Paul Stachour is a long-time software engineer.
He receives his personal EMail and does his personal network-style unix
computing here at Winternet.  Paul's family (wife Fran, daughter Val)
receive their EMail on earthlink, while his son Frank gethis EMail
at through juno.

Paul began computing by writing assembly language programs (on paper
tape, of course) for a school-built computer in the EE department
at Iowa State University in Ames, Iowa, in 1962.  He thus had the
opportunity to get in on "the ground floor", and has been writing
programs for over 35 years.  Paul's favorite early machine was the
IBM 7074 (20,000 10-decimal-digit words), and his favorite computer
is the MIT/GE/Honeywell Multics machine (a very programmer-friendly,
reliable and secure machine, with PL/I and cleanups-on-abort, and
power-down shutdown easily performed by any programmer).  His
early HLL was FORTRAN II, and his current favorite is Ada95.

Paul believes that most software failures are directly traceable
to the fact that, in general, we in the software industry neither
measure ourselves nor learn much from experience.  We couple that
with hardware and software architectures that make house-of-cards
collapses easy, then seem surprised about memory-faults and other
failures.

Paul believes that reliability comes only from disciplined defect 
prevention steps (e.g., requirements agreement between developer
and customer) and defect removal steps (e.g., measurement of
unit-test coverage) coupled with re-use of debugged requirements,
designs, and code.  Paul believes that security and reliability
are duals, and that a system which is not designed with securable,
controlled, and isolated partitions can never be reliable.

Paul is a member of the ACM, Twin Cities Chapter of ACM, Twin Cities
Chapter of ACM SIGAda, and the Twin Cities Quality Assurance Association.
To browse further, try the URL  www.winternet.com/~stachour