[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Issue 44 proposal amendment
Last week I had requested for more time to think about the impact of the proposal for issue 44 [1]. Currently, how artifacts get resolved within a contribution is under the control of the deployer of the contribution through the export/import mechanism (modulo issue 8 resolution). One impact of the proposed resolution for issue 44 is that intents/policy sets/implementationType/bindingType QNames in contribution A may be resolved in a different contribution B without A having any control over it. This, I don't think is a problem for intents, implementation types and binding types (other than the security issue). But may be a problem for policy sets. The solution to that would be for contributions to use their own distinct target namespace for all the policy sets that they use (or care wrt resolution). I think it would be worthwhile to add a note regarding this. I also think that in addition to saying that the QNames in the definitions files must be unique within the domain we need to say that runtimes MUST generate an error when there is a conflict. -Anish -- [1] http://www.osoa.org/jira/browse/ASSEMBLY-44
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]