home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!pnl-oracle!lovie!d3g325
- From: d3g325@lovie.oname (DT McGowan)
- Newsgroups: comp.software-eng
- Subject: DoD2167A/Objects/Ada
- Message-ID: <1992Jul29.191938.7760@oracle.pnl.gov>
- Date: 29 Jul 92 19:19:38 GMT
- Article-I.D.: oracle.1992Jul29.191938.7760
- Sender: news@oracle.pnl.gov
- Reply-To: d3g325@lovie.oname
- Followup-To: comp.object
- Organization: Sun Microsystems
- Lines: 40
-
-
- Here is the dilemma:
-
- Given:
- Object Oriented project constrained to implementation in Ada requiring
- DoD-STD-2167A documentation (yuck!)
-
- Question:
- Any recommendations for what Ada construct is mapped to Computer Software
- Components (CSCs) and Computer Software Units (CSUs)?
- Also, what recommendations would you have for tailoring 2167A (SDD, IDD
- and backward to the SRS, IRS) in order to facilitate this view of CSCs and CSUs?
-
- Position-to-date:
- Gleening info from "Experiences in Preparing a DoD-STD-2167A Software Design Document
- for an Ada Project" (C.A. Meyer, et. al. 1989 ACM publication) we are leaning toward
- having CSCs be wholly logical entities (not mapping to any physical Ada construct) and
- having CSUs represent the lowest level of object present in our object diagrams. Since
- our definition of an object includes operations performed by that object, the Ada
- construct associated with the current choice for a CSU would be a package or a task.
- Another example of a CSU would be an object class that is utilized by at least one
- other additional object. This class would then map to the physical representation
- of a generic package in Ada.
-
- This position would require a tailoring of the SDD to redefine "separately testable"
- in a far more liberal fashion. Also required would be the removal of "execution control"
- from the descriptions of CSCs and would have to explicitly call out components of the CSU
- when presenting detailed descriptions of data elements,PDL, etc..
-
- Any comments/suggestions regarding the "dilemma" and/or the position-to-date
- presented would be greatly appreciated. Thanks.
-
-
- Dan McGowan
- d3g325@lovie.oname (DT McGowan)
-
- Research Scientist
- Battelle
- Pacific Northwest Laboratories
- Richland, WA 99352
-