ps.map and Linux

[Richard E Nairn]

Ok, on to the next problem... I am using ps.select and ps.map.. If I
don't have a ps painter set up it will use the normal 8.5 / 11
definition. and it will turn out fine... However if I use ps.select to
select another page size and run the same script, the raster image
will come out blank...

Does anyone else have this problem??? This happens if I compile it
myself, or if I use the binary compiled version from Markus Neteler.
--
Richard Nairn
Geological Survey of Canada
** Email r...@spans.gscc.nrcan.gc.ca **
** nai...@cadvision.com **

[Richard E Nairn]

>
> Ok, on to the next problem... I am using ps.select and ps.map.. If I
> don't have a ps painter set up it will use the normal 8.5 / 11
> definition. and it will turn out fine... However if I use ps.select to
> select another page size and run the same script, the raster image
> will come out blank...
>
> Does anyone else have this problem??? This happens if I compile it
> myself, or if I use the binary compiled version from Markus Neteler.

Ok, after a little debugging, I found the problem with ps.map. The
C code that parses the device files is missing a header declaration,
so it produces garbage. It uses atof and atoi but it forgot to use
#include <stdlib.h> for the definitions. When i made the change it
works very nicely. The file involved was
src.alpha/ps.map/ps.map/read_cfg.c.

PS, I notified Markus so that he could patch the official sources.