DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20642>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20642

Lack of ctor definitions in XMemory prevents build on VC++ 7.0 with WPO enabled





------- Additional Comments From [EMAIL PROTECTED]  2003-06-10 17:40 -------
I've now been able to perform a build with WPO disabled, and it failed with the same 
error messages:

[...]
XercesLib error LNK2001: unresolved external symbol "private: __thiscall 
xercesc_2_3::XMemory::XMemory(class xercesc_2_3::XMemory const &)" ([EMAIL 
PROTECTED]@@[EMAIL PROTECTED]@@Z)
XercesLib error LNK2001: unresolved external symbol "private: __thiscall 
xercesc_2_3::XMemory::XMemory(class xercesc_2_3::XMemory const &)" ([EMAIL 
PROTECTED]@@[EMAIL PROTECTED]@@Z)
XercesLib error LNK2001: unresolved external symbol "private: __thiscall 
xercesc_2_3::XMemory::XMemory(class xercesc_2_3::XMemory const &)" ([EMAIL 
PROTECTED]@@[EMAIL PROTECTED]@@Z)
[...]


That is, it is not related to WPO as I first believed. I think it has more to do with 
how XMemory is used as a base class.

>This is an very common idiom in C++, and Microsoft's WPO option is going to
>break a lot of existing code.

Even if it had been releated to WPO, I don't see why WPO would break 'a lot of 
existing code'? Adding a dummy definition can't break any compiler (if it did, that 
compiler would be severly broken anyway), and once this is done the whole Xerces 
library builds fine and passes all tests. The optimization result with WPO is quite a 
lot better than without it.

>I think they ought to fix their compiler, rather than have huge amounts of
>source code modified for the sake of one compiler.

It's not a matter of changing huge amounts of code; it's only those two declarations 
that lacks definitions in Xerces. I'm sure MS will fix it, if it's an error; 
personally I would have to verify it with the C++ standard first. It might already 
have been corrected in VC++ 7.1 (VS.NET 2003), which I haven't been able to compile it 
with as of this.

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to