Mapserver Error Code: 33

[pw]

Frank Warmerdam wrote:
> On Wed, 16 Feb 2005 13:46:11 -0800, pw <p.willis at telus.net> wrote:
>
>>Hello,
>>
>>I get the following error from mapserver:
>>
>>msProcessProjection(): Projection library error. no system list, errno: 33
>>
>>What is error 33?
>
>
> Peter,
>
> Looking it up in /usr/include/asm/errno.h, it seems that error 33 is
> EDOM (Math argument out of domain of func).  Any of the errors above
> 0 are supposed to be system errno values.
>
>
>>The map I am trying to render is a lambert conformal conic
>>above 60 degrees latitude.
>>
>>Do proj4 or mapserver have problems if the map traverses the
>>pole in lambert C.C. or Lambert equal area?
>
>
> I am suspecting some sort of numerical problem related to the pole, or
> your parameters.
>
> Best regards,

Hello,

I tested things out a bit with proj4 to see if the proj parameters
that were going into the mapserver map file were causing problems for
coordinates close top the poles.

PROJECTION
  "proj=lcc"
  "ellps=WGS84"
  "lat_0=90"
  "lon_0=0"
  "units=m"
END

PROJECTION
  "proj=laea"
  "ellps=WGS84"
  "lat_0=90"
  "lon_0=0"
  "units=m"
END


ie:

[proj-4.4.9]# proj +proj=laea +ellps=WGS84 +lat_0=90 +lon_0=0
0 90
0.00    -0.07

[proj-4.4.9]# proj +proj=lcc +ellps=WGS84 +lat_0=90 +lon_0=0
0 90
0.00    0.00


I got things to work using laea. I think lcc has problems
mapping *over* top off the globe. Even though the distance
isn't that much, at the pole, the angles are. (polynomials ??)

I haven't tried mapping any rasters over the poles yet. I need
to find a GOES image or something that covers the pole.

My new problem is the graticule shapefile I have is in latlong.
It appears that some sort of polar oriented projection is preferred
once I get above 60 degrees. The north pole of my graticule
was shifting and rendering somewhere in Sweden (??). This problem
appears to be covered in the archive for this list.

I spent a few minutes to make a program to make a 5 degree line
graticule, for the northern hemisphere, with arcview and then I
converted it to laea centered on 90 north.
This will get loaded as the graticule once the viewer gets above
60 degrees latitude. Same for southern hemisphere. The
standard graticule from  60S to 60N will just be laea for
the current map nadir.
I will make one for southern hemisphere tomorrow as well.

Thanks for you help,

Peter