Although they are fun, these heated discussions should not be taken too seriously. However, from the tone of the pro-Eiffel postings I get the impression that we are trying to replace C++ with Eiffel. As noble as that goal maybe, I think it is not realistic - C++ is here to stay - instead we should set our goals higher. Eiffel should not replace C++, it should replace COBOL!
First of all COBOL is still being used more than C++ in commercial software development, so, if it replaced COBOL, Eiffel would have much larger presence in the computing world. Moreover, now is our chance, as COBOL's position is changing. Many companies are beginning to move away from mainframe systems to client/server architectures running on UNIX workstations or PCs and are contemplating re-engineering their applications using object-oriented methods. Most of these organizations do not want to fire their entire programming staff and hire a group of young OO gunslingers for reasons not altogether altruistic. The resident COBOL programmers understand the business and they know how the current systems keep their companies running. They may not know what a doubly linked list is, but the know how to compute interest on an overnight repo, or what a vanilla floater is. Without knowing they already understand "business objects" - just listen, they talk about customers, accounts and inventory items.
The central problem faced by the companies moving to OO, is that of teaching of OO and C++ to their COBOL programmers. As it turns out, most articles written on this topic recommend strongly against using C++ as the first OO language. For example, in October '94 issue of JOOP (pg. 55), Lewis J. Pinson writes "One should consider teaching Smalltalk or Eiffel. Let programmers taste success (...) before jumping into C++." But once COBOL programmers learn Eiffel, why force them to learn C++, why not let them loose with Eiffel and see what happens. Unencumbered by the low level details of C++, they will be able to concentrate on solving their business problems promptly, and isn't that what the management wants in the first place?
It is up to all of us, seasoned Eiffel enthusiasts, to bring the Eiffel message to the COBOL programmer and to show how the next generation of business systems can be built using object oriented technology. And of course, deployed systems speak louder than piles of multi-colored presentation transparencies.
Richie Bielak (March 1995)
Writing TOC Last modified: Fri Jun 17 07:55:13 EDT 2005