In usability and software design there's an informal design meme called paving cowpaths which is a kind of retroactive design philosophy where you nail the design down late after observing how people actually use what you're designing and adapting your design to that.
Some frown on this as a kind of 'undesign', others embrace it (and others yet again would combine the two by calling perl 'undesign')
There's an incident apocryphal story on how a university campus was built without footpaths. Instead the designers just let people walk and only later paved the natural paths people had worn into the campus lawns. As is typical of this kind of urban legend various people add various actual universites to the story to spice it up.
Jon Udell mentioned this fact in a recent blogpost and I went looking for any true story I could find and it turns out there actually is one. Peter Merholz has the good post on the subject. First off, a delightful collections of photos from a campus that actively tries to prevent this kind of design, but in the comments a solid reference to this book in which Cristopher Alexander - father of pattern languages - apparently tell the story of actually doing this when building a campus for The University of Oregon.
Thanks so much for this post! I have heard of Larry Wall's talk about campus footpaths being paved for quite some time now, but I have had a rather difficult time finding a verifiable source. I will certainly be purchasing a copy of "The Oregon Experiment". :)
Posted by: Scott Johnson on June 23, 2005 1:06 AM