Search

Ada offers a great deal to the MIS community. There are a number of language features that provide attractive tools to the programmer. Ada's support for Object Orientation is far superior to that currently afforded by IBM mainframe languages such as COBOL and PL/1. This translates into more maintainable code, which is the key to MIS Ada acceptance.Unfortunately, Ada introduces problems that do not currently exist in MIS. Decimal arithmetic and string manipulation adequate to support reporting are inherently non-existent. I/O support is inadequate as are DBMS bindings. Conversions are forced on non-ASCII hosts. The compiler has the option of reordering fields within records, rendering non-Ada utilities useless. There is an inexcusable lack of understanding of the targeted market.Virtually all of the known weaknesses of Ada were to be solvable by standard packages to be marketed by compiler vendors as a result of “market forces”. This is how compiler vendors would differentiate their products. This strategy has failed and no new strategy has replaced or augmented it.The Ada 9X Language Revision Process does offer some hope. Native character sets, decimal arithmetic and what is loosely called “secondary standards” offer the potential to make MIS Ada acceptable without burdening other disciplines.Ada 9X doesn't have all the answers. Bindings are still a significant problem. This is especially true if there is interest in converting any of the 80 billion lines of COBOL code to Ada. While the Ada-SQL binding will solve the Ada-SQL problem, the binding of the SQL specification to the proprietary DBMS is still left to those mysterious “market forces”.In my opinion, the Ada community needs MIS support to impact “market forces”. The Ada community has not demonstrated an understanding of MIS requirements and therefore cannot market to the MIS community. There is no MIS inclination to even acknowledge Ada, therefore any change to the status quo must come from the Ada community. The Ada 9X Language Revision Team has started to address this problem.There may be an MIS Ada awareness in the 1990's, but bringing Ada into the MIS mainstream is more than a decade away.

The problems associated with asking a COBOL1 programmer to learn Ada can be likened to the problems you might anticipate when an American employee is asked to learn Japanese. Significant motivation must be provided to overcome some of the obstacles. First, learning Japanese is difficult. The Japanese and English languages have no common ancestry. Second, societal values vary. Third, it is difficult to generate much interest in learning anything when its anticipated practical use is minimal.Of course, with proper motivation, anyone can learn either Japanese or Ada. In either case, the anticipated benefit must exceed the anticipated learning effort. An anticipated transfer to Tokyo could conceivably spur some interest in learning Japanese. But what could possibly motivate the COBOL programmer to learn Ada?