April 22, 1997
by Magnus Kempe
Here are answers to some questions we raised after the announcement by Secretary Paige of the end of the US DoD's Ada-only policy. The answers were provided by AJPO Director LTC Drew Hamilton.
Q: If the DoD abandons its policy known as the "Ada mandate", what are the available and chosen alternatives to ensure that the problems (language proliferation and high maintenance costs) that led to the creation and adoption of Ada do not reappear?
A: The policy changes being contemplated do not constitute abandonment of Ada. Programming language choice is not the only factor that leads to high maintenance costs. As for program language proliferation, the NRC Report specifically addresses non-standard, proprietary languages. It will be very difficult to justify the use of a non-standard, proprietary language during an SEPR.
Q: How soon will the effects of this decision be visible?
A: Translating the NRC Recommendations into appropriate policy for an organization as large as the Department of Defense is a non-trivial task and will take some time to develop and implement.
Q: What will be the consequences for DoD-supported activities that focused on Ada so far? For instance, will the AJPO be closed down, as had previosuly been scheduled for June 1997? If so, when?
A: As noted in the Winter 1997 issue of AdaIC News, the AJPO will remain open through the end of FY 97. With the implementation of the NRC Recommendations, it is likely that the AJPO will remain open for some time.
For other DOD activities, the technical merits of Ada have not changed. Training activities that use Ada are likely to continue to do so. The pedagogical advantages of Ada are well known and documented. Dropping the mandate should have no effect here.
For projects already successfully using Ada, there is no reason to change languages because of any potential policy changes. There has never been a requirement to use Ada where it did not make engineering sense, that is why there was a waiver process in the first place.
Q: How will the DoD's decision be perceived?
A: That is unknown and unknowable. There are many Ada advocates who have believed that Ada should be selected based upon its technical merits. Some PMs and engineers have taken the view that dropping the mandate merely shows confidence that where a rational software engineering process exists, Ada95 can win the engineering arguments.
Q: Has the mandate served its purpose?
A: The mandate successfully focussed attention on some very real programming language issues. The Ada mandate focused attention on (to quote Jean Sammet's famous paper) "Why Ada is not just another programming language." One very real outcome of the whole Ada effort is that many of the language features pioneered in Ada are now appearing in other commercial languages.
Q: Are there new software development problems that call for a change in DoD policy?
A: We know more about software engineering than we did fifteen years ago. Language selection is but one important engineering decision in software development. It does not make sense to focus solely on language selection and ignore other critical software engineering decisions.
Q: Is it a change of tactics, or will Ada remain a preferred language in the coming years?
A: While from a software maintenance perspective particularly, commercial languages still have not caught up to Ada 95, the trend of Ada-like features migrating into other commercial languages is a good one and one we hope continues. It is not the intent of the DOD to be in the programming language business. We are in that business because currently commercial programming languages do not fully address our often unique requirements for high assurance, maintainable software development.
Page last modified: 1997-04-22