[fdo-internals] automake/autoconf/libtool preferred version?

[Frank Warmerdam]

Greg Boone wrote:
> I am successfully building on Fedora 5 with the following:
> 
> autoconf (GNU Autoconf) 2.59
> automake (GNU automake) 1.9.6
> ltmain.sh (GNU libtool) 1.5.22 (1.2220.2.365 2005/12/18 22:14:06)
> 
> I have not seen any warnings or errors to speak of.

Greg,

I have upgraded to automake 1.9.5 and things are working substantially
better now.  At the very least this resolved the makefile problem.  Now I'm
running into this:

  g++ -DPACKAGE_NAME=\"FDO\" -DPACKAGE_TARNAME=\"fdo\" 
-DPACKAGE_VERSION=\"3.2.0\" "-DPACKAGE_STRING=\"FDO 3.2.0\"" 
-DPACKAGE_BUGREPORT=\"https://fdocore.osgeo.org\" -DPACKAGE=\"fdo\" 
-DVERSION=\"3.2.0\" -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 
-DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 
-DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -I. 
-I. -I../../Inc -I../Nls -I../Common -I../Geometry 
-I/home/warmerda/fdobase/Thirdparty/apache/xml-xerces/c/src -MMD -MT 
-DFULLPROTO -D__USE_GNU -DLINUX -DENABLE_BINRELOC -MT 
FeaturePropertyReaderImpl.lo -MD -MP -MF .deps/FeaturePropertyReaderImpl.Tpo -c 
Xml/FeaturePropertyReaderImpl.cpp  -fPIC -DPIC -o .libs/FeaturePropertyReaderImpl.o
Xml/FeaturePropertyReaderImpl.cpp: In member function `virtual FdoBoolean 
FdoXmlFeaturePropertyReaderImpl::XmlEndElement(FdoXmlSaxContext*, const 
FdoString*, const FdoString*, const FdoString*)':
Xml/FeaturePropertyReaderImpl.cpp:536: error: no matching function for call to 
`xercesc_2_5::Base64::decode(const XMLByte*, FdoSize*)'
/home/warmerda/fdobase/Thirdparty/apache/xml-xerces/c/src/xercesc/util/Base64.hpp:127: 
note: candidates are: static XMLByte* xercesc_2_5::Base64::decode(const 
XMLByte*, unsigned int*, xercesc_2_5::MemoryManager*)
/home/warmerda/fdobase/Thirdparty/apache/xml-xerces/c/src/xercesc/util/Base64.hpp:148: 
note:                 static XMLCh* xercesc_2_5::Base64::decode(const XMLCh*, 
unsigned int*, xercesc_2_5::MemoryManager*)

This patch seems to get things building (cast the FdoSize * "&len"
to be an unsigned int *), but I suspect that this isn't legitimate
if FdoSize is actually 64bits in size on 64bit machines.  Is that
the case?

warmerda at amd64$ svn diff Fdo/Xml/FeaturePropertyReaderImpl.cpp
Index: Fdo/Xml/FeaturePropertyReaderImpl.cpp
===================================================================
--- Fdo/Xml/FeaturePropertyReaderImpl.cpp       (revision 2498)
+++ Fdo/Xml/FeaturePropertyReaderImpl.cpp       (working copy)
@@ -533,7 +533,8 @@
      case ParsingState_base64Binary:
          {
              FdoSize len;
-            XMLByte* decoded = XERCES_CPP_NAMESPACE::Base64::decode((const 
XMLByte*)(const char*)m_dataProperty, &len);
+            XMLByte* decoded = XERCES_CPP_NAMESPACE::Base64::decode((const 
XMLByte*)(const char*)m_dataProperty,
+                                                                    (unsigned 
int *) &len);
              if (decoded != NULL) {
                  curFeatureHandler->FeatureBinaryData(m_featureContext, 
decoded, len);
                  XERCES_CPP_NAMESPACE::XMLString::release(&decoded);

-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org