Steve, I too have taught courses in 'modern COBOL' in client shops. I have stopped doing so. What I accomplished was to turn the brightest younger programmers into disaffected employees because their managers were unsupportive. Some of them were, I suspect, already disaffected; but I was blamed for their departures when they sought more interesting employment.
The approach I suggested is, I think, a better one. Teach COBOL to some established, competent C or assembly-language programmers. This can be done in three weeks iff, to use a wonderful word I recently encountered, they are properly 'incentivated'. It does need a different approach. With such programmers, one need only provide answers to the classic three questions: o What are the data types? o What operations can be performed on them? o How is the path of control among these operations specified? emphasizing branch tables, iteration, and recursion, all of which are now possible in COBOL, in answering this last question. Then one provides a lot of competitive, ruthlessly criticized practice in writing COBOL subroutines for which you have test bed/drivers in place. The culture of most COBOL shops is so complacent that disruptive technology is very hard to teach in them. It is not, I am sure, impossible; but it is almost certainly uneconomic. . John Gilmore, Ashland, MA 01721 - USA ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
