Sat | 3 May 2003
Design Papers
11:55 EST |
These papers deal with questions of design. Most of the design papers collected so far are some commentaries on papers by Paul Graham. Over the course of the past year he has written and released a couple of interesting and provocative notes on his web site. I have written parallel-ogues on each of these papers. These documents are paragraph-by-paragraph commentaries on the matters being discussed.
While I disagree with much of what Graham says, I find his remarks always provocative and therefore worth discussing.
Graham on Beating the Averages
This is a paean to Lisp. The problem is that it offers no evidence whatsoever on any issue other than the fact that Graham likes Lisp and is quite apparently able to use it effectively. While I don't doubt this, it is of little use to anyone else, and Graham offers no evidence that others would find the choice of language to be significant in the process of deveoping systems and successful startups.
Graham on Hackers and Painters
In this paper Graham tackles the task of describing things that he feels painting shares with `hacking'. I find it to be a very careless paper, too casual about the use of words, and with little contribution to either hacking or to painting.
Snctns isnt Pwr
This paper is an extension of my remarks about Graham's Succinctness is Power paper. The technology of the parallel-ogue, while good for responding to points made in a paper, is not very helpful in dealing with problems that arise because a paper doesn't deal with all of a problem. It is good technology for discussing what someone says, but not for discussing what they fail to say. In this paper Graham's proposition about succinctness is dealt with in somewhat more depth.
Graham on Succinctness is Power
In this paper, Graham discusses the relationship between succinctness and power. I find the discussion is interesting, but hardly compelling. There seems to me to be lots of confusion between the language itself and expressions of programs in the language. The paper also pays scant attention to the problems of interaction effects between language and personal programming style.
Graham on A Hundred-Year Language
Graham's paper argues that by trying to develop a view of the `programming language of the future' we can learn something about what we ought to be doing today. I am not so sure.
Graham on Design
In this paper Graham attempts, in my opinion less than successfully with attempting to relate design efforts and research efforts. I find the paper somewhat unclear on whether it is talking about whether it is dealing with research on programming language design, research on programming or rather more general relationships between programming and research. But it is an interesting paper and well worth discussing.
Graham on Taste
In this paper Graham raises issues associated with the development of taste, particularly in the context of design. He draws heavily on examples from Renaissance art, and while I find some of this discussion successful and almost all of it interesting, I also find a substantial part of the discussion mis-directed and wrong. Arguments about Art, Science and Engineering are often very interesting, but I am not sanguine about the liklihood that these discussions are likely to lead to any design consequences that have much immediate impact.