I'm not as fluently in speaking english as John Gilmore is, of course, because I'm not an english native speaker, but that's exactly what I would have answered to Mr. William Blair, too, if I could.
Of course, in my classes, which last 3 or 4 days, I only can teach the very basis of z/OS, which are necessary to start with dump reading, and the experience has to come from practicing. The participants who really want to gain experience start reading all the dumps from their department, which consists of some 50 developers - that is: a dump every day. And there are some strange situations, when programs write over the code of other programs etc. and the save areas are all zeroed out - nice exercises. Thanks, Bernd Am 26.07.2013 14:42, schrieb John Gilmore:
William Blair's post is in most respects a superb one. There is, however, an important omissis. I already had IBSYS dump-reading skills; I learned to read OS dumps beginning with OS/PCP, i.e., in a radically simpler world; I was able to come to terms with increasing complexity incrementally, over many years; and I had others to talk to about my perplexities. 'Learn it alone and unaided!' is now a counsel of despair. If you want to learn these skills you need to apprentice yourself, at least part-time, to someone who already has them. Find a situation in which you can consult him or her several times a day, and make sure that he or she has something of a vocation for teaching. I have found it possible to impart these skills to two inexperienced but admittedly very able teenagers using such an apprenticeship scheme. In doing so I avoided introducing them to dump-reading aids (other than hexadecimal calculators) until very late, but I did give much attention to the many ways in which all IBM systems are similar, to frequently encountered control-block chaining schemes, stacks, pools, queues and unholy, impure mixtures of them. I have not compared notes with anyone else who has imparted/taught these skills successfully; but there are certainly other equally viable ways to do so that differ in substantial detail from mine. I suspect, however, that the availability of a mentor or mentors is crucial. They figure centrally in other professional training schemes, be they for surgeons or steamfitters. Exclusively do-it-yourself schemes will fail; homo sapiens is too short-lived a species; and preternatural dedication must anyway not be expected or required. John Gilmore, Ashland, MA 01721 - USA
