[Gdal-dev] Build failure with VC6 (and continued failures with mingw)

[Andrea Aime]

Hi,
today I tried to build gdal again with VC6 to find it does not
build anymore due to changes in the Integraph module:

icrosoft (R) Program Maintenance Utility   Version 6.00.8168.0
Copyright (C) Microsoft Corp 1988-1998. All rights reserved.
         cl /nologo /MD /EHsc /Zi /Fd..\..\gdal.pdb -I..\..\port 
-I..\..\ogr -I..\..\gcore  -I..\..\alg -D_USE_MATH_DEFINES 
-I..\gtiff\libtiff /c intergraphdataset.cpp
intergraphdataset.cpp
IngrTypes.h(62) : error C2632: 'long' followed by 'long' is illegal
IngrTypes.h(63) : error C2632: 'long' followed by 'long' is illegal
intergraphdataset.cpp(445) : error C2374: 'i' : redefinition; multiple 
initialization
         intergraphdataset.cpp(410) : see declaration of 'i'
intergraphdataset.cpp(451) : error C2374: 'i' : redefinition; multiple 
initialization
         intergraphdataset.cpp(410) : see declaration of 'i'
NMAKE : fatal error U1077: 'cl' : return code '0x2'
Stop.

I also tried going on with mingw, which compiles fine gdal and ogr,
but then starts complaining in the swig binding directory about
issues I don't remember about with the java OGR bindings.
I also tried building a swig version specific for MinGW, as suggested
on the SWIG site, but failed because of autoconf issues...
the msys and mingw guys do not seem to be getting along very well 
apparently, there are two autoconf versions around, one distributed 
along with mingw, that does not work out of the box because I miss
perl, which is distributed with the msysdtk.... which installs its own
version of autoconf and automake too.

Oh well, I had enough for today. If I have some half a day to throw
away next weekend I'll try again the visual c++ express path (which
I have to reinstall from scratch, find the right windows SDK, the
right incantations to pacth atl, and so on... would be nice to have
a guide).

Cheers
Andrea