[GRASS-dev] Problem compiling ximgview in the latest grass_trunk

[Pierre Roudier]

Glynn,

Is there an easy work-around this? It's a bit weird as the compilation
goes well on my older machine. It seems to fail only on this newer
configuration. After a bit of googling, I *think* that the xutils-dev
package, which provides imake, does not provide the xmkmf script
anymore.

I don't think I'm making any use of the ximgview module at the moment,
so I'm happy to try and drop it from my final install. I do need
grass7 to compile for testing Eric Momsen's GSoC project.

Any help greatly appreciated,

Cheers,

P

2012/7/13 Glynn Clements <glynn at gclements.plus.com>:
>
> Pierre Roudier wrote:
>
>> : && gcc -L/usr/local/src/grass_trunk/dist.x86_64-unknown-linux-gnu/lib
>> -L/usr/local/src/grass_trunk/dist.x86_64-unknown-linux-gnu/lib
>> -Wl,--export-dynamic
>> -Wl,-rpath-link,/usr/local/src/grass_trunk/dist.x86_64-unknown-linux-gnu/lib
>>  -o /usr/local/src/grass_trunk/dist.x86_64-unknown-linux-gnu/bin/ximgview
>> OBJ.x86_64-unknown-linux-gnu/color.o
>> OBJ.x86_64-unknown-linux-gnu/main.o    -lgrass_gis.7.0.svn -L -lX11
>                                                              ^^
>
> The problem is the bare -L switch, which causes -lX11 to be treated as
> an argument to -L rather than a separate switch.
>
> Unfortunately, the AC_PATH_XTRA test which sets X_LIBS is part of
> autoconf, and not something which can easily be worked around.
>
> First it tries using imake (xmkmf), which may not exist on modern
> systems. If that fails, it tries a fixed set of plausible library
> directories, all of which use "lib" rather than e.g. "lib64", as the
> latter wasn't in common use when autoconf-2.13 was released (Jan
> 1999).
>
> It may be time to think about moving to a newer version of autoconf.
> It's much more stable now than it was in the period immediatley after
> 2.13.
>
> --
> Glynn Clements <glynn at gclements.plus.com>



-- 
Scientist
Landcare Research, New Zealand