Bottom line: The system was never used in production. Because nothing was actually delivered until the end of the project, the user's commitment was never tested.
Thanks, John. The entire system development effort was done online, not offline. (I was using IBM's CMS and most of this was written in CICS). But more importantly, there was nothing stopping me from breaking the project into two phases, as I noted as a lesson learned. That would have allowed us to test the user management commitment before committing to the much greater effort to develop the entire system up front. In fact, I did exactly that in my very next project, which I wrote about here: https://enterprisespectator.substack.com/p/the-most-significant-system-developmenthtml. That had a much better outcome.
This story is a vivid example of the major shortcoming of the waterfall approach. (Of course, that approach was largely driven by the high cost of mainframe -- most common then -- computer time, so we had to do as much development work off the computer in advance of actually putting hands on keyboards.) Waterfall enforces rigor, but I 100% agree an agile, iterative, collaborative approach is better. (Now, except for AI, computer use is almost free and developers are very expensive.)
Thanks, John. The entire system development effort was done online, not offline. (I was using IBM's CMS and most of this was written in CICS). But more importantly, there was nothing stopping me from breaking the project into two phases, as I noted as a lesson learned. That would have allowed us to test the user management commitment before committing to the much greater effort to develop the entire system up front. In fact, I did exactly that in my very next project, which I wrote about here: https://enterprisespectator.substack.com/p/the-most-significant-system-developmenthtml. That had a much better outcome.
This story is a vivid example of the major shortcoming of the waterfall approach. (Of course, that approach was largely driven by the high cost of mainframe -- most common then -- computer time, so we had to do as much development work off the computer in advance of actually putting hands on keyboards.) Waterfall enforces rigor, but I 100% agree an agile, iterative, collaborative approach is better. (Now, except for AI, computer use is almost free and developers are very expensive.)