[omniORB] objectId problem with PERSITENT SYSTEM_ID POA

[Duncan Grisby]

On Saturday 12 October, bjorn rohde jensen wrote:

>   I have a bit of a problem with the objectId's omniORB4 is
> generating for objects activated on a POA with policies PERSISTENT
> and SYSTEM_ID.  Maybe i have made some sort of error, but it seems
> to me, that the objectId's generated are just a simple count of
> object activated in the POA so far and not at all unique across all
> instantiations of the POA as specified in section 11.3.7.4 of the
> specs.

You're right. I'm surprised nobody has noticed until now. I have
checked in a fix to omniORB 4's CVS. In case someone was relying on
the old behaviour (by assuming the length of a system id was 4), there
is a new poaUniquePersistentSystemIds configuration parameter. The
default is true, meaning ids are unique across all runs; setting it
false gets back the old behaviour.

Note that omniNames relied on the old behaviour. I've changed it, but
it's important that you don't run an old omniNames against a new
omniORB.

Cheers,

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan@grisby.org     --
   -- http://www.grisby.org --