Current (hour or so ago) cvs Win32 build problems with VC5


Subject: Current (hour or so ago) cvs Win32 build problems with VC5
jeremyd@computer.org
Date: Sat Aug 25 2001 - 17:15:57 CDT


Hello,

These errors are from cvs versions about an hour or so ago and
using make distribution ABI_BUILD_VERSION=x.y.z method.

In addition to the ut_rand.cpp problem described in previous messages,
the recent (somewhere between just before 0.9.2 was released and
about an hour or so ago) changes to abi\src\af\util\xp\ut_iconv.cpp
cause it to complain about unexpected end of file.
The problem is with the UT_TRY & UT_CATCH macros, with there new
placement. (adding {} to the extern "C" changes the error message
to something similar to not all paths return a value). Commenting
the UT_TRY line and corresponding UT_CATCH section results in it
compiling fine. I'm not sure the proper fix.

abi\src\wp\impexp\xp\ie_exp_HTML.cpp should not include unistd.h,
as it is not available. Instead for mkdir (and possibly other
functions) direct.h should be included. Additionally, it only
takes 1 argument, so
                        int result = mkdir(fname, 0750);
needs to be
                        int result = mkdir(fname);

abi\src\wp\impexp\xp\ie_imp.cpp complains:
"ie_imp.cpp(173) : error C2440: 'reinterpret_cast' : cannot convert from 'int' to
 'int'
                               Conversion is a valid standard conversion, which
can be performed implicitly or by use of static_cast, C-style cast or function-s
tyle cast
ie_imp.cpp(174) : error C2440: 'reinterpret_cast' : cannot convert from 'int' to
 'int'
                               Conversion is a valid standard conversion, which
can be performed implicitly or by use of static_cast, C-style cast or function-s
tyle cast"
and will compile changing:
        if (s->supportsFileType(reinterpret_cast<IEFileType>(a+1)))

                return reinterpret_cast<IEFileType>(a+1);

to:
        if (s->supportsFileType((a+1)))

                return (a+1);

Otherwise, it builds fine (not counting warnings) and seems to run
correct (given tiny testing) once installed. If any of these are
fixed, sorry for the duplicate information, it takes me a while to
get the source from cvs and to do the build.

Thanks,
Jeremy Davis
jeremyd@computer.org



This archive was generated by hypermail 2b25 : Sat Aug 25 2001 - 17:53:41 CDT