Re: Object Reference counting idea.

[Michael Meeks <michael helixcode com>]

Hi all,

        Here is my proposal ( based on a discussion with George Lebl in
Toronto about what happens in the panel ); which is simple, almost
entirely foolproof, trivial to implement and yet gives us a fair whack of 
protection. 
  
        We partition refs and unrefs into 2 sorts; local and
remote.

        * Remote is defined as coming through impl_Bonobo_Unknown_ref.
        * Local is defined as coming immediately though bonobo_object_ref.

        NB. 'Remote' calls thus can happen in-process for shared
libraries, or even by doing Bonobo_Unknown_ref/unref in the server
process.
 
        So; we have 2 reference counts; We flag errors and refuse to do
more Remote ref/unrefs than we have refs / unrefs. We have some policy for
the local case that could be:
 
        * Warn on too many unrefs but do it anyway
        * Warn and don't do it
        * Don't warn and do it.
 
        I don't much care about the latter policy, but I'd prefer Warn and
don't do it.
 
        So; the advantage of this is that remote scripts have no chance of
crashing the system by unreasonable ref / unrefs ( unless they try and
wrap the ref-count :-).
 
        I've been meaning to do this for some time. Of course, the obvious
advantage is simplicity, I'll hack up a patch sometime soon and see if it
impacts much existing code.
 
        Of course; if we try to do anything more for Gnome 2.0, I'd like  
to read up on distributed garbage collection, everyone keeps claiming this
is an unsolved problem and clearly there would be a load of issues. If    
anyone has some nice links on the subject I'd much appreciate them.

	Regards,

		Michael.

-- 
 mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot