From: "Shot (Piotr Szotkowski)" <shot@...>
Date: 2008-02-22T04:01:46+09:00
Subject: Re: Object#freeze as a basis for caching of method results?

--pKvQgaf82iZAytyH
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Robert Klemme:

> Demarshalling was the option I had thought of. A copy through
> marshalling _might_ use less memory than the marshaled form
> (the String).

Right, but then I=E2=80=99m back to defining a sensible
#=3D=3D, #eql? and #hash for the class in question.

On the other hand, with serialization I=E2=80=99m just silently handing ove=
r the
decision to Marshal, so I might as well (from the logical point of view,
barring the resource overhead) assume they=E2=80=99re defined as

def =3D=3D other
  Marshal.dump(self) =3D=3D Marshal.dump(other)
end

def eql?
  Marshal.dump(self).eql? Marshal.dump(other)
end

def hash
  Marshal.dump(self).hash
end

Hmmm. :)

I=E2=80=99ll go with using the Marshal strings as keys and see how
expensive it turns out to be. Thanks a lot again, Robert!

-- Shot
--=20
A database is a black hole into which you put your data. If
you're lucky, you'll get it back again. If you're very lucky,
you'll get it back in a form you can use.    -- Charlie Gibbs

--pKvQgaf82iZAytyH
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQFHvb8ki/mCfdEo8UoRApUkAJ44H83ciqQyqbH/N4rXilArbjfz6QCfT2sX
HbmOtl/0lySNkNZiS5v1qEo=
=DZ4k
-----END PGP SIGNATURE-----

--pKvQgaf82iZAytyH--