Re: bug in BinaryDeserializer?

On Mon, Apr 28, 2003 at 04:03:44PM +0100, Matthias Radestock wrote:
> [email protected] wrote:
> 
> >Since its a JavaObject, we probably actually want the 
> >ser/deser methods or the contract of JavaObject to hint to the 
> >serialization that this needs to be enforced.
> 
> It's the contract on the Hashtable that matters since generally it's 
> perfectly ok to ser/deser JavaObject the way we do.
> 
> What I don't understand is why the (de)serializer does what it does. Is 
> it employing a breadth first strategy in order to avoid deep stacks?

Yes.  
> 
> In that case having a {read/write}DirectExpression method that uses 
> depth first instead would work.
> 
Yeah, thats what I'm going for. 

	Scott

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+rWXgr9IW4v3mHtQRAvDaAJoC+dH5JG2SoPgbO3K7YgPozegCYACffEKl
OGUitlVOWrygyim/KmUVR2M=
=6EFM
-----END PGP SIGNATURE-----