[GRASS5] ERROR: MAPSET fakestart not found

Huidae Cho grass4u at gmail.com
Wed Nov 3 16:47:22 EST 2004


On Wed, 3 Nov 2004 13:39:49 +0100, Markus Neteler <neteler at itc.it> wrote:
> On Wed, Nov 03, 2004 at 01:26:05AM -0600, Huidae Cho wrote:
> > Hi,
> >
> > I got the following error when making r.digit:
> >
> > GRASS_FAKE_START=1 GISBASE=... dist.i686-pc-cygwin/bin/r.digit
> > --html-description
> > ERROR: MAPSET fakestart not found.
> 
> This error happens in several modules. Someone will have to debug
> it and add further fakestart tricks into libgis.
> 
> In this case the line in main.c
>  polyfile = G_tempfile();
> is the problem (the function needs to be extended as it was done
> for e.g. G_gisinit().
> 
> To debug it, just set in a shell all the variables such as
> 
>  export GRASS_FAKE_START=1
>  export GISBASE=/path/to/dist.i686-pc-linux-gnu
>  export LD_LIBRARY_PATH=/path/to/dist.i686-pc-linux-gnu/lib
>  ddd /path/to/bin/r.digit --html-description
> 
> > r.digit does not call G_parser(), so it cannot generate usage. But it
> > fails with "Programmer error: no flags or options" error after
> > inserting G_parser() to make a usage file. "No flags or options" can
> > be intended by programmers, should it be something like fatal errors?
> 
> You mean above problem? No, that should not stop compilation.

I mean "do we need the 'Programmer error: no flags or options' error?"
If no options or flags are defined, it is intended by developers, so
the module should continue instead of exit even if G_parser() is
called to generate usage..

Huidae Cho

> 
> Markus
>




More information about the grass-dev mailing list