> -----Ursprungligt meddelande----- > Från: IBM Mainframe Discussion List > [mailto:[EMAIL PROTECTED] För Steve Comstock > Skickat: den 14 november 2007 17:47 > Till: [email protected] > Ämne: Re: SV: SV: SV: COBOL COPY statement w REPLACING... > > Thomas Berg wrote: > >>-----Ursprungligt meddelande----- > >>Från: IBM Mainframe Discussion List > >>[mailto:[EMAIL PROTECTED] För Howard Brazee > >>Skickat: den 14 november 2007 17:34 > >>Till: [email protected] > >>Ämne: Re: SV: SV: COBOL COPY statement w REPLACING... ... > > > > Old programs that are recompiled get the "new" copys/versions. > > And fails in compilation if they contains the ":tag:" format > > (if they - as I mentioned - are not using REPLACING). > > And that is not acceptable. > > > > Thomas > > You keep saying that ("that is not acceptable"), and I keep > wondering: why not? Sometimes things must change. IBM has > done a pretty amazing job in supporting old code with no > changes; code that was written and linked over 30 years > ago will likely still run today with out being re-compiled > and re-bound. Try that on Windows or UNIX. > > But if you want to use new features, you likely need to > re-write, re-compile, re-bind. With some careful planning > you can produce code that will run without more modification > (unless you want to use more new features) for 30+ more > years. > > But nothing is free. I thought Bill Klein's suggestion was > pretty clever and workable. But obviously we are not seeing > the line between "acceptable" and "not acceptable" in the > same way your user is.
We (the project) can't force other parts of the organization (that is not part of the project) to do work that for which they have no budget. (That is: change, compile, *test* and update the production.) And if a programmer is called in in the middle of the night to fix a production problem and is seriously delayed by the compilation error we will be hanged by the CIO or CEO. (Again, the availability of the production applications goes before *anything*.) The problem we tried to solve was that we have to either make the "new" copys conforming to our current standard that is needed for (among other things) to work with Datamanager *OR* make them compatible with the old programs. (We had a preprocessor - MetaCOBOL - earlier that have a COPY REPLACING functionality that would have solved our problem.) My initial point when I started this thread was that if You make an effort to create a functionality, why doing it badly ? Thomas _____________________________________________________________ Thomas Berg Specialist IT Utveckling Swedbank AB (Publ) ---------------------------------------------------------------------- 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
