503

It would be very nice if could get rid of that one I think.


Regards
Nicklas



On Thu, 2011-12-15 at 10:11 +0000, Mark Cave-Ayland wrote:
> Hi all,
> 
> For those of you that haven't been following the thread on -hackers: 
> http://archives.postgresql.org/pgsql-hackers/2011-12/msg00627.php.
> 
> Basically in order to remove floating point precision errors on x86 
> platforms, we need to compile with -fexcess-precision=standard. This 
> would probably be a good idea for GEOS too.
> 
> 
> ATB,
> 
> Mark.
>