[postgis-devel] Issue 89 in postgis: transform() grid-shift 2nd chance logic defective

[codesite-noreply at google.com]

Comment #4 on issue 89 by william.... at acm.org: transform() grid-shift 2nd  
chance logic defective
http://code.google.com/p/postgis/issues/detail?id=89

Mark,

Yep, I've had exactly the problem you cite: the Ubuntu and Debian PostGIS  
packages I
was using lacked the grid-shift files and I was scratching my head  
wondering why my
NAD83->NAD27 transformation did nothing.  Had to build my own PostGIS to  
fix it --
and that was long before I patched PostGIS for this issue.  However, I  
believe my
patch provides adequate safeguards for this problem.  I believe the other  
reasons I
give for the behavior I advocate are of equal or greater importance.  To  
briefly
summarize them:

* My patch throws a warning, so the user knows something is wrong unless  
they have
actively suppressed warnings.  (Improving that warning by identifying the  
offending
geometry would be an important improvement to my patch.  I'll try to do it  
if the
team doesn't have the resources to do so, though your C skills are surely  
far
stronger than mine.)

* In all other cases (AFAIK), PostGIS simply "wraps" proj.4 behavior and a
knowledgeable user expects to see that behavior propagated without change.   
PostGIS
does not abort transformations outside the defined grid-shift region; its  
cs2cs
utility glides past this problem without even a warning.

* I'm just a newbie to cartographic grid shifts, but from reading the proj.4
documentation, I don't think that NAD27 is unusual in being defined over a  
limited
region.  Anyone who's using grid shifts should know the limitations of what  
they're
doing.  Thus, if they present geometries for transformation outside the  
defined
region, they presumably wish to obtain the transformation in all cases,  
even if it
cannot be grid-shifted.  I believe this behavior manifests the "principle  
of least
surprise."

Whatever y'all decide, thanks so much for what you're doing!

-- Bill

--
You received this message because you are listed in the owner
or CC fields of this issue, or because you starred this issue.
You may adjust your issue notification preferences at:
http://code.google.com/hosting/settings