[postgis-devel] 1.3.4rc1

[Mark Cave-Ayland]

Courtin Olivier wrote:

>> So this only leaves the question of would changing the meaning of the 
>> function in this way cause
>> you great problems? Given that most data should be within a similar 
>> area, I think specifying a
>> slightly different fixed precision should give you a similar result as 
>> you have now.
> 
> Problem mainly occurs when i begin to reproject with PostGIS, for example:
> 
> SELECT st_asgeojson(transform(geometryfromtext('POINT(1000 2300000)', 
> 27582), 4326), 5);
>  {"type":"Point","coordinates":[-5.61668,47.42663]}
> 
> In this case, i 'know' that output will be in Decimal Degrees, so 5 
> decimal digits enough.

Right, I think I see what you're trying to do. I would personally prefer 
using significant figures, but I suspect that the GeoJSON code may have 
a slightly different use case for use over the web.

I think the best way to get around this for 1.3.4 will be to re-write 
the regression tests so that each one specifies a precision that 
restricts the floating point numbers to a maximum of 15 significant 
digits for each individual test. I'm fairly tight on time at the moment, 
so would you be able to post a patch with the altered regression tests 
along with a documentation section detailing the change of precision 
from significant figures to decimal places? I will then try this on 
Win32 later this evening.


ATB,

Mark.

-- 
Mark Cave-Ayland
Sirius Corporation - The Open Source Experts
http://www.siriusit.co.uk
T: +44 870 608 0063