[GRASS5] configure checks, fortran, static libraries

Markus Neteler neteler at itc.it
Wed Aug 8 04:16:24 EDT 2001


On Tue, Aug 07, 2001 at 11:24:50PM +0100, Glynn Clements wrote:
> 
> Markus Neteler wrote:
> 
> > > > > > PS: An idea for postgresql support: In case the postgres stuff is not found,
> > > > > >  maybe the support should be switched off? Then the NVIZ can be compiled
> > > > > >  straight forward, the user doesn't need to run configure again with
> > > > > >  --enable-postgres=no. Just an idea...
> > > 
> > > An excellent idea. (speaking as someone who's been trying to compile
> > > grass5 and wondering what I have to do to make it work, and
> > > --enable-postgres=no is probably what I need to do ..)
> > 
> > good news: This has been implemented today in CVS :-)
> > by Glynn. Seems we have synchronized ideas.
> 
> Clarification:
> 
> I've modified configure[.in] so that a failed check for a PostgreSQL
> header or library has the same effect as using --enable-postgres=no,
> i.e. USE_POSTGRES is set to empty.
> 
> I haven't modified anything else, so USE_POSTGRES itself is still
> being completely ignored.
> 
> Also, if the idea was to conditionalise PostgreSQL support in various
> programs (rather than controlling whether or not to build
> src.garden/grass.postgresql), USE_POSTGRES should probably be an
> AC_DEFINE (which affects config.h) instead of (or as well as) an
> AC_SUBST (which affects Makefile and src/CMD/head/head).

However, at time the NVIZ checks if PGLIB is present in the head
file or not. The test is at:
src.contrib/GMSL/NVIZ2.2/Gmakefile and
                         doconfigure

So at least the NVIZ compile is controled now.

Markus



More information about the grass-dev mailing list