<...snip...> Doesn't each STARTBR tie up a VSAM string on the file? Probably, but wouldn't it only be one string per *FILE*? So I would only tie up seven or so, even if I did issue hundreds of STARTBR?
It's been a long time, but I remember that we had problems of this sort in the past. I'm not the CICS person anymore. In any case, that is the only thing that I can think of that might be a problem. Well, other than "elegance" in the code. Or, as has happened around ehre, some other code LINKing multiple times to your code and exhausting CICS resources. We had a subroutine that did GETMAINs and no FREEMAINs because it __knew__ that it was only called once per transaction. The somebody called it about 1000 times in one transaction and exhausted DSA (SOS and locked up the CICS region). I am a firm believer in "if I got it, then I should give it back". I do realize that there are times when this may be difficult or maybe even impossible without a lot of rewriting of code. It is just a personal philosophy of programming of mine. Yes, I understand that philosophy and agree, but all of this is down in the bowels of our API library. We control all of that, so memory management and whatnot is not a problem. My only real problem is this was originally ported from M$, so the file I/O methodology is PC based, and it makes it really hard to do efficiently on MVS, and CICS isn't any better. David Logan ---------------------------------------------------------------------- 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
