Peter Farley astutely points out: >It seems to me that the first reason for the PDSE requirement >was their choice to use the Java backend code generator....
Precisely. OK, let's summarize: 1. There is always a migration effort to upgrade anything, anywhere. It will cost millions or billions of dollars for the iPhone community to upgrade from iOS 6 to iOS 7 starting next week, yet it will be done. And for most people most of the time it'll be a smooth, low cost, value-adding experience because a lot of technical people did a lot of work. 2. Enterprise COBOL 5.1 has a PDSE prerequisite for important technical reasons which Peter Farley has helpfully explained. (Thanks, Peter.) 3. You may have a migration effort to get to PDSE. If so, there will be some cost/effort. I don't remember characterizing the size of that cost/effort except that it's far from the biggest in history. (Y2K? z/Architecture? VS COBOL II? Sysplex?) I do remember recommending an assessment of that cost/effort via a trial (for example). 4. The migration effort must be assessed along with the benefits -- and wow there are many of the latter. What's the continuing cost to run XX% less efficient code than now available? John Gilmore opines: >I have made no secret of my view that the management of too many >mainframe shops is compulsively risk-averse, suspiciously unanimous in >its rejection of innovation, in a word, reactionary. >These attitudes are destroying the mainframe.... I hope that's not going on here, this time. The existence of a cost/effort is not a sufficient justification for inertia and inaction -- agreed, John. I humbly suggest we now, constructively, focus on what's involved in moving to PDSE (if you haven't already), techniques for reducing the costs/risks/effort to move, and develop advice (and/or point to existing advice) on how to get the job done as quickly and efficiently as possible. We've got a prerequisite, yes. We've also got a fantastic new compiler, hurray! And if there's something IBM could or should do better to make PDSE better, yes, please let IBM know (the official ways). But don't wait for IBM unless that's the only business-justified choice available. Who's delivering friendly, responsive customer service? Who's delivering relevant, valuable innovation? Are you? Or is somebody else? *Somebody* will satisfy user and business demands. How about this community? How about mainframes *and the talented people who operate them*? Let's roll up our collective sleeves, figure this out, and get the job done, OK? Because the folks upgrading 10,000 blade servers from Windows Server 2003 to Windows Server 2012 are working, and they don't even have coexistence/fallback available (to pick an example). -------------------------------------------------------------------------------------------------------- Timothy Sipples GMU VCT Architect Executive (Based in Singapore) E-Mail: [email protected] ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
