Mailing-List: contact dev-help@cxf.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cxf.apache.org
Received-SPF: neutral (athena.apache.org: local policy)
From: Daniel Kulp <dkulp@apache.org>
To: dev@cxf.apache.org
Subject: Re: CXF-1891 another revenant from XFire
Date: Mon, 3 Nov 2008 12:09:54 -0500
User-Agent: KMail/1.9.9
Cc: "Benson Margulies" <bimargulies@gmail.com>
References: <61b5d9410811030517x2fc62decm3ecd2b43731640d5@mail.gmail.com>
 <20302927.post@talk.nabble.com>
 <61b5d9410811030628q473ecd13vac46eeb4be088f05@mail.gmail.com>
In-Reply-To: <61b5d9410811030628q473ecd13vac46eeb4be088f05@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200811031209.54232.dkulp@apache.org>

On Monday 03 November 2008 9:28:41 am Benson Margulies wrote:
> In this particular case, it might be worth considering whether the
> particular use case holds water. 'Gee, I have these XMLBeans over here
> for some of my types, but I can't/don't want to bother to either build
> out XMLBeans for all of my types or build JAXB for those. Couldn't I
> mix and match? It seems incompletely crazy to me (and much like I've
> been learning about JAX-RS), but it also seems like a big job to do it
> coherently.

With XFire, it was even more complex.   There was a circular dependency 
between the xmlbeans databinding and the aegis databinding which was just 
nuts.   The xmlbeans databinding also delegated down into aegis for anything 
it didn't know about (in particular, anything "primitive").   I DIDN'T do 
this for the cxf version of XMLBeans as I didn't want aegis to be a required 
dependency to use XMLBeans.  (and didn't want Aegis to depend on XMLBeans)

Thus, if I was going to tackle this, I would actually tackle this from within 
the XMLBeans databinding.   If the xmlbeans databinding cannot get an 
xmlbeans type for something, possible callout to other databindings to see if 
they want to handle it.

Actually, this could be an opportunity to refactor the databinding stuff 
slightly to make it more usable by the JAXRS stuff.   If the databindings 
could be registered someplace on the bus, and also have a more generic 
factory type API, all the databindings could be slowly updated to delegate 
out if something is unknown and JAXRS could use the same stuff to reuse some 
plugability.

Dan


>
> On Mon, Nov 3, 2008 at 8:40 AM, Glen Mazza <glen.mazza@gmail.com> wrote:
> > I share your sentiments that you placed in put in that bug report, namely
> > that CXF, like Metro, does not have a requirement or an architectural
> > need to do everything that XFire does.  BTW, for the benefit of others,
> > here's the link: https://issues.apache.org/jira/browse/CXF-1891
> >
> > Benson Margulies-4 wrote:
> >> Any other committers care to weigh in on this?
> >
> > --
> > View this message in context:
> > http://www.nabble.com/CXF-1891-another-revenant-from-XFire-tp20302600p203
> >02927.html Sent from the cxf-dev mailing list archive at Nabble.com.


-- 
Daniel Kulp
dkulp@apache.org
http://dankulp.com/blog