Ed I haven't been following this long-running thread but some some devil whispered in my ear to see what it was all about - mainly I expect because a long-running thread tends to drift off course and so I get curious as to where it may have wandered.
I had got to the end just as my normally noisy grandson was enjoying having his nappy changed - he's back to being noisy again! Having digested this post I was able to reflect that, ideally USS messages are composed within the installation and so, if they are inscrutable, the person to whom to complain should be available with an "in-organisation" telephone call - assuming the original author hasn't been "let go". However, I remember well a comment from a CE colleague in around 1977 or so regarding the supplied USS message 7: "SESSION NOT BOUND".[1] "What am I supposed to do with that" he mused! Chris Mason [1] This did get improved but to something that only the "techies" could really handled: "luname UNABLE TO ESTABLISH SESSION runame FAILED WITH SENSE sense". On Wed, 15 Jul 2009 22:33:38 -0700, Ed Gould <[email protected]> wrote: >--- On Thu, 7/9/09, Patrick O'Keefe <[email protected]> wrote: > >From: Patrick O'Keefe <[email protected]> >Subject: Re: IBM error messages getting worse? >To: [email protected] >Date: Thursday, July 9, 2009, 2:29 PM > >On Thu, 9 Jul 2009 08:21:22 -0400, David Andrews ><[email protected]> wrote: > > > >I think you make a very good point. There have always been >absolutely inscrutable MVS messages and there will probably >be new ones. But there are many prefectly clear ones, too, and >there is that message id you can look up, run searches on, submit >RCFs on (if those still exist ... and are read), or even open PMRs on >if they are so bad as to be considered a defect. I don't think it is >getting worse. > >Pat O'Keefe > >- >Pat:I am not sure I agree , but I think that most USS messages are at best inscrutable. Take TCP as an example most of the messages I had to look up at the time did not follow the IBM convention as to importance (W,I,E,C) and then they didn't set the condition code to match the message. The condition code IMO was the worse and it looked to me like they threw the dice when it came to setting codes. Frankly I gave up and after reading the message in the manual 2 or 3 times and it still did not make sense I opened a PMR. I very rarely got anyplace with the PMR as the "USS" people live in the own universe separate from IBM. To me they decided to thumb their nose at the rest of IBM and said we are going to operate as we like to hell with IBM and their rules. >I think the LE people weren't quite as bad, but they are up there. Do *NOT* get me going on COBOL messages. Their so called self describing messages must have been made up on a bad acid trip. >In the past, say 1970 or so we can all agree that some messages like "call your systems programmer" were nightmares especially at 3AM and there was no IBM support to call back then. >Since then (thanks to GUIDE anyway) we made the messages a MAJOR issue and I can still remember 1 GUIDE where the pubs people came into GUIDE and promised to do a better job. It actually did work, thank goodness. Messages actually started to mean something and they were reasonable english straightforward and it might take you a bit to understand the famous VSAM messages that gave you a bunch of possibilities at an answer if you could discern if it was a FC or other type of RC or whatever. If you read it carefully enough it did make sense (most of the time). That was about the time that (sorry I do not remember the name of the IBM product) but IBM shipped you a searchable database every month or so and you could play with search args to find something you couldn't make heads or tales out of. Of course now its IBMLINK (when it is up) and it functions the same and with reasonably more up to date issues than the once a month tape shipment. >The pubs people might have lost their way as it seems in the late 80's (especially with USS) components (I DO NOT MEAN UNFORMATTED SYSTEM SERVICES so if anyone wants to get anal about the meaning I do not care. >Ed ---------------------------------------------------------------------- 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
