Re: cglib port?

Jon Tirsén <[email protected]> Fri, 17 Oct 2003 20:52:41 +0200 (CEST)
Newsgroups gmane.comp.java.nanning.devel
Message-ID <[email protected]>
 --- Kevin O'Neill <[email protected]> skrev: >
On Fri, 2003-10-17 at 16:45, Jon Tirs=E9n wrote:
> > Yes we looked into that, and we got it working.
> Only
> > problem was that serialization didn't work
> properly
> > (it worked for POJOs with interface, but not for
> POJOs
> > without). Since Nanning is so close to a 1.0
> release
> > we did not want to fiddle with the internals in
> ways
> > we didn't understand.
>=20
> Actually what I want to do is intercept (no mixins
> at this point)
> prevayler commands. So if it can't do serialization
> I'm kinda stuffed
> :).

Yeah... Actually looking into how serialization works,
it is *impossible* to instrument ordinary classes and
have serialization working at the same time.

> > Then again, if there's sufficiently popular demand
> for
> > it, I could implement it again. It's really not
> that
> > difficult and if you don't need serialization it
> works
> > fine.
>=20
> Is this something that the developer could configure
> at runtime?

Actually, the first prototype supported both at the
same time, ie. it used CGLIB when it couldn't use
dynamic proxies. A little bit messy but it worked.
I'll look into supporting that again.

> >  --- Kevin O'Neill <[email protected]> skrev:
> >
> > Reading the archieves from last month there was
> some
> > > talk about porting
> > > nanning from dyn proxies to cglib. Is that still
> the
> > > plan?
> > >=20
> > > If it is does it mean that nanning will be able
> to
> > > "aspect" pojos not
> > > just pojos with interfaces?
> > >=20
> > > -k.
> > > =20
> > > --=20
> > > If you don't test then your code is only a
> > > collection of bugs which=20
> > > apparently behave like a working program.=20
> > >=20
> > > Website: http://blogs.codehaus.org/people/kevin/
> > > Pub Key:
> > >
> http://blogs.codehaus.org/people/kevin/public.asc
> > >=20
> >=20
> > > ATTACHMENT part 2 application/pgp-signature
> > name=3Dsignature.asc
> > =20
> >=20
> > H=F6strusk och gr=E5 moln - k=F6p en resa till solen p=E5
> Yahoo! Resor p=E5 adressen
> http://se.docs.yahoo.com/travel/index.html
> > _______________________________________________
> > nanning-dev mailing list
> > [email protected]
> >
>
http://lists.codehaus.org/mailman/listinfo/nanning-dev
> --=20
> If you don't test then your code is only a
> collection of bugs which=20
> apparently behave like a working program.=20
>=20
> Website: http://blogs.codehaus.org/people/kevin/
> Pub Key:
> http://blogs.codehaus.org/people/kevin/public.asc
>=20

> ATTACHMENT part 2 application/pgp-signature
name=3Dsignature.asc
=20

H=F6strusk och gr=E5 moln - k=F6p en resa till solen p=E5 Yahoo! Resor p=E5=
 adressen http://se.docs.yahoo.com/travel/index.html