Say what?

[plonq]

We recently switch from the Roscoe mainframe environment over to TSO/ISPF because it is cheaper better. All through the process, our IT people kept lying assuring us that the transition in environments would be a nightmare seamless. All that we would need to do is sacrifice a goat to the old gods move our libraries over to the new environment and we'd be eternally damned to a meaningless life of heartache and drudgery up and running like nothing had happened.

They were as wrong as a chocolate Jesus right for the most part, in that things are running just fine. Mind you, they are "running" in the sense that they are taking up CPU time and resources but generating no output. Since most of the people who use this output are idiots find blank reports to be less-than useful, this has understandably led to some mixed frustration and amusement on my part consternation.

We've had a couple of the legacy guys (who also happened to be the IT champions of this new system) looking into it for us over the past couple of weeks to see if they can figure out why we seem to have created the reporting equivalent to Schrödinger's Cat - that is, the reports run and do not run at the same time. They have been merrily ignoring my correspondence running test cases, and moving things in and out of the schedule to see if they can replicate the error. Replicate it?! The fact is they can't not replicate it. Today I got some email confirming that they had finally discovered on their own what I had painstakingly outlined for them in a detailed email almost two weeks ago made some important headway in identifying the problem.

1) The jobs that are scheduled to run in the JCL library are launching properly and generating output.
2) The jobs that they are subsequently calling in the user library are running just fine, but sending out blank reports.
3) They all chain and generate output properly when they are triggered manually, but not when the system triggers them.
4) They can find no errors in either the program, nor the Job Control Language (I knew that, but I am a bit biased, having written it)
5) This all worked perfectly in Roscoe.

The fifth item was not one of their conclusions, rather it is a fact to which I rather tersely called their attention when it became obvious that the new system sucked they were ignoring my correspondence.

I like many things about the new system, and I generally agree that it is better than the old, but I wish they had ensured that things actually worked in it before throwing the switch.

Comments

[duncandahusky.livejournal.com]

Well, I'm certainly glad it's you and not me sorry to hear that you having to deal with this steaming pile challenge, but I'm sure you'll be screaming up and running through the office beating people with a stapler in no time.

[furahi.livejournal.com]

How is it better than the old?
Is it because it doesn't work? ;)