>>On Fri, 2006-12-01 at 13:56 -0600, Ann Robinson wrote: >>> Hi, Sanjiva, >>> The proposal could be committed with little risk to the rest of Axis2. >>> The save/restore of a message context is not an automatic action and >>> so it would not affect current message handling. >> >>Yes but the code that's committed will not work according to Axis2 works >>and anyone else using it will find that out the hard way. >> >>That's not an acceptable commit right? >> >>Sanjiva. > >On Mon, 2006-12-11, Ann Robinson wrote: > Sanjiva, > > Committing the code as-is will not disrupt anyone else. The function has to > be explicitly called. I expect to provide an update > to address your concerns over the ServiceContext fairly soon, so I expect that the > window in which someone could use it and hit that problem will be short. In this > respect, this is no different from someone using Axis2 and encountering a problem > and opening a JIRA issue. > > If finding someone to do the commit is a problem I would suggest Bill Nagy, who > already has some familiarity with the message context save/restore code. > > Ann >
Sanjiva, This proposal has been available for review and comment since 2006-10-31. I know the shutdown for Axis2 1.1 release took a lot of attention away from other items such as this. However, the prototype for this proposal has been tested and used internally so the prototype has reached a level of stability. I think the risk of introducing problems to the mainstream message processing is very minimal. This proposal does not change Axis2 architecture but instead deals with the existing structure of the Axis2 message context object graph and message processing. I would really like to get this proposal at a point where it could be included in the Axis2 1.2 release. I am currently working on a solution for the point you made regarding the ServiceContext and I will provide an updated patch. What are the concerns/issues that are blocking the acceptance of this proposal? Ann
