On Wed, 18 Nov 2009 12:59:17 -0500, Don Williams wrote: > >So if many decades ago, IBM would have asked me (LOL), I would have >suggested an option like READONLY, RO, PROTECTED, or something along those
My understanding is that is what REFR was intended to mean. (Etym.: If a memory page physically failed, that page could be REFReshed from a pristine copy, perhaps in a page data set.) >lines to cause a program to be loaded in protected storage. Then the RENT >option would mean that multiple instances of a program would be allowed >regardless of it READONLY/NONREADONLY attribute. For READONLY programs, all >execution instances could use the same physical copy; for NONREADONLY >programs, each instance would need its own physical copy (since each >instance might alter itself). For the NORENT option, only one execution >instance would be allowed regardless of its READONLY/NONREADONLY attribute. > I agree with your analysis: REFR and RENT should be independent. One can contrive scenarios to populate each quadrant of the truth table. IBM, with binder, however, has made the definitions hierarchial: REFR implies RENT, and RENT implies REUS. -- gil ---------------------------------------------------------------------- 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
