>>>>> "Allan" == Allan Rae <[EMAIL PROTECTED]> writes:
Allan> JMarc has already run into this one. It's been reported to Karl
Allan> (sigc++ maintainer) but it looks like cxx reads every line of
Allan> the file even the ones it knows it doesn't have to interpret --
Allan> looking for preprocessor directives it doesn't understand so it
Allan> can give you a warning.
Yes, it seems that's the reason. Karl hinted he'll use #error there.
Allan> Do either of you (Angus and JMarc) have SGI's STL installed or
Allan> are prepared to install it? I'm wondering if it's a compiler
Allan> problem or a problem with the way Compaq/Digital have written
Allan> their library. Try compiling both with and
Allan> without-included-string with SGI's STL.
Sorry, I won't have time to do that.
>> Compiling with gcc 2.95.2 =================== Can compile
>> FormCopyright. Compilation fails in FormPrint:
>> ../../../xtl/xdr.h:32: endian.h: No such file or directory
>> ../../../xtl/xdr.h:33: byteswap.h: No such file or directory
Allan> I would have thought that gcc should have included these
Allan> headers or something like them in its distribution.
I do not think there is anything standard in these headers. They do
not exist on the systems I have access to.
>> cxx: Warning: ../../../xtl/objio.h, line 391: "typeid" is reserved
>> for future use as a keyword decl_oobj_method(2)
Allan> That's an interesting warning. I should have a look at the code
Allan> there and see what it isn't doing as a result.
Probably releated to the fact that the script I (and Angus) tried sets
-nortti :) Is rtti a must have here?
JMarc
