From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: *
X-Spam-Status: No, score=1.8 required=5.0 tests=BAYES_40,FROM_STARTS_WITH_NUMS,
	INVALID_DATE autolearn=no autolearn_force=no version=3.4.4
Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!necntc!ames!sdcsvax!ucbvax!NAVPGS.BITNET!4526P
From: 4526P@NAVPGS.BITNET ("LT Scott A. Norton, USN")
Newsgroups: comp.lang.ada
Subject: Re: Assembly language (was: Re: Another 1.3 wish.)
Message-ID: <8708241851.AA01472@ucbvax.Berkeley.EDU>
Date: Mon, 24-Aug-87 14:29:19 EDT
Article-I.D.: ucbvax.8708241851.AA01472
Posted: Mon Aug 24 14:29:19 1987
Date-Received: Tue, 25-Aug-87 04:28:21 EDT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
List-Id: <comp.lang.ada>

With all the "my language is better than your language" noise going on,
I thought I should provide some perspective.  There is one overriding
concern in DoD's world of software embedded in weapon systems.

              MAINTENENCE

Consider that most DoD weapon systems have a life cycle of over 20 years,
and that that lifetime is full of changes.  ( Don't quote me on this
number, since I don't have my references handy, but I think that avionics
software has 50% of the code rewritten every 6 years. )

My previous assignment was as Tactical Data Systems Maintenance Officer on
a 20-year old guided missile cruiser.  The NTDS program was 25 years old,
since it was deployed on a previous class of ship.  In 25 years, just
imagine the changes that took place:

  Replacement of analog fire control systems by digital.

  Introduction of Harpoon surface-to-surface missiles.

  Interface with digital sonar.

  Automatic tracking radars and IFF systems.

  Replacement of electronic warfare systems.

  Support for the F-14 fighter, with a two-way data link and multiple
  intercept capability.

Also realize that in a 25-year lifecycle, many participants have a
hand in the software.  Components are produced and maintained by Navy
activities, contractors, and research labs.  Univac wrote one module,
which is maintained by NavSeaTechRep, St Paul; Johns Hopkins APL wrote
another, which was turned over to missileers to maintain.  The data
link was written by a contractor to conform to a joint Army-Navy-Air
Force standard, and then maintained by FCDSSA, a Navy activity.  So,
the program must stand on its own, without the benefit of corporate
knowledge or the "Lord High Fixer", who was there when it was
originally written.  Doug Bryan wrote
<12328308981.12.Bryan@Sierra.Stanford.EDU>
>...  there is nothing about military software that makes it any
>harder or easier to implement than many, many other kinds of
>software.

The software that Ada was meant for, embeded in weapon systems, is
harder to implement than most, since it operates under strict
constraints of time, size, and correctness.  But most important,
its lifetime is as long as any COBOL banking application, and yet
as full of changes as any operating system.

LT Scott A. Norton, USN
Naval Postgraduate School
Monterey, CA 93943-5018
4526P@NavPGS.BITNET