Ant, just what I was wanting to mention - you beat me to it.  I really
think additions may not be a disturbance to the SPIs.  This could be a
start to looking at how much of the extension-helper we could start
moving into the SPIs. But am a bit curious about where this sort of
thing will go into the SPI.  Do we intend another package
org.apache.tuscany.sca.assembly in the spi module ?

Thanks

- Venkat


On 8/8/07, ant elder <[EMAIL PROTECTED]> wrote:
> The ExtensionHelper module was originally intended as just an interim thing
> while we were sorting out the SPIs anyway, so if we  go ahead with this its
> fine to move those abstract classes into the SPI module. Just adding
> something new to the SPI module seems fine to me as its not going to break
> anything so I'm still +1 on doing this.
>
>    ...ant
>
> On 8/8/07, Simon Laws <[EMAIL PROTECTED]> wrote:
> >
> > On 8/8/07, Venkata Krishnan <[EMAIL PROTECTED]> wrote:
> > >
> > > Hi,
> > >
> > > I just observed that, not all bindings are extending from
> > > AbstractBinding though all implement the interface Binding.  The
> > > AbstractBinding class provides an implementation for the Binding
> > > interface and in my opinion should be used by all bindings to avoid
> > > duplication of code that implements the BInding interface.
> > >
> > > I feel the pinch for this now as I am adding a couple of things for
> > > policies and seems a bit odd to go an copy over the same things across
> > > a dozen classes.
> > >
> > > Let me know what people feel about this and I can go and fix all the
> > > bindings for this.
> > >
> > > Thanks.
> > >
> > > - Venkat
> > >
> > > ---------------------------------------------------------------------
> > > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > > For additional commands, e-mail: [EMAIL PROTECTED]
> > >
> > >
> > I like the idea but it gives me a problem with the sca binding. Adding it
> > to
> > SCABinding would imply a cyclic dependency as AbstractBinding is
> > implemented
> > in ExtensionHelper (as putting it in the SPI would imply and SPI change I
> > guess) which in turn depends on quite a lot of other stuff. If we could
> > separate out AbstractBinding somehow it would work for me. Maybe needs  to
> > wait until our SPI sweep. +1 to using it where ever it can safely be used.
> >
> > Simon
> >
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to