[GRASS-dev] Re: [GRASS GIS] #1131: Global LFS for wingrass

GRASS GIS trac at osgeo.org
Mon Aug 16 09:45:49 EDT 2010


#1131: Global LFS for wingrass
--------------------------+-------------------------------------------------
 Reporter:  mmetz         |       Owner:  grass-dev@…              
     Type:  defect        |      Status:  new                      
 Priority:  critical      |   Milestone:  7.0.0                    
Component:  libgis        |     Version:  svn-trunk                
 Keywords:  LFS,wingrass  |    Platform:  MSWindows 7              
      Cpu:  All           |  
--------------------------+-------------------------------------------------

Comment(by mmetz):

 Hmm. This avoiding old names with

 {{{
 #define _NO_OLDNAMES
 }}}

 is causing more troubles than expected, too many old names are required.
 Trunk compiles on Windows without LFS-related errors, but there are now
 too many warnings about implicit declarations.

 {{{
 #define lseek lseek64
 }}}

 does not work with old names in use, compiler error because
 <grass/config.h> needs to be included before all others.

 Alternatively, we could not redefine existing functions/types, but define
 functions/types for grass, e.g.

 {{{
 #define g_off_t off64_t
 #define g_fseeko fseeko64
 #define g_ftello ftello64
 #define g_lseek lseek64
 /* use _stati64 compatible with MSVCRT < 6.1 */
 #define g_stat _stati64
 #define g_fstat _fstati64
 }}}

 but then these new names must always be defined, also on other platforms,
 and the codebase needs to be modified, each occurrence of lseek would need
 to be replaced with the new name, same for all other new definitions
 (there is a workaround for off_t, that could maybe stay). This is too much
 modification for my taste. I would like to postpone LFS support for
 wingrass in trunk a bit.

 Markus M

-- 
Ticket URL: <https://trac.osgeo.org/grass/ticket/1131#comment:13>
GRASS GIS <http://grass.osgeo.org>



More information about the grass-dev mailing list