AdaIC News Fall 1997


Letter from the AJPO

| The end of the Mandate | NRC report | Reliability Counts |
| Tri-Ada'97 | Fighting misinformation | Another transition |
| Continuing tasks | Thanks | Ada lives |


The end of the Mandate

I continue to applaud Mr. Paige's bold decision to end the Ada Mandate and implement the remaining recommendations of the National Research Council (NRC) report. If you have not yet read the NRC report, I would urge you to do so. You can link to an on-line version via the AdaIC Web page (http://archive.adaic.com/docs/present/nrc/).

The NRC report

The NRC's solid explanation of and support for Ada's reliability and suitability for warfighting software must not be ignored. Ada should be used where it makes engineering sense. In military applications, reliability makes engineering sense. For military systems with twenty-year-plus lifecycles, a maintainable language is essential. Ada will prevail where engineering factors are part of the decision process.

Reliability counts

Make no mistake, war is about killing people and breaking things. Military weapons systems are designed to be lethal and must be reliably controlled. Unreliable military software is frightening. We often hear of Ada's strong typing. An implicit type conversion that results in a 1 degree rounding error will, at a range of 40 kilometers, put ordnance 700 meters off target. In a close combat situation, a 700-meter error can result in friendly casualties. Reliability is important. I don't want to have to tell a Gold Star Mother that her son or daughter was killed by friendly fire due to a software error. The fact that this software was procured using "best commercial practices" and was determined to be "good enough" for military use is likely to be of small comfort. Military software is a life-or-death proposition. People who don't understand this should not be in the business.

I would stress that during my tenure at the AJPO I have never received any engineering-based complaint about Ada. The solid work done by the Ada vendor community has done an effective job of silencing critics who have complained about Ada compilers. One objection that is often parroted is the complaint about "lack of tools." I have been unable to get the complainants to state exactly what "tools" are lacking. The Ada Resource Association has also aggressively sought to find out. The reality is that high-level languages require fewer tools than low-level languages. Another complaint is that there are "no Ada programmers." It is well known that there is currently a serious shortage of skilled software personnel. Any hard-skill computer scientist can pick up Ada very quickly.

Tri-Ada '97

Contrary to many people's perception, Ada is a very easy language to learn. At Tri-Ada '97, hard data will be presented to demonstrate this. A controlled experiment conducted at the United States Military Academy shows that students can go further, faster in Ada than in Pascal, a language specifically designed for educational use.

I would like to encourage everyone in the Ada community to attend Tri-Ada '97. Noted Ada author Norman Cohen has done an outstanding job putting together a very strong technical program. Given the policy changes this year, results from the educational community, and progress in industry, this year promises to be one of the most productive Tri-Ada's ever, covering each of Tri-Ada's three constituent groups: government, industry, and education.

Fighting misinformation

Our biggest challenge comes from the "magazine management crowd" who base what should be engineering decisions on what they read in popular computer trade journals. Unfortunately, some of these journals continue to publish wildly inaccurate information about Ada and the Ada Program. That being said, I would like to thank the reporters who have taken the time to contact the AJPO to check facts or ask questions.

As a former journalist myself and a currently active researcher, I have great respect for journalists who take the time to go to primary sources for their information, and I thank them for their trouble. I also appreciate the many members of the Ada community who have brought these erroneous articles to my attention and have taken it upon themselves to fight some of this misinformation that has grown into a mythology of its own.

Another transition

As of this writing, I am preparing to return to the faculty of the United States Military Academy. Once again, the future of the AJPO is in doubt. Decisions still have to be made regardingthe funding of the NRC recommendations for DOD support to the Ada infrastructure. In this respect, the Ada program is no different than most other Defense programs these days. Given current funding constraints, many critical programs are living a day-to-day existence.

Continuing tasks

In my view, maintaining a consistent, fair, and stable compiler validation process remains one of our highest priorities. Another critical mission is U.S. representation on the various national and international standards bodies. I see these tasks continuing uninterrupted regardless of funding constraints.

Thanks

The Ada community is a great community, and the greatest aspect of this assignment has been interacting with the many selfless Ada volunteers. The government and contract personnel of the AJPO have made this a delightful assignment. I would particularly like to thank Rear Admiral John A. Gauss for his inspiring leadership, keen insights, and for the freedom and support he gave me while serving at the AJPO.

Ada lives

Inside the Beltway, there is no stopping a catchy analogy. Ada is often compared to Betamax VCRs. The analogy being that there is no question of the technical superiority of the Beta format over VHS, yet Beta format VCRs failed to capture the consumer market. Like many analogies, this does not go far enough. In fact, Beta format VCRs continue to be used by video professionals specifically because of the technical superiority of the Beta format. Just because Blockbuster doesn't stock Beta format videotapes doesn't impact on the engineering decisions made by video professionals.

Military software is not the same as consumer software. You cannot buy "off-the-shelf" fire-control systems. Ada was designed to build software to military specifications. Where reliability counts and where software engineering considerations factor into project management, Ada will continue to thrive.

John A. Hamilton, Jr.

Lieutenant Colonel, U.S. Army
Chief, Ada Joint Program Office


Previous Page Contents Next Page