President's letter to the membership

Of the approximately 6000 members who voted on the Association's proposed change of name, 63 percent favored the change. Since a two-thirds vote was required, we remain the Association for Computing Machinery. Personally I am very sorry. This is the first proposed change that the membership has ever turned down. However, you have voted to change the membership requirements for ACM. After 30 June 1966 a new Member must satisfy certain educational and experience requirements, and have them attested to by two members of ACM. You have established a nonvoting category of Associate Member without such requirements. As I wrote in the July Communications, I do not have a strong opinion about this, and am looking forward to learning how it works out. Naturally your administration will work hard to carry out your expressed wishes. Before our Cleveland meeting I spent two days at San Dimas, California, with the ACM Conference on the Pragmatics of Programming Languages. This was a closed four-day conference sponsored mainly by the System Development Corporation. It seems undemocratic for our Association to get involved with a conference not open to every one, and yet wide experience demonstrates the need for such affairs. A group of say 50 experts can make astonishing progress in an area, if they are thrown close together and presented with reasonable agenda. As compensation to you for the closed nature of the conference, very early deadlines have been set, so that the proceedings of the conference should appear in the February 1966 issue of the ACM Communications. I was invi.ted to San Dimas only for .honorific purposes, and know very litle about the field. In fact, I wasn't quite sure what "pragmatics" means! However, I took advantage of the 10 minutes allotted to me for a welcome, to urge the designers of languages to pay a great deal of attention to the needs of ordinary users. In particular , I urged atention to the need for "fail-soft" language systems, i.e., systems which can respond graciously to programming errors. In spite of a stupid error in one part of my program, I should like to learn all I can about the rest of my program. Some language systems are good at this, but some are so unaccommodating that a single error seems to upset them completely. We have heard a great deal lately about how computing in the future will be like …