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=1.6 required=5.0 tests=BAYES_20,REPLYTO_WITHOUT_TO_CC
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,eea22f75441ee810
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 1995-03-28 06:33:46 PST
Path: 
 nntp.gmd.de!news.rwth-aachen.de!news.rhrz.uni-bonn.de!news.uni-stuttgart.de!rz.uni-karlsruhe.de!xlink.net!howland.reston.ans.net!cs.utexas.edu!news.sprintlink.net!uunet!in1.uu.net!portal.austin.ibm.com!bocanews.bocaraton.ibm.com!watnews.watson.ibm.com!usenet
From: walt@lfs.loral.com
Newsgroups: comp.lang.ada
Subject: Re: Ada and the COCOMO model.	Question....
Date: 28 Mar 1995 14:33:46 GMT
Organization: Loral Federal Systems, Owego New York
Message-ID: <3l96oa$k5t@watnews1.watson.ibm.com>
References: <796093965snz@gelert.demon.co.uk>
Reply-To: walt@owgvm3.vnet.ibm.com
NNTP-Posting-Host: walt.endicott.ibm.com
X-Newsreader: IBM NewsReader/2 v1.04
Date: 1995-03-28T14:33:46+00:00
List-Id: <comp.lang.ada>

In <796093965snz@gelert.demon.co.uk>, jay@gelert.demon.co.uk (jay roughley) writes:
>I believe that there are products out there that already do this (CB CoCoMo for
>Windows, Costar(V.4.0)), but does anybody use them?  How accurate are they in
>making predictions for s/w planning?

Jay

You must remember that the COCOMO model as described by Boehm in 1982
was based upon the compilation of data from the late 1970's and early
1980's in the commercial environment. Much water has gone under the
bridge since then.  He also indicated in the text that calibration was
necessary to make his concepts work...ie you need to calibrate his model
type to your own experience and type of software.

As a result, all the major software estimation tools have a calibration
toolset embedded in the offering to adjust the coefficients to make them
apply to the customer's environment.

Here at Loral Federal Systems, we have been collecting metrics on our
software since Tom Watson was a child, and have calibrated all of the
software estimation tools we utilize. With this calibration we are
getting very good correlation between extimated development effort and
the resulting development effort as well as in the schedule area.

The various tools and models have their individual strengths...we
utilize several models in developing an estimate for our organization.

Also, Dr. Boehm has started his COCOMO II Affiliates program to revise
COCOMO to reflect the realities of 3rd->4th generation development
technology. Remember, model technology today addresses the development
of software applications, not the integration of existing applications
into a system solution to a need. This is where the cutting edge is
today.

Walt Johnson
Loral Federal Systems Group
Owego, New York 13827
waltj@lfs.loral.com
(607-751-2158)
FAX(607-751-6223)