home *** CD-ROM | disk | FTP | other *** search
-
-
-
-
-
- PROCEEDINGS FROM THE GRAPHIC ADA DESIGNER
- CRITICAL DESIGN REVIEW
-
- The Critical Design Review (CDR) for the Graphic Ada Designer
- (GAD) was conducted 22-23 August 1985 at SYSCON Corporation, San
- Diego Division. Attendees to the CDR were as follows:
-
- Maj. T.Courtwright - WIS/JMPO
- Maj. J.Maxwell - WIS/JMPO
- Capt. J.Choiniere - WIS/JMPO
- Mr. G.Hawkins - FBI
- Mr. A. Brintzenhoff - Mgr. Ada Technology Group,
- SYSCON Corporation, SD
- Mr. S.Christensen - GAD Project Manager
- Mr. J.Reddan - GAD Technical Manager
- Mr. B.Marek - GAD Technical Staff
- Mr. J.Baker - GAD Technical Staff
- Mr. J.Long - GAD Technical Staff
-
- The agenda for the CDR and chronology of the presentation is
- given below as follows:
-
- Introduction A Brintzenhoff
- Project Overview S Christensen
- Design Objectives and Requirements S Christensen
- Macroscopic Design Updates S Christensen
- Microscopic Design Overview J Reddan
- Comments On Documentation J Reddan
- Hands On Demonstration
- Project Issues J Reddan
- Informal Discussion
-
- The questions, issues and discussion that arose during the
- presentation are presented in the sections below which
- are Critical Design Review Comments, Documentation Review Comments,
- WIS JPMO Action items.
-
-
-
-
-
- CRITICAL DESIGN REVIEW COMMENTS
-
- The following topics were discussed during the formal
- presentation and hands-on demonstration.
-
- 1. Virtual Package Concept - This conceptual entity provides a
- method of expressing high-level architectural constructs in the
- design. The software engineering principals of deferring
- architectural details, and promoting abstraction and information
- hiding are provided by the concept. The virtual package entity
- is consistent with emerging concepts in the literature such as
- SUBSYSTEMS and ENCLOSING_ENTITIES. (Reference MIL-STD 2167
- concerning the implications for Virtual Packages.)
-
- 2. Project Statistics - The WIS/JMPO representatives all
- indicated it would be extremely useful to gather project related
- statistics (time/effort in the various phases).
-
- 3. Limitations of the DEC Ada Compiler - It was pointed out by
- Maj. Courtwright that the DEC Compiler does not allow single
- character input from the terminal without a concluding <CR>.
- Also, there are significant differences between Versions 4.1 &
- 4.2 of the DEC Ada Compiler in multiple areas. Any problenms or
- issues will be addressed during the rehosting to the DEC Ada
- compiler.
-
- 4. Imported Entities - The question was raised if imported
- entities should be retained on a design diagram if there are no
- data or call connections to them.
-
- 5. Generic Instantiations - The question was raised on how to
- handle generic instantiations (eg. [1] split (on same line with
- instantiation name) or [2] show place holder and access PDL via
- editor).
-
- 6. Seperate Compilation (Subunits) - The question of support for
- seperate compilation (subunits) was raised. The current design
- does not support a separate compilation attribute. However, an
- "is seperate" statement is used to provide compilable PDL for
- exports which are not connected to actual entities.
-
- 7. MMI Flow Problem - There is a general problem with forcing
- the user to return to the Main Menu when errors occur (there
- should be an alternate way of exiting, to present a menu or
- giving choices via mouse buttons).
-
- 8. Items 3, 4, 5, 6 and 7 above will be assessed during SYSCONs
- internal evaluation of the GAD; changes, where appropriate,
- will be made prior to the evaluation in which WIS JPMO personnel
- will participate.
-
-
-
-
-
- DOCUMENTATION REVIEW COMMENTS
-
- The following comments were received concerning the GAD User's
- Manual, Design Document and Microscopic Design listings.
-
- 1. User's Manual
- a. Maj Courtwright - Needs to include/expand glossary of
- very specific terminology that is utilized concerning
- the tool.
