| Age | Commit message (Collapse) | Author |
|
createOperations(): fix double vertical unit conversion from CompoundCRS to other CRS when the horizontal part of the projected CRS uses non-metre unit
|
|
|
|
other CRS when the horizontal part of the projected CRS uses non-metre unit
Fix issue reported on https://lists.osgeo.org/pipermail/proj/2019-October/008939.html
|
|
Fix two sentences that seem to be copy-and-paste forgotten.
|
|
|
|
Use dedicated logo for website
|
|
Fixes #1670
|
|
Database: update to EPSG v9.8.3
|
|
doxygen build script: don't output to /tmp
|
|
'RGSPM06 to Danger 1950 height' grids
|
|
The doxygen log-files was previously put in /tmp. This is usually fine
on UNIX-type systems but when building the docs on Windows with Cygwin/MSYS2
tools the script can fail since that path is guaranteed to exist. This
commit makes sure that the doxygen script outputs its files in a path
relative to the current working directory.
|
|
Note: a hack has been added into customizations.sql to cancel deprecatation
of USA geoid2012 grids by geoid2018 grids, as the later are not yet available
in proj-datumgrid-northamerica (https://github.com/OSGeo/proj-datumgrid/issues/55)
|
|
|
|
|
|
Remove sentence saying that coordinate system might by null
|
|
return null.
It is true for datum(), but does not apply to coordinateSystem().
|
|
ob_tran doc: fix/clarify semantics of o_lat_p/o_lon_p
|
|
proj.pc.in: add Libs.Private
|
|
rouault/fix_custom_compound_crs_with_NAD83_2011_and_geoidgrid_to_WGS84_G1762
proj_create_crs_to_crs(): remove elimination of Ballpark operations that caused transformation failures in some cases
|
|
verticalCRS to a 2D CRS
|
|
caused transformation failures in some cases
|
|
Add -lstdc++ to Libs.Private to allow packages such as libgeotiff to
retrieve this dependency when linking statically with proj
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
|
|
Add logo to website
|
|
|
|
Courtesy of Sylvain Beorchia. The logo is put under the CC-0 license.
|
|
The PROJ Conda package is now simply called "proj". Also mention that the grid packages can be installed with Conda.
|
|
Add a proj_crs_demote_to_2D(). Useful if forced to export a 3D CRS to a best approximate as WKT1 that doesn't support it
|
|
aeqd: for spherical forward path, go to higher precision ellipsoidal case when the point coordinates are super close to the origin (fixes #1654)
|
|
approximate as WKT1 that doesn't support it
|
|
|
|
when the point coordinates are super close to the origin (fixes #1654)
|
|
string
|
|
|
|
This is a super confusing topic, but from experiments, and in particular with
the example of Fairgrieve projection, o_lat_p and o_lon_p are the coordinates
of the North pole of the unrotated geographic CRS expressed in the rotated
geographic CRS.
This is different from defining a 'new pole', which would be the coordinates
of the rotated North pole expressed in the unrotated CRS.
|
|
Only call pkg-config in configure when necessary
|
|
Only call pkg-config in the configure script when the variables $SQLITE3_CFLAGS and $SQLITE3_LIBS are not set.
This allows building PROJ when pkg-config is not available.
|
|
Add API and WKT mapping for 'nsper' to EPSG Vertical Perspective method
|
|
Relates to https://github.com/OSGeo/gdal/issues/1856
|
|
|
|
|
|
Fix some Cppcheck complaints in geodesic routines
|
|
|
|
|
|
Improve vertical transformation support
|
|
PROJ string CRS ingester: recognize more unit-less parameters, and general handling of +key=string_value parameters
|
|
|
|
handling of +key=string_value parameters
|
|
|
|
When we had a transformation between a compoundCRS and a target geographicCRS,
we did not take into account that in the vertical->other_geog_CRS transformation
we used, the other_geog_CRS was an implicit interpolation CRS. Thus before
doing vertical adjustment, we must go to this interpolation CRS.
The workflow is thus:
source CRS -> interpolation CRS + vertical adjustment + interplation CRS -> target CRS
|
|
rouault/fix_proj_create_crs_to_crs_if_only_ballpark_transf
proj_create_crs_to_crs(): fix when there are only transformations with ballpark steps
|