<service> --> <interface>
<extend-service> --> <implementation>

I like that, except that <service> is a "super-element" of <implementation> and I want 
to keep it
that way, since that's the normal usage.  That is, you can define a service and 
provide the
implementation (core service impl plus interceptors) all in one go.  And that loops me 
back to
<service-point> and <extend-service>.  <contribute-service>?

I think many people only give lip-service to OO ... they still think in terms of 
massive, central
"brain" objects with a bunch of helpers attached. I tend to think in terms of lots of 
smaller
objects, peers, working together ... and HiveMind lets me structure it all that way.

--
Howard M. Lewis Ship
Creator, Tapestry: Java Web Components
http://jakarta.apache.org/tapestry
http://jakarta.apache.org/commons/sandbox/hivemind/
http://javatapestry.blogspot.com

> -----Original Message-----
> From: Bill Lear [mailto:[EMAIL PROTECTED] 
> Sent: Wednesday, September 17, 2003 10:34 AM
> To: Jakarta Commons Developers List
> Subject: RE: [HiveMind] Extension to adder example
> 
> 
> On Wednesday, September 17, 2003 at 10:08:14 (-0400) Howard 
> M. Lewis Ship writes:
> >...
> >Yep, that's the whole point ... build the two jars and you 
> have Adder 
> >and Subtractor ready to go.  HiveMind will locate and parse all the 
> >hivemodule.xml's into one consistent registry.  If you get 
> >hivemind.examples.Adder it will combine the interface from the first 
> >jar with the implementaion contributed by the second jar.
> >
> >So ... obviously I've been failing on getting the message out about 
> >what HiveMind does if this is coming as a surprise now.
> 
> I think the issue is very simple: programming in this 
> paradigm is a radical departure for us not used to it.  I'm 
> not used to gluing things together with such little effort.  
> Not only are the concepts different, but, as one might 
> expect, the language one uses to convey the concepts is 
> altogether new.  In such cases, it's not so much a question 
> of being clear, as providing concrete examples that cover the 
> initial case, and then the initial case plus one --- sort of 
> teaching by induction, if you will (prove case 1, then prove 
> if case N is true, case N+1 is true, etc.).
> 
> So, I don't think you have missed wildly (failing); I think 
> that the examples could encompass this (and, critically, I 
> think examples need to be *structured*, starting from simple 
> to simple + 1); and, finally, I think a tweak to the 
> language, as we have discussed, would also help.
> 
> By having structured examples, it allows intellectual 
> movement from simple cases to those beyond, which is what 
> learning new things like this is all about.
> 
> [Just a thought: why "service", "service point"?  I realize 
> that "service" is a useful term, but why not "interface" and 
> "implementation"?  Those terms seem to be broad enough and 
> easy to understand...]
> 
> 
> Bill
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 


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

Reply via email to