On Oct 5, 2006, at 11:58 AM, Ignacio Silva-Lepe wrote:
Sorry, I think I misunderstood. You meant update the testcases to
point to the new location for AbstractOutboundInvocationHandler right?
No, the test cases should reside in the same module as the
implementation, in this case SPI.
Don't know why but "move the testcases down to SPI" sounded
different to me at the time I read it. Apologies for the confusion.
I may have not worded it clearly enough too.
Jim
----- Original Message ----- From: "Jim Marino"
<[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Thursday, October 05, 2006 2:39 PM
Subject: Re: axis2 binding "compile" dependency on core
On Oct 5, 2006, at 11:10 AM, Ignacio Silva-Lepe wrote:
Hmm, not sure I follow.
Btw, I'd like to clarify that when I say "we basically agree" I
did not mean that as a forgone conclusion, apologies if I gave
the impression that I was trying to go around the process. I
was basically reporting our conversation on this topic (which
happened to come up on a bug-related discussion). My intent with
my reply was to put our, let's call it preliminary, agreement
out for consideration.
O.K. now I am really confused...I was agreeing that moving the
class down would work (for now, we may want to revisit later as
we refactor things at some point) and you should just go ahead
and do it unless anyone objects. I was also just adding that if
this is done, make sure to move the test cases as well.
Jim
So what is the thought, would promoting the
AbstractOutboundInvocationHandler function (not necessarily the
class line by line) to the spi work?
----- Original Message ----- From: "Jim Marino"
<[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Thursday, October 05, 2006 1:52 PM
Subject: Re: axis2 binding "compile" dependency on core
O.K. We'll need to move the testcases down to SPI as well then.
Jim
On Oct 5, 2006, at 10:31 AM, Ignacio Silva-Lepe wrote:
Hi Jim,
I had a quick conversation with Raymond about this. We
basically agree that we need to promote the function performed
by AbstractOutboundInvocationHandler to the spi as an
extension point, which is also option 2 in my response to him.
This is also what his statement means, ie,
AbstractOutboundInvocationHandler is "performing an
invocation over an interceptor chain", to quote him, and his
proposal is to extract that to the spi.
Thoughts?
----- Original Message ----- From: "Jim Marino"
<[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Thursday, October 05, 2006 1:01 AM
Subject: Re: axis2 binding "compile" dependency on core
On Oct 4, 2006, at 9:58 PM, Raymond Feng wrote:
Hi, Ignacio.
I just found one of your patches added a "compile"
dependency on "core" to "binding.axis2". It's from
"Axis2CallbackInvocationHandler extends
AbstractOutboundInvocationHandler". I don't think it's
appropriate to have binding.axis2 depend on core. Can we fix it?
We had some discussions before to extract the invocation over
interceptor chain into SPI. It seems to be the right way to go.
Hi Raymond, could you explain what you mean by this?
Thanks,
Jim
Thanks,
Raymond
--------------------------------------------------------------------
-
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]
---------------------------------------------------------------------
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]
