axis-java-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Glen Daniels <>
Subject Re: [Axis2] Geronimo integration?
Date Wed, 24 May 2006 15:39:13 GMT
Hi dims, Sanjiva:

Yup - tradeoffs and optimizations.  The Axis SAAJ impl presents itself 
as normal SAAJ to the rest of the world, but internally we use our 
knowledge of Axis-specific APIs in order to get easier code and better 
performance.  This could change, but we'd need to do a serious review 
before considering it.


Davanum Srinivas wrote:
> Q: "Does Axis1 treat its SAAJ impl differently from pure SAAJ"?
> A: YES.
> No clue about how others are doing it.
> -- dims
> On 5/24/06, Sanjiva Weerawarana <> wrote:
>> On Wed, 2006-05-24 at 08:03 -0400, Davanum Srinivas wrote:
>> > Sanjiva,
>> >
>> > Problem that jongjin is referring to is that when you drop our Axis2
>> > saaj impl into Axis1, it will blow up.
>> I'm prolly being thick but why? "Drop our Axis2 SAAJ impl into Axis1"
>> must imply that the person dropping it did the work to integrate it via
>> the right factory and so forth .. and in that case its not obvious to me
>> why it should fail. Does Axis1 treat its SAAJ impl differently from pure
>> SAAJ??
>> Anyway this conversation seems to be moot: If JAX-RPC requires the old
>> pre-DOM SAAJ per Nick's message then we are indeed doomed (not DOOMed ..
>> which would of course imply that we're using the DOOM SAAJ). We can of
>> course impl the old SAAJ stuff using (DO)OM underneath too.
>> I wonder how XFire and Celtix deal with this issue- do either of them
>> support JAX-RPC and JAX-WS simultaneously?
>> Sanjiva.

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

View raw message