Re: [stp-dev] Idea on Integrating a New Policy tool for GSoc'09'

[Oisin Hurley <oisin.hurley@xxxxxxxxx>]

Just to clarify this bit...

> About UI part for the policy derivation tool , File>> Import >> Policies
> ,Kind of thing would work really fine rather than keeping a separate UI for
> the component and redirecting it to the Editor. But, I didn't really get ,
>>>or there may be an advanced way that would allow them
>>>to be redirected elsewhere
>  Is it like feeding the  policy instance into other unopened project? or
> another related STP component ?

What I was thinking about here was an extensible way to 'write out'
the WS-Policy models that the derivation step has produced.  One
way to do it is to write each WS-Policy document to a file. You might
also want to write it to a database, for example.

This springs from my abstract thinking of the operation of the policy
derivation. First, there is a thing which produces a series of messages
for analysis. Second, there is a thing which builds one or more standard
model representations of the policies. Third, there is a thing which
iterates over the model to transform it to a durable form. In my mind,
each 'thing' is represented by an extension point. That's the architecture
and API part, and the first point of the design. Then, you create exemplar
extensions using these extension points, and see how it all works. For
example, you could have a 'produce messages from a directory of
files', 'build policy models from SOAP messages' and 'write policy
models to XML files' implementation as the primary goal. Does that
make sense?

> Regarding Standards support.. I agree with the UI aspects you have
> mentioned.As we progress i'm sure we may come up with lot of other options
> as well.  I will go through XEF examples you have mentioned asap. Seems like
> i will have lot's of points clear up when i get through it.

I'm always keen for feedback on XEF, because I think it's cool :)
Suggestions for improvement and new features and patches
very welcome.

> <deleted> However i didn't have a clear idea about da second one -
> exensible generation framework..Is it some kind of a runtime configurator
> for Policy Editor that can be configured using set of  custom made policies
> ?

That's pretty much it :)  I helped construct one of those before -
basically you get a collection of policy models and behind the
scenes there is a scriptable generation framework that allows
you to construct runtime configurations and that kind of thing.
It may be a bit out-of-date these days what with the evolution
of runtimes that can consume WS-policy instructions directly.

You'll note though, that this generation approach would be an
exemplar for the third extension point I mention above.

> <deleted> I hope we could discuss more regarding
> this and come up with a good action plan .

I'm happy to have more discussion on the topic.

>In the mean time i will try to
> familiarize things with Eclipse and Policy Editor  stuff   from now on ,so
> it will be lots more easier when the actual work happens.

Very cool. Let us know your feedback!

 cheers
  --oh