Re: Any more regressions?

[Bowie Owens <bowie owens csiro au>]

Hi Murray,

I think orbitcpp is behaving itself pretty well now that the changes 
have settled. So there doesn't seem to be any problem there.

I have spent a little time cleaning out bugzilla. This included getting 
the any/simple test to work properly. So what we are left with is:

   * everything.idl doesn't compile because we don't support arrays
     within unions properly, but this didn't work before. This will
     require a bit of work. We need a new interface and virtual
     functions for correctly generating the accessor functions for
     union members. See the comments in the bug report.
   * WRITE ME bug indicating things that aren't finished.
   * Exceptions not defined problem. I need to spend some time
     reflecting on the standard and what other idl compilers do.

Since the 1.3.6 release the following improvements have been made:

   * Simplified c++ output.
   * Improved any support.
   * Reduced the number of leaked references.
   * Eliminated a number of dereference null pointer type errors.

It might be worth making a new release soon. So we can bump the required 
version of orbitcpp in libbonobomm to match.

Murray Cumming wrote:

> So, Bowie you seem to have fixed various problems that I introduced with
> my reorganisation of the code. Are there any regressions left that I
> should try to fix? The "everything" test doesn't build, but I don't know
> if it did before.
>
>  


--
Bowie Owens

CSIRO Mathematical & Information Sciences
phone  : +61 3 9545 8055
fax    : +61 3 9545 8080
mobile : 0425 729 875
email  : Bowie Owens csiro au