Re: bug in egcs with C++?

[teunis <teunis at computersupportcentre dot com>]

On 11 Dec 1998, Alexandre Oliva wrote:

> On Dec 11, 1998, Joe Buck <jbuck@Synopsys.COM> wrote:
> 
> > However, perhaps we could give a better diagnostic for this case;
> 
> > foo::__vtbl                     ttt.o
> > [Hint: try checking whether the first non-inlined, non-pure virtual function of class foo is defined]
> 
> I don't think this message would help.  I've already told this several 
> times to Teunis, but it didn't help solving the problem, probably
> because of misinterpretation of what I mean :-(

Sorry - wrong.  I -do- understand.  My C++ experience may be 5+ years out
of date but the code I've been writing is not unusual except perhaps in
the number of virtual objects.  I -do- have many objects with
unimplemented (=0) functions but they are never referred to in such a
way to cause this.  (I've also discovered that polymorphic objects aren't
happy with multiple base objects with the same root object...  I take it
this is expected?  I've assumed as much and changed my code style.  As
this code is kinda ported from Java I've been using a base "Object" that I
based everything on.  Now all pure virtual (read "interface") objects are
not based on "Object")

I'll try to make an official bug report.
That was the -first- thing I looked for actually.

But as I've said -many- times, all functions were implemented, both
virtual and non.

Though all of the cases contain -only- virtual functions + a constructor.
To repeat myself before I get sent another "you've failed to implement
this function" message I wish to say *again* that all functions were
implemented.

I shall try to duplicate in a smaller codebase.

G'day, eh? :)
	- Teunis

PS: now that I think about it I -do- use a garbage collecting lib...