[mapserver-users] mapscript queryByPoint - tolerance depends ofresolution ?

[Steve Lime]

Tolerance computations are based on cellsize. Resolution affects only scale computations, so it
is not clear to me how one could affect the other. I just went through the code to verify. Is there
any chance that layer scale limits are coming into play here and you're not querying the layer you
think you are? Does the behavior persist in 5.x?

If you can create a concise test case I'll gladly take a look.

Steve

>>> On 7/9/2008 at 7:41 AM, in message <2A7451FF53BE4CF0BC85B033346C7A94 at maunakea>,
"Oliver Christen" <oliver.christen at camptocamp.com> wrote:
> Hi
> 
> I have a problem when doing a queryByPoint with phpmapscript 4.10.1
> 
> By default, the resolution of my map is 96 dpi.
> I do a queryByPoint on a layer containing polygonal areas. The point is 
> right in the middle of one area.
> I use a default tolerance (-1) corresponding, if I read correctly the 
> documentation, to 3 pixels.
> 
> The result obtained is the area containing the coordinates I used, it's 
> absolutly normal and I get what I expect.
> 
> Now, if I dynamicaly change the resolution of the map to 150 dpi and then do 
> 
> exactly the same queryByPoint as before,
> I get nothing!
> Mapserver consider there was nothing at the coordinate I provided.
> 
> Then, if I redo the query, at 150 dpi, BUT with a tolerance set to 10 
> (instead of -1), I get a correct and expected result.
> 
> So, somehow, Mapserver seems to make some very weird calculation with the 
> tolerance and resolution which make it to believe there is nothing when then 
> 
> is something.
> 
> have you ever heard of such behaviour?
> is that a bug?
> 
> any help would be very welcome.
> 
> thanks in advance
> Oliver 
> 
> _______________________________________________
> mapserver-users mailing list
> mapserver-users at lists.osgeo.org 
> http://lists.osgeo.org/mailman/listinfo/mapserver-users