# from David Golden
# on Tuesday 03 March 2009 12:02:

>> I would expect application of the Science to be found in the
>> Software Engineering program, but we've got a ways to go yet before
>> it becomes a professional discipline -- complete with licensing and
>> responsibility.
>
>You might be interested in this article by Peter Deming, "Is Software
>Engineering Engineering?
>
>http://cs.gmu.edu/cne/pjd/PUBS/CACMcols/cacmMar09.pdf

Yes.  Interestingly, he mentions the 1968 NATO conference (where the 
term "Software Engineering" first appeared.  From what I've read in the 
proceedings of that conference, one key debate was about software 
pricing (and IIRC, a substantial portion of that was about whether or 
not vendors should be "allowed" to charge for software.)  Yet, Deming 
doesn't have much to say about the relative lack of openness in 
software and computer science compared to other disciplines.  I'll also 
note that most of the world's population does not have any difficulties 
relating to race conditions in the flight controller of their fighter 
jet (and anyway, the factor of safety for most warcraft is roughly 0.9 
to compensate for conservative assumptions hidden in the material 
specs.)

But steel and concrete are priced per-pound (or a proxy of weight such 
as volume aka truckload.)  And even installation cost is roughly linear 
per pound assuming typical methods.  The engineer gets a percentage of 
this.  Perhaps we should charge by the pound for CPAN modules?  Not 
lines of code, but some kind of metric which appropriately conveys heft 
(toughness/strength/finish are, of course, premium.)

One should also note that the importance factor for most software is 
about 0.1.  The very slow (yet massively parallel) death of "wasted 
time" apparently does not count as "lives are at stake."

I think we figured out long ago that the design algorithm (given the 
classic "pick-two" variables $good, $cheap, and $fast) is basically:

  if($good) {
    goto &engineering;
  }
  else {
    print q(print "hello world\n";);
  }

And nobody ever picks $good, yet they wonder why they get what they get.

--Eric
-- 
Issues of control, repair, improvement, cost, or just plain
understandability all come down strongly in favor of open source
solutions to complex problems of any sort.
--Robert G. Brown
---------------------------------------------------
    http://scratchcomputing.com
---------------------------------------------------

Reply via email to