Mailing-List: contact geronimo-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: geronimo-dev@incubator.apache.org
Message-ID: <40B44B87.8090502@coredevelopers.net>
Date: Wed, 26 May 2004 00:47:19 -0700
From: Jeremy Boynes <jeremy@coredevelopers.net>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US;
 rv:1.6b) Gecko/20031205 Thunderbird/0.4
MIME-Version: 1.0
To: geronimo-dev@incubator.apache.org
Subject: Re: AW: Changes to OpenEJB interceptor stack
References: <003301c442ee$fafde750$7201a8c0@Tabletti>
In-Reply-To: <003301c442ee$fafde750$7201a8c0@Tabletti>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

hbaxmann wrote:

> IMHO this is not a security issue at first. 
> If one divide the "security" into authenfication, authorization and
> auditing, then we have a iddentification issue here. The same problem will
> at least arise if one tries to establish something what is called today AOP:
> the 'turn-one-key-opens-all-doors' syndrome.
> 
> I would vote for establishing an identity interceptor as the first in the
> message flow. He is marking the call with the identity of the caller. So one
> is able, even in threadlocal, to identifying who is in.
> 
> absolutely wrong??
> 

If I understand you correctly, I think that is what we are doing, except 
we don't use an actual ThreadLocal, we associate the Subject with the 
thread's AccessControlContext.

The question is where this should happen and how do we ensure it is done 
for all invocations including callbacks.

--
Jeremy