Title: From the Coalface

The project is sold to me as an OOA/D implemented in C++. I think, "Sounds good some solid experience and there isn't anything else interesting happening here. Let's do it".

Not very much later I'm working through scenarios for the various use cases we have identified. The project manager says, "Why are you doing that? It's far too detailed. We haven't got time to go into that much detail". I start worry-ing that something isn't quite right. Some months into the project we don't possess a complete set of use cases and scenarios, an object model or any dynamic models. The customers have been "sold" the product. I'm getting very worried.

Still on the same project. We have to use a relational database rather than an OO database. Fair enough. I crack on with designing a data model (on the basis that even if we don't know what the objects will look like at run time at least we'll know what data we need to store). Various snide comments about the desirability of this exercise later I am treated by the project manager to what I think is destined to become a classic quote, "Of course we do things differently. You like to start with a data model and build on it. I prefer to start with the user interface and see what data structures I need to support it and add to them as I go, but of course we still end up with the same result." My jaw still drops in disbelief whenever I think about this.

The same project manager, in response to a change arising out of a meeting with customers to discuss the data model and its suitability, "We can't do that, it doesn't fit in with our interface design." His emphasis.

I go to the next person up the chain (having tried by various subtle means to indicate to the project manager that I'm concerned things aren't terribly well "planned") and explain my worries about this project and suggest that at the very least a formal review of progress and plans is required. The senior person says, "<project manager> says everything is in hand". I do this twice more and get the same response, almost verbatim. In November I start looking for another job.

So I'm leaving, two of the three contractors who make up the rest of the team aren't going to renew their contracts, for the sort of reasons outlined above. This leaves the project manager and one other. I am hoping that the mass desertion will wake the management up to what appears to be a complete disaster. I hope for the customers' sake that the project is dropped. They will have lost money, but at least they won't be saddled with an unsuitable product that is likely to be delivered very late and almost certainly over budget.

One last quote. The same project manager, different project. "I'm the best C++ coder I know." Draw your own conclusions.