Parnas mentors

Everard M. Wiliams
Wiliams taught me that:
- We should not use the title "Engineer" to describe anyone who builds things, but only to refer to those whose education and experience prepared them to base their practice on sound scientific and mathematical principles.

Alan J. Perlis
He was the first to show me that the intellectual discipline and mathematical rigor that I enjoyed in engineering , could also applied to software.
From him I learned to look for methods that would get the job done reliably, not mechanisms (he is referring to AI) that were modeled on the way that people or animals appear to solve a problem. .. Imitating human behavior is not the path to reliable software.

Leo Aldo Finzi
Finzi taught us that we were designing in a multidimentional space.. and that no single metric (or figure of merit) could capture what was meant by "good design"... He also showed us it was always possible to design a product that had a good value of a metric, but was still a lousy design.

Harlan D. Mills
One of Mills's most important lessons had to do with the difference between management and engineering. He did not try to to tell us how to manage unmanageable projects;.he showed us how to make projects manageable. He understood that only well-designed and well-documemted software could be managed properly. He taught us how to design products in ways that made management easier. Today, when so many papers discuss software engineering as if the problem was simply project management, I miss Harlan Mills.


No comments: