> This happens on SPARC too so it's no compiler specific issue.

As Sun is probably using the same C++ frontend on both platforms, it probably is a 
compiler issue. I'm not quite sure if it is a compiler bug or if they are 'strict, but 
just' with regard to the C++ standard.

> "/home/sapdb/maxdb-76/sys/wrk/incl/SAPDB/SAPDBCommon/SAPDB_Ran
> geCode.hpp\
> ", line 180: Error: Different types for "?:" 
> (SAPDB_ToStringClass and co\
> nst char*).

The  compiler seems to be of the opinion that the second and third operand of the ?: 
operator have to be of the same type. Other compilers are satisfied if both can be 
converted to the target type.

This can probably be fixed by the following lines:
@@ -177,8 +177,8 @@
                                              , SAPDB_ToString((const void 
*)destination)
                                              , SAPDB_ToString(destination_position)
                                              , SAPDB_ToString(lengthInBytes)
-                                             , (fillBytes ? 
SAPDB_ToString(fillBytes[0]) : "null fillbytes")
-                                             , (fillBytes ? 
SAPDB_ToString(fillBytes[1]) : "") ) );
+                                             , (fillBytes ? (char*) 
SAPDB_ToString(fillBytes[0]) : "null fillbytes")
+                                             , (fillBytes ? (char*) 
SAPDB_ToString(fillBytes[1]) : "") ) );
             SAPDB_HandleMoveErrorMessageList(errList);
 
             errorCode = e_move_error;

> "/home/sapdb/maxdb-76/sys/wrk/incl/SAPDB/SAPDBCommon/SAPDB_Ran
> geCode.hp\
> p", line 172: Error: Formal argument MessageList of type 
> const SAPDBErr_\
> MessageList& in call to 
> SAPDBErr_MessageList::AppendNewMessage(const SAP\
> DBErr_MessageList&) is being passed int.

I guess this goes away when the first problem is fixed. 
=> the compiler doesn't recognise the types of two arguments of SAPDBErr_MsgList
=> doesn't find the proper function definition
=> doesn't know the return type
=> uses int as default
=> gets a type mismatch for the call to AppendNewMessage

Do all the errors have their origin in SAPDB_RangeCode.hpp?

Daniel Dittmar

P.S. Sorry that I didn't integrated the Solaris/i386 patches yet. But several of the 
patches consist of removing the comments of assmbler files. And I rather add a script 
to our build tools to do this during make.

-- 
Daniel Dittmar
SAP Labs Berlin
[EMAIL PROTECTED]


-- 
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]

Reply via email to