The segment on the fitted kitchen was interesting. A fitted kitchen seems like a natural and obvious thing now, but the idea had to be invented. The discussion of the Frankfurt Kitchen in the programme was the start of an interesting thread. Users of the kitchen tended to be a bit ill-disciplined. Certainly the tended to disregard the labels permanently attached to the custom–made drawers and put any old thing in them. Users found that the kitchen was built to support well only certain workflows, workflows that they didn't like, didn't understand and couldn't change. Workflows devised by an architect who couldn't cook.
Meanwhile, another Bauhaus architect, Le Corbusier, is being given free reign to redesign entire cities, up to the point of making models, anyway. Filling them with great towers full of his “machines for living in”. And we know how that worked out once people started taking it seriously.
If you are a regular reader of this blog you probably know where I'm going next.
Commentators on software development often seem to exhibit a lot of discipline envy. Two common themes are that 1) our projects should exhibit the reliability of those in the “established” branches of engineering, and 2) our projects should exhibit the conceptual integrity attained by building architects.
That conceptual integrity can be a dangerous thing. Lihotzky's kitchens had a lot of conceptual integrity (and a lot of research to back that up), Corb's vision of mass housing (and its implementation by later architects) had a really astonishing amount of conceptual integrity. Neither leads to much in the way of joy for users (*). The Bauhaus architects designed a lot of chairs, none are comfortable to sit in.
One of the designers interviewed in the programme explained the problem along these lines: architects tend to be ‘I’ shaped, by which he means they have a very deep knowledge and skill in their craft, but not a lot else going on. Designers tend to be ‘T’ shaped, deep in craft but also with a breadth that touches many other disciplines. And from that breadth comes the ability to design objects that people can comfortably incorporate into their lives.
I think that the application of this thought to the software world is clear.
(*) The very few dense housing projects that Le Corbusier himself built have proven to be resilient and popular. It's the shoddy works inspired by his ideas and executed without his art that are the problem.