[Proj] Numerical Precision in Proj, WKT, and everywhere

Glynn Clements glynn at gclements.plus.com
Fri Apr 2 17:08:21 EST 2010


Gerald I. Evenden wrote:

> The only hassle occurs when the the fractions are irrational such as pi, then 
> there is a problem. But in these situations (and probably others) the 
> continual bouncing back and forth between decimal and binary should be 
> avoided.

Such is the reality of processing data using a multitude of tools
connected by textual file formats. Most such file formats use decimal
representation, while most software uses binary (e.g. IEEE-754
floating point) internally.

Coupled with the fact that so few of the programmers who use floating
point have a good understanding of the issues, it's ultimately
necessary to retain some "headroom" in order to minimise the impact of
the inevitable flaws.

-- 
Glynn Clements <glynn at gclements.plus.com>


More information about the Proj mailing list