Shell Method™ Software Engineering Process Repository
The revision management process manages the creation and editing of individual configuration items as well as the combination of items into one or more builds. Builds that pass all required reviews and/or tests become stage baselines.
The actors associated with revision management include:
The analysts and/or developers associated with the development team,
The PDR, acting in the role of Configuration Manager, or a specifically assigned Configuration Manager, and
Reviewers and/or testers charged with evaluating stage work products.
Each stage of the SDLC is composed of a standard set of pre-defined stage processes. The stage processes that include revision management activities include:
Informal iteration, where new configuration items are created (often from templates or prototypes), and checked in to the appropriate CM system as defined in the project SPMP.
Formal iteration, where configuration items are finalized for review and/or test, and incorporated into a release build for the stage.
In-stage assessment, where review and/or testing is conducted and the build is either rejected for re-work or accepted as a stage baseline.
The informal iteration process is focused on elicitation and refinement of requirements and project artifacts, resulting in the generation of configuration items that may or may not become candidates for the stage build. For this reason, new items are classified as either “Tentative” or “Real.” This status becomes important in the next stage process.
The Configuration Manager is responsible for setting up the project and stage structure in the CM system, which includes establishing the specifications for the stage build. This process is further described in the vendor’s operating manual for the selected CM system, as defined in the project SPMP.
The formal iteration process is focused on refining and finalizing stage deliverables. The analysts and/or developers involved in the generation of stage deliverables check out, edit, and check in items until they consider the items ready for inclusion in the stage build. The Configuration Manager performs the stage build when all items are flagged as ready for the build. The formal iteration process concludes when the build is successful.
The in-stage assessment process focuses on reviewing and/or testing (as appropriate) the stage deliverables. In the case of reviews, the end result is either a concurring review or a non-concurring review with associated comments. In the case of testing activities, each test case execution results in the generation of either a Test Completion Report (TCR), or a Test Incident Report (TIR).
If one or more non-concurring reviews and/or TIRs are generated, the Configuration Manager evaluates the item and assigns an analyst or developer to correct the problem. The stage reverts to the formal iteration process when the review comments and/or TIRs are passed to the designated analyst or developer. At this point, the build is revised to include one or more of the changed items and the stage deliverables are re-built after the analysts and/or developers have finished revising the appropriate configuration items.
Note: Several builds may be executed before all TIRs are fully resolved. For example, one build may be devoted entirely to resolving a set of related TIRs, while another build may address a single TIR.
If all reviews and/or tests are successful, only Concurring reviews and/or TCRs are generated and passed to the Configuration Manager. Once all reviews and TCRs are in, the Configuration Manager promotes the build to a released stage baseline. At this point, the stage transitions to the Stage Exit process, as defined in the SDLC.