[robin2]

An odd thing about Alexander's work on pattern languages is that, despite people's enthusiasm, Alexander himself considered it to be a failure. There's a chapter in Richard Gabriel's "Patterns of Software" (http://dreamsongs.org/Files/PatternsOfSoftware.pdf) that has an interesting discussion on this.

[brudgers]

Alexander considered his work prior to A Pattern Language to be a failure. Notes on the Synthesis of Form was largely his Phd work from Harvard, and his self-criticism was that the process was too formal.

My personal opinion is that it better corresponds to current software architectural practice because whereas A Pattern Language is entirely focused on habitability for end users, both current software architectural practice and Notes on the Synthesis of Form emphasize habitability for the architect/programmer - e.g. three tiered systems make the job of the programmer easier, they don't necessarily improve user experience, but light on two sides does.

[dmm]

Read Alexander's foreword to that book. He considers his earlier work to be a failure. In the sense that design patterns didn't work to allow people to make living buildings. However he seems to be very satisfied with his new Nature of Order series.