IBM Mainframe Discussion List <[email protected]> wrote on 12/16/2008 09:57:39 PM:
> >... the mid 1990s, and somewhere around then production > >of PLMs (internal and external) ceased in order to reduce > >development costs. > > Uh, reduce cost by eliminating Program Logic doc? I assume the PLM > was replaced by something else for internal use. Eliminating doc for > program logic seems like a good way to also eliminate program logic. > PLMs were not internally replaced by any coherant strategy. Some components maintained theirs for a while in TSO data sets, other components did other things (component workbooks, whatever). Others did nothing. Now with worries about an aging carbon-based life form knowledge base, and Wikis being all the rage, now and then there is some excitmemnt about trying to organize that kind of information for internal use. A lot of labor went into writing, maintaining, and producing PLMs, so eliminating them of course immediately reduced some costs. Now, whether or not the benefits of PLMs provided a positive return on investment, and more was lost than was saved by eliminating them, I don't know how to measure that. I am just trying to relate a little of the history as I recall it. > BTW, I vaguely recall the before the external PLMs disappeared, > some of them had changed format, with diagrams of logic flow being > replaced with something else. "HIPO" diagrams or something like > that? That helped lessen the blow of the PLM's demise. First > replace it with something useless; then ease the pain by removing > the PLM altogether. [insert a "scarcasty" here if you have one.] > Was I the only one that did not like that change? For assembler programs, there was program called FL1 or some such thing that produced flowcharts from special block comment comments imbedded in the code. That is why you would see comments like */* P RESET SVC FLIH SUPER BIT, RESTORE NORMAL FRR STACK */@G381PXU */* P ENABLE PSW FOR I/O & EXTERNAL INTERRUPTS */@G381PXU */* D (NO,SVCENTPT,YES,CMSCK) ANY MORE LOCKS */@G381PXU I never coded this stuff myself - I would guess that P produced a "process" box, and D produced a "decision" box. This was already falling out of use when I got here in 1979, and the amount of assembler source was continuing to diminish in favor of PL/whatever. HIPOs were becoming the rage in the IBM internal programming classes for new hires, and I probably still have a green plastic HIPO template (similar to to old plastic flowchart templates) stuffed in a drawer somewhere. There was a program called "the logic tool" that produced HIPO-ish stuff for PLMs from block comments in PL/whatever source code. That is why you may have seen block comments starting with Title: or Logic: , and line comments starting with L: . Some development groups were very much into this stuff, others were not. The PIG (Processor Interface Group, that owned MVS reconfiguration, SADMP, machine check handler, etc in the mid 1980s) loved this stuff, and would develop new modules by first writing the block comments, and then hold design reviews in their meeting room (the PIG pen) of the logic tool output, with great attention to detail on the esthetic beauty of the results. Of course, the ratio of amount of code to programmers was drastically lower in those days. Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
