Today I tried running my latest version of the on-time delivery tracker at the site with the highest volume. It didn’t exactly work because the latest version of the upgrade was not ready for roll-out; it was work in process. Even if I had thought it was finished there would have been glitches, so not even thinking of it as finished there were enough problems I never got through a complete and fair test.

My brother helped me identify that my biggest obstacle here is probably the network performance, which my personal experience on-site confirms is very poor. I suppose when one takes an industrial manufacturing facility and converts it into the business hub a certain lack of information infrastructure may become evident.

My hunch is that the solution (operating the analysis on the local sytem, without handling files from the network) will adequately address the problem. Since my boss clearly does not want me to spend more time on this project, a successful outcome of this test will result in him saying “Okay, whatever you tested, let that be the working system and let’s move on.” I cringe in anticipation. Please refer back to remarks about this not being a finished version. It is painful trying to do maintenance on a system that was never actually set up to keep running.

Another thing I did today was haul out more than a week’s worth of data from the version running at my own site, and send out my reminder to everyone who was supposed to check over it and fill in the gaps. Nobody here really uses the data from this analysis; we package it up and send it off like we are supposed to, but we don’t try to do anything with it. So what outcome of this stop/go analysis will I consider a success? I ought to be clear on my bias before performing the test, don’t you think?