Saturday, 17 January 2009

Facts and Fallacies of Software Development


It’s been said by people a lot cleverer that me (DeMarco and Lister 1999 – Peopleware) that most developers don’t read books. So with a typical level of humdrum insight I’ve worked out that if I read a few programming books a year then it will lift me a bit above the norm – keep a positive career calculus.

Facts and Fallacies of Software Development by Robert Glass is the latest book to have taken up residence on my bedside table. It’s a collection of little known “facts” with a smaller number of falsehoods about software engineering – unsurprisingly. How though can there be a “fact” in a profession that seems to owe more to superstition and religion than it does to science and engineering. The conceit it that the facts are items of research and less often the authors personal experience. So at it heart it is a book popularising academic research in a series of short (1 -5 page) articles.

So does it work? I have enormous sympathy with the project of academic popularisation and God knows – computer science needs it more than most. It’s very engaging, very readable and gives food for thought. For me though the nature of the book makes it a bit disjointed. I appreciate IT books are novels – but that said I do like them when they have more of a beginning, middle and an end. That said – for dipping into its perfect. So buy it when you’ve a long journey – sitting on a train, sipping an overpriced coffee and reading this book – that’s the way to travel.

And my favourite fact – well it’s “[Software] Maintenance is not a problem it’s a solution”. When you’ve completed your project and your client comes back with 4 pages of desired enhancements it isn’t because your product is rubbish – it’s because your client has total engaged with it and can now see the potential. And if they pay for it – then everyone’s a winner.

No comments: