[GRASS5] 5.7: r.terraflow and gcc 2.96 fails

Hamish hamish_nospam at yahoo.com
Wed Nov 17 21:31:35 EST 2004


> Currently the automated 5.7 CVS Linux binary snapshot
> fails:
> 
> bash-2.04$ cd
> /hardmnt/grass0/neteler/grass51binarybuild/grass51/raster/r.terraflow/
> bash-2.04$ make Makefile:78: warning: overriding commands for target
> `clean'../../include/Make/Rules.make:35: warning: ignoring old
> commands for target `clean' mkdir -p OBJ.i686-pc-linux-gnu/FLOAT ;
> true mkdir -p OBJ.i686-pc-linux-gnu/SHORT ; true
> c++ -c -g -O2 -I./IOStream/include -DUSER=\"neteler\" -DNODATA_FIX
> -D_FILE_OFFSET_BITS=64
> -I/hardmnt/grass0/neteler/grass51binarybuild/grass51/include
> -DELEV_FLOAT main.cc -o OBJ.i686-pc-linux-gnu/FLOAT/main.o In file
> included from IOStream/include/ami.h:61,
>                  from common.h:49,
>                  from main.cc:59:
> IOStream/include/empq_impl.h:44:19: ostream: No such file or directory
> In file included from plateau.h:50,
>                  from water.h:51,
>                  from fill.h:46,
>                  from main.cc:60:
> direction.h:46:19: ostream: No such file or directory
> make: *** [OBJ.i686-pc-linux-gnu/FLOAT/main.o] Error 1
> 
> bash-2.04$ gcc -v
> Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/2.96/specs
> gcc version 2.96 20000731 (Red Hat Linux 7.2 2.96-112.7.1)
> 
> Probably the recent fixes in 5.3-CVS addressed this problem?

No.

gcc 2.96 is missing the <ostream> include header. <ostream.h> is still
available in gcc 3.3 (not sure about 3.4?) for backwards compatibility.

Therefore we have to use  #include <ostream.h>  and live with the
compiler warning that ostream.h is depreciated.

This has not been reverted yet (for a second time) in either 5.3 or 5.7:
 http://freegis.org/cgi-bin/viewcvs.cgi/grass/src.contrib/DUKE/r.terraflow/common.cc

the recent fixes to 5.3 were before the 5.7 copy-mix, they are the same.

see also
 http://grass.itc.it/pipermail/grass5/2004-November/015840.html


I'm just about to head out to sea for a while so I can't deal with this now..



Hamish




More information about the grass-dev mailing list