I've been travelling. On a recent visit to London, England I took the opportunity to talk to a twenty-year veteran Systems Developer with a large financial institution in the City. The conversation went something like this

JW “I'm interested in finding out about how you do development projects so I can be sure that what I tell my students is up-to-date. What methodologies are you using?”

SD “Well, we don't actually use any methodology as such. It depends on who is doing the development, and what the development is about”.

JW “So what DO you do?”

SD “We generally work in small teams of two (but sometimes just one). If the two get on well they talk to each other about what they are doing all the time – they're usually co-located across a desk from each other.”

JW “And if they don't get on well?”

SD “Well that's when we have problems. Sometimes it works and sometimes it doesn't. We might have to change the team, or sometimes just abandon the development.”

JW “Who makes that decision?”

SD “It's up to the project manager then, if they're any good. Or sometimes the customer.”

JW “So do the project managers have a methodology they follow?”

SD “Some of them probably do, the better ones do. We've got some who have done PMI training and some PRINCE2. They tend to get paid more, and some of them are worth more, but there are no rules about this.”

JW “Sounds pretty chaotic to me – a bit hit or miss?”

SD “It's not as bad as it sounds. We only do small projects. It's unusual to have anything run more than six months elapsed time (or one man-year), and we break down bigger projects into smaller chunks. The business is pretty fast moving.”

JW “Who coordinates the small projects?”

SD “The project manager. Good project managers will be looking after a dozen projects. Mind you, there's not many that good, most have trouble coping with three projects max. A lot of them are still developers at heart.”

JW “So the developers and the project managers do things any way they like and rely on the project managers to pull it together?”

SD “I suppose that's about it. It would be difficult to get everyone to agree on a standard way of doing things.”

JW “So what should I be teaching my students?”

SD “ Project Management is very important, so learning something about it would be useful. On the development side, it depends on the people, the problem and the timescale. I suppose we do refer to methodologies – Agile, Scrum, Prototyping, Iterative development (as long as the number of iterations is controlled), and even waterfall have their place, depending on the circumstances.”

JW “It sounds like what I'd call 'Pragmatic Programming'.”

SD “Yeah, that would be a good name for it. They need to find what works for them and adapt it. Recognise when it's appropriate to use a particular approach. Oh, and as the song says 'Do what you do do well, boy'...”

I heard similar tales from others in different places – people continuing to muddle through, making the same mistakes over and over again.

Outside, my vision of highly efficient and effective development practices in financial institutions in tatters I came across an interesting multi-story building...

I used to work in the City way back in the sixties, and I remembered the building from then, now in the same partially complete state for over forty years. The story I heard was this:

In an attempt to beat the English weather and avoid delays, a top-down approach was adopted. The central core with lift shafts, stairwells and services was built using traditional methods. The roof was then cantilevered out to provide shelter for workers underneath, and subsequent floors were hung from the top down. Unfortunately something went wrong and they had to stop, and the building has been like this ever since. (Of course another explanation is that the building was completed, passed its useful life and was being demolished from the bottom up).

It seemed somehow appropriate. A new technique that didn't quite work, abandoned with noone knowing quite what to do about it (or knowing but being unprepared to make a decision).

It's not just IT projects that run into trouble, but there are things that can be done, like spreading the word about PMI.