Castell de Montgri

I recently visited Castell de Montgri, in Catalonia, Spain. It's a really nice castle, built between 1294 and 1301. It's small, only 31 m by 31 m, but really well made - you could see the craftsmanship which had gone into it, for example, on the careful facing on the stonework. It was in amazing condition too, considering it's over 700 years old. It was built in the style of the crusades, apparently, a square footprint, with round towers on each corner - it looked really good. It's location was, dramatic, perched on top of a 300 m rocky hill, visible for miles around. Three sides of the hill were really steep. You could of climbed them, but it would have been jolly hard work. The fourth side, which we went up, was not easy either - a good hours hard walk and scramble to get to the top. They had the "impenetrable fortress" requirement nailed.

There were however, how should I put it....a few problems. There was no well - so no drinking water. Hell there wasn't even any soil - just rock. And no other buildings, no town to defend. Just the castle alone on the top of a hill. The "defend something valuable" requirement had been forgotten. The blurb inside confirmed my suspicion, construction was abandoned before building was complete, "due to changes in its strategic importance". I think what they actually meant was "because it had no strategic value whatsoever".

All this got me thinking about the projects I've worked on. I spent two years working on a "Castell de Montgri" - a project which had no value whatsoever. It's not an experience I want to repeat. I recently read the Tom DeMarco article Software Engineering: An Idea Whose Time Has Come and Gone?. In it he considers two dramatically different kinds of project:

• Project A will eventually cost about a million dollars and produce value of around $1.1 million.
• Project B will eventually cost about a million dollars and produce value of more than $50 million.

I've been considering this, and I can say I've worked on lots of projects of type A, which make about 10% profit. In DeMarco's words "projects which deliver marginal value". I have yet to work on a type B project....but maybe the one I'm on now could turn out to be a type B - I hope so.