- b. Maj Maxwell - The document failed to communicate to the
- user how to navigate the MMI Menu Network. A step-by-
- step (CAI) scenario is required to provide the necessary
- sequencing and interaction information. Additionally,
- the entire menu structure needs to be included in the
- USER's Manual.
- c. Capt Choinere - The organization of the document was
- satisfactory, however, it needs to include more detail and
- an example and subsequent scenario to illustrate the use
- of the tool.
- 2. Design Document
- a. All references to the earlier Harris GKS implementations
- will be omitted.
- b. Adhere more to the GKS terminology (ie COLOUR vs COLOR).
- c. The organization and content was generally satisfactory.
- 3. Microscopic Design
- a. Areas of the listings where TeleSoft Compiler
- Version 1.5 idiosynchroncies occur should be annotated
- for quick and easy identification for possible future
- modification to more appropriate Ada syntactical forms.
- b. Although their occurances were not numerous, some
- instances of criptic names did appear (ie CONN_PTR vs
- CONNECTION_POINTER). These occurances should be
- renamed.
- c. The occasional use of renames caused some confusion.
- This should be avoided whenever possible.
- d. The menu package should be more "generic" in nature to
- allow the use of a wide range of general icon
- conventions for the Menu Window.
- e. When the design support package dependency is removed,
- all occurances of TBD types in the resulting PDL
- generated from a design diagram should cause an error
- when compiled. At present, this is not the case (ref.
- enumeration type TBD).
-
-
-
-
-
- 4. Tool Design Action Items - As per Maj. Courtwrights request,
- the following summarizes the action items taken on the
- functionality and design of the GAD tool.
-
- a. GKS_Prime vs GKS - Use a GKS Interface as near to the
- ANSI standard as possible. The ability to accomplish
- this is dependent upon the availability of a DEC Ada
- compiler prior to delivery of the tool. The existing
- terminal driver interface and specific driver software
- will be utilized. (SYSCON received its Ada compiler on 3
- September and will have it operating by 6 September. The
- exact sequence for transition to the compiler will be
- determined after reviewing other GAD factors during the
- week of 9 September.)
- b. Provide a Comment Insertion Capability - A feature
- should be added which will permit the user to insert
- approximately 1-3 lines of text, which will be treated as
- Ada comments, into each package (specification and body) and
- subprogram and task bodies when the PDL is generated. The
- text will be editable via the same editing capabilities
- which are used for editing identifiers. The text will not
- be dislplayed on the object oriented design diagrams;
- however, it will be displayable in the prompt/status/error
- area of the CRT for observation or editing.
- c. Complete the implementation of major functionality.
- This includes the FILE, PRINT and MOVE (in scope move
- only) functions. For the movement of entities outside
- the scope of an enclosing entity, at a minimum the
- project is to identify the problems and issues with
- achieving this capability.
- d. New functionality - It is extremely important that a
- RESIZE or RESCALE function be added to assist tool users
- with improving the asthetics of a design diagram without
- having to delete and redraw incorrectly sized entities.
- A REPLICATE function was deemed not critical to the
- success of the GAD tool.
-
- WIS JPMO ACTION ITEMS
-
- 1. Mjr. T. Courtwright will a copy of the ANSI GKS Standard
- to SYSCON.
-
- 2. Mjr. T. Courtwright will be responsible for locating a
- suitable "alpha" test organization for early use of the GAD
- during the development process. In selecting an organization, it
- was mutually agreed by both SYSCON and Mjr. Courtwright that the
- organization selected must be acceptable to both the WIS JPMO and
- to SYSCON. In addition, it was agreed that proper controls would
- need to be placed on the organizational personnel involved and be
- properly enforced to ensure that the evaluation was meaningful
- and that due account would be taken of the fact the the tool was
- still undergoing development and subject to both refinement,
- enhancements and changes.
-
- 3. Mjr. Courtwright agreed to authorize expansion of SYSCON's
- directory on the TECR machine at NOSC to approximately
- 10000-15000 blocks.
-