[GRASSLIST:3532] Re: r.mapcalc3

Glynn Clements glynn.clements at virgin.net
Fri May 28 10:16:56 EDT 2004


Maciek Sieczka wrote:

> when I compile Grass 5.3 CVS or the official source release from 15.05.04 I
> get the following in the end:
> 
> GRASS GIS compilation log
> -------------------------
> Start of compilation: pia maj 28 13:35:04 CEST 2004
> Errors:
> Compilation error in module: src/raster/r.mapcalc3 (ignored)
> End of compilation:   pia maj 28 13:54:45 CEST 2004
> DONE generating GRASS GIS binary code
> 
> r.mapcalc and r3.mapcalc seem ok so what is that r.mapcalc3?

That is where the actual r.mapcalc program comes from. The code in
src/raster/r.mapcalc has been superseded by the version in
src/raster/r.mapcalc3, and isn't compiled by default.

In general, you need to examine the actual error message in the output
from make to determine why compilation failed. However, in this case,
it's almost certainly because src/libes/g3d hadn't been built at that
point.

The src/raster/r.mapcalc3 directory has recently been updated to build
both r.mapcalc and r3.mapcalc from (mostly) the same source files. The
r3.mapcalc program requires the G3d library (src/libes/g3d), which
wasn't built until later (I think that this has been fixed now).

However, the old r3.mapcalc (in src.contrib/GMSL/g3d/src3d/raster/r3.mapcalc)
is currently still being built; that will overwrite the version from
src/raster/r.mapcalc3. Also, src/raster/r.mapcalc3 builds r.mapcalc
first, so that should still be present even if r3.mapcalc fails to
compile.

IOW, the error should be harmless.

-- 
Glynn Clements <glynn.clements at virgin.net>




More information about the grass-user mailing list