axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "David Illsley" <>
Subject Re: [AXIS2] Adding security phase to OutFaultFlow
Date Mon, 17 Dec 2007 11:35:42 GMT
I disagree. It complicates the chain and our configuration files for
no benefit to a large number of users.

The lack of dynamic phases also requires people to repetitively edit
axis2.xml files when adding something we haven't yet added a phase
for.... it's really naff from a user perspective, especially given all
the huffing and puffing in articles about how great our module
architecture is.

Can I ask if there's something specific you're worried about wrt
dynamic phases, or if it's just something you don't see a need for?


On Dec 17, 2007 9:25 AM, Deepal jayasinghe <> wrote:
> Eran Chinthaka wrote:
> > I also agree with David here.
> >
> > Nandana, can you please explain why you need to hard code security
> > phases in to axis2.xml? Does this mean we need to have n number of
> > phases if Axis2 is supporting n WS-* spcs?
> Yes for the most commonally used module like Addressing , Sandesha and
> Security.
> > If that is the case then I feel there is something wrong with Axis2
> > architecture itself.
> :)
> >
> >
> > I like/prefer to keep Axis2 to the minimum. I for one, and believe
> > most of other users, do not bother to use security at this time. I can
> > quote you very large scientific systems working with Axis2 without
> > WS-Security but with just transport level security.
> Yes I do agree with you , but just having security phase in the chain
> will do nothing unless you have handlers in the phase.
> > Why do we wanna have some set of phases, which I never need, hard
> > coded in to the system.
> >
> > I know, our security guru (you know who he is ;) ) will tell me that
> > it is not a harm to add one more phase in to Axis2. But why do we
> > wanna hack the system, without properly implementing it with dynamic
> > phases or whatever it is.
> I strongly believed and believe dynamic phase rules  are more than what
> we want , its better if we can find a simple solution than that.
> -Deepal
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

David Illsley - IBM Web Services Development

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message