Our trade is divided in many ways: experienced and novice, business and engineering, major and mini. Men and women, too - thank heaven, as Maurice Chevalier used to say! One of the more difficult divisions concerns technical papers. On the one hand we have our academic brethren, who write profusely.
Sometimes repetitively, sometimes without much content - but always with the full apparatus of references and referees. And on the other hand there are the silent installations, where tough problems abound but no guidance for others is recorded. Talk at the local DPMA chapter meeting, sure; advice within the parent corporation, perhaps. But nothing for the wider community.
The reasons for the, difference are clear: the professors benefit
from publication, in the esteem of their peers and in terms of promotion. Other peoples' papers can be used to pad out one's own citations and footnotes and bibliographies. There is a praiseworthy urge to read the latest advance, to discuss at next week's seminar, to try out new algorithms. But for the business data processor or the engineering computer specialist, publication is often contraindicated. The boss thinks it a waste of time or even giving away of comercial advantage. The peer group is less vocal. There are no eager disciples, no subservient thesis grubbers.
Yet the nonpublishers do want more than they get. The DPMA and ASM journals are thin, the articles too narrow or too broad, the authors often clumsy and lacking references and follow-on material. The ACM and IEEE journals are full of computer science: artificial intelligence, peculiar languages like Algol, disputes about minutiae. They abound in solutions to problems no one cares about; the business DP magazines list real problems, but provide few answers.
In despair, practitioners with tough problems that need new powerful techniques go off to Guide or Share or Univac group. But that's a limited universe, and one pretty much inhabited by only one kind of advice: buy only one kind of advice: buy a little more core, add a few more program packages - the builder knows best.
The trade press tries: Computerworld, Datamation and their ilk. But temporal pressures interfere, and spot news, and advertising. We need something better a Scientific American of computerdom. In my view, the only way to get such a magazine is to have it be the journal that goes to all the members of a 200,000-member North American Computer Society. And we are a long way, alas, from that.