Roger André wrote:
I've been mercilessly beating a dead horse on a couple of lists about
some weird artifacts that I'm getting when doing reprojections in
Mapserver from epsg:4326 to epsg:900913. It's recently come to my
attention that something was discussed in an IRC chat that might be
pertinent to this issue, and I was hoping I could test it.
Specifically, it relates to SOURCE_EXTRA, and although the conversation
at the time dealt with issues at the International dateline, I think it
might also manifest itself in other edge cases (literal edges). ;).
I would like to know if it is possible to set SOURCE_EXTRA to something
other than 1 at compile time? I do not see it listed as an option in
the configure documentation, but perhaps I could edit a file, or pass in
the flag anyhow?
Roger,
I haven't followed your problems closely (I recall something about datum
shifts). The SOURCE_EXTRA only helps with the GDAL warper. If you wanted
you could modify the following code in gdal/alg/gdalwarpoperation.cpp
and recompile:
/* -------------------------------------------------------------------- */
/* Allow addition of extra sample pixels to source window to */
/* avoid missing pixels due to sampling error. In fact, */
/* fallback to adding a bit to the window if any points failed */
/* to transform. */
/* -------------------------------------------------------------------- */
if( CSLFetchNameValue( psOptions->papszWarpOptions,
"SOURCE_EXTRA" ) != NULL )
{
nResWinSize += atoi(
CSLFetchNameValue( psOptions->papszWarpOptions, "SOURCE_EXTRA" ));
}
else if( nFailedCount > 0 )
nResWinSize += 10;
Perhaps change the nResWinSize += 10 line to nResWinSize += 100 or something
like that.
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, [EMAIL PROTECTED]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | Geospatial Programmer for Rent
_______________________________________________
gdal-dev mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/gdal-dev
