Conversely, the organization could continue with its own
modeling language, substantially weakened by the lack of computer support, or even
build its own CASE tool, a heavy investment in a venture in which it had no
experience, and could often only make ?¬?nancially feasible by attempting to sell the
resulting tool to others.
These con?¬‚icts, of course, did nothing to improve the image or practical bene?¬?t of
the expected CASE revolution (Yourdon, 1986). CASE had been unrealistically
trumpeted to be the ???silver bullet??? that would solve all software development
problems. All too often, the response to the failure of CASE to provide such a solution
was to blame the modeling language or the tool. This of course led to the development
of yet more languages and tools: hardly likely to improve the situation.
14.2.2 1990s Overview
As we have observed, building a whole modeling tool from scratch is a large and
complicated project, signi?¬?cantly too slow to keep pace with modeling language
development. The solution to this conundrum thus lies in a tool that can be customized
to support any modeling language. Two possible approaches to this were perceivable
in 1990: effectively our levels 2 and 5 above. A modeling tool could be designed and
built modularly, so that the minimum coding effort was required to change the part
concerned with a particular modeling language.
Pages:
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685