Thou shalt have no other gods before the ANSI C standard 1528

Bit Gods may be irrelevant to science, but conversely science is irrelevant to Bit Gods. The LSG at DEC was not in the business of producing science. That is the bailiwick of MIT, CMU, ISI et al. Why, even within Digital, they had operations like DECWRL.

*Therein* lies the rift between Science and Practice.

In SCIENCE: To be accepted into the corpus, something must be provable.

In PRACTICE: One does not need to show that a solution is provably optimum - one needs to show that it meets the requirements. Only when the requirements are "be scientifically provable"1 do JMF and other ersatz bit gods need to show the proof.

Business is a *consumer* of science - if Dr. J. Random Borgwart from the Upper Helsinki University for Computronics publishes a well received paper that the Finnish Parbreastioning Algorithm is the state of the art in parbreastioning n dimensional trees - and my project requires parbreastioning n dimensional trees, I will implement his algorithm. If it meets my overall performance goals (which could be as simple as: is better than the crufty old Brute Force and Spit algorithm), I have done my job, my customers are happy, I get to eat.

That said, a wise business operation might find it in their own best interest to support science either directly (Bell Labs, DECWRL) or indirectly ( all of the dozens of university laboratories who receive direction and funding from industry ).

Like Bill Leary, upthread, I often when presented with a problem will give an intuitive answer. Fortunately for me, I work in an organization where I can say to my handler, "Well -- I ain't fersure -- but my gut tells me xyz ..." Perhaps even *more* fortunately for me, my gut has been right sufficiently more often than wrong, that my say-so is sufficient. *That* is called experience, and it is why noteverySoftware Engineer (see upthread for my macro expansion of software engineering) has been replaced with a recent grad who will work for a third the price.

Appeal to Authority is a fallacy in forensics, but it is strong medicine in "getting my product out the door and some beans in my larder."

Footnotes: 1 Of course, if you are making the case that "be scientifically provable" is, or should be, a requirement more often, that is a separate debate. It all comes down to a cost-benefit analysis, and deserves a new thread.

-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- Computer software consists of only two components: ones and zeros, in roughly equal proportions. All that is required is to sort them into the correct order.