It is good that there was fallout. Someone didn't do their job, or wasn't doing their job. The remedial action seems curious, especially if the side-by-side products have the same control cards.
If you keep SORT control cards "simple", boy is your site missing out. But, since they pay, it is their choice, so what can I say? Things which exist in one product and don't exist in the other are somewhat easy to find. The final act of tossing a JOB into the overnight run is not the place to do it. There's not much that looks the same but operates differently, but wouldn't you expect that everything needs to be checked (programatically)? An interesting new difference: current Syncsort supports REPEAT= on PARSE. But only 100 PARSEd fields, whereas DFSORT has support for 1000. Of course, there could be a patch out there, and REPEAT= is not documented. On Wednesday, 3 February 2016 01:12:33 UTC, Ed Gould wrote: > Ed: > > That is why I have always resisted the extra features of SORT as they > *ARE* sort dependent. > A long time ago a company I was at set a standard of what sort was to > be used. The LPAR was using syncsort (nothing wrong with that IMO) > and they installed DFSORT. The conversion went reasonably well but > then a yearly job ran 8 months later and it failed because of > incompatibility of sort control cards. > > Luckily the people (IBM IFIRC) helped the programmer change the > control cards so the job was successful. But there were a *LOT* of > political fallout as a result. As a result there had to be a year > overlap on future program product changes so a fall was available. > > On the other hand we were ordered to get of one vendors software and > installed another vendors software and the control cards were 100 > percent compatible. > > SO it seems to work sometimes, just hope you don't get into a > political fight. > > Ed ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
