Jean-Claude,
I got an update about the compiler issue... don't know what you want to do about it (change the code back and edit the Makefiles, or just leave it changed for now), but here is the answer that works:
=============================
Do you get the same problem if you compile with this option:
-features=%all,no%iddollar,tmplife
The compiler assumes -features=%all,no%iddollar,no%tmplife when no -features option is given, and this affects when temporary objects are destroyed. Interestingly, there is the following comment on the man page:
The behavior of a program might change when you
use -features=tmplife. Testing whether the pro-
gram works both with and without the
-features=tmplife option is one way to test the
program's portability
=============================
Givin the last statement, I think that you should keep the code changed, but that's just my opinion.
Later,
Dan
Jean-Claude Wippler wrote:
Dan Gregory wrote:I can compile ok, but the test cores on b14(actually running in the debugger can get it to run correctly) when the view is accessed via GetAs(). This tries to invoke ~c4_string and when it dereferences p, it cores. If are having this problem then you can replace this > method:
const char* c4_HandlerSeq::Description()
{
return _field != 0 ? UseTempBuffer(Definition().DescribeSubFields()) : 0;
}
with this one:
const char* c4_HandlerSeq::Description()
{
const char* returnPtr = (_field != 0) ? UseTempBuffer(Definition().DescribeSubFields()) : 0;
return returnPtr;
}
I think the problem lies somewhere with the destruction of temporary variables. The temporary c4_String variable returned from DescribeSubFields get descructed before Definition() or DescribeSubFields() gets called. It seems totally hosed... I am in the process of discussing this with sun to see if it's a bug or not.
I've made changes in the source - better safe than sorry. In CVS now. Thanks. -jcw _______________________________________________ metakit mailing list - [EMAIL PROTECTED] http://www.equi4.com/mailman/listinfo/metakit
_______________________________________________ metakit mailing list - [EMAIL PROTECTED] http://www.equi4.com/mailman/listinfo/metakit
