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=-1.9 required=5.0 tests=BAYES_00 autolearn=ham
	autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,60e2922351e0e780
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 2003-11-13 11:11:15 PST
Path: 
 archiver1.google.com!news2.google.com!news.maxwell.syr.edu!newsfeed.mathworks.com!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi_feed4!attbi.com!attbi_s02.POSTED!not-for-mail
From: tmoran@acm.org
Newsgroups: comp.lang.ada
Subject: Re: Clause "with and use"
References: <bov8qg$22e2$1@msunews.cl.msu.edu>
X-Newsreader: Tom's custom newsreader
Message-ID: <lHQsb.194893$e01.711812@attbi_s02>
NNTP-Posting-Host: 12.234.124.41
X-Complaints-To: abuse@comcast.net
X-Trace: attbi_s02 1068750673 12.234.124.41 (Thu, 13 Nov 2003 19:11:13 GMT)
NNTP-Posting