🚀 go-pugleaf

Failure is just as real for software as it is for hardware. The only
difference is that the reason for failure is strictly due to the design
rather than a combination of the design and manufacture. When talking
about MTBF of hardware components we are usually talking about the
probabilities of those components (or subcomponents) will fail. In
software the MTBF is more closely related to the defect density and can
be derived from an estimate in defect densities. This is based on the
assumption that all software is created equal (that's something I never
bought into, though :-)

cheers,
-Jim.


In article <RJC.92Mar4031453@brodie.cstr.ed.ac.uk> rjc@cstr.ed.ac.uk (Richard Caley) writes:

   Path: mips2!news.bbn.com!olivea!uunet!mcsun!uknet!edcastle!aisb!cstr!rjc
   From: rjc@cstr.ed.ac.uk (Richard Caley)
   Newsgroups: gnu.misc.discuss
   Date: 4 Mar 92 03:14:53 GMT
   References: <9202282313.AA03177@wubios> <1992Mar3.021139.17963@wpi.WPI.EDU>
   Sender: news@aisb.ed.ac.uk (Network News Administrator)
   Distribution: gnu
   Organization: Centre for Speech Technology Research
   Lines: 12

   In article <1992Mar3.021139.17963@wpi.WPI.EDU>, Lawrence C. Foard (lcf) writes:

   lcf> What does the MTBF mean for software? All software is perfect
   lcf> copies of the original?

   But software is an abstraction, real computer systems are real
   physical objects and interact with the real world and so fall over now
   and again.

   --
   rjc@cstr.ed.ac.uk			_O_
					    |<