RE: Engineering versus Science, Anecdote versus Evidence ... [Was:Design

["Cox, Bruce" <Bruce.Cox@USPTO.GOV>]

I would have said that the goal is to move away from software engineering-as-a-craft to software engineering-as-an-industrial activity.  I'm sure there are artists among our developers, but that isn't the problem.  Robust, maintainable systems can't depend on the skill of a single craftsman.   Our goal here has been to standardize key processes, behaviors, and materials so that any professional engineer can join a team and take over where their predecessor left off.  By "professional" I mean someone with a deep understanding of a common body of knowledge and recognized skill in putting that knowledge to use. 

It's been my experience in attempting to develop "best practices" guides for developers here that most of the effort goes into making choices among many alternatives based on the judgment of acknowledged professionals.  It appears to be most successful when those who will follow the conventions we establish have a hand in reviewing them.  Some conventions are imposed by external forces (this IS the Federal Government, after all), but even there we have choices to make about where, when, and how to follow them.

It's difficult to imagine how a strictly scientific approach would be used in this context.  I'm expecting that the best practices we agree to will evolve based on feedback from their use.  I doubt that there will be any controlled experiments producing that feedback; more likely there will be "lessons learned" and other post-mortems where we document our experiences in order to guide those who follow.  I think this is good enough for most business environments, even if it is "anecdotal."

Bruce B Cox
USPTO/OCIO/AED/SAED
571-272-9004


-----Original Message-----
From: Costello, Roger L. [mailto:costello@mitre.org] 
Sent: 2011 December 29, Thursday 14:18
To: xml-dev@lists.xml.org
Subject: Engineering versus Science, Anecdote versus Evidence ... [Was: Designing an experiment to gather evidence on approaches to designing web services]

> I'd suggest software design is closer to engineering than science

But isn't the goal of every engineer to move steadily away from engineering-as-an-art to engineering-as-a science? 

Stated differently, shouldn't we endeavor to approach engineering problems as scientists?

/Roger 

P.S. Fascinating discussion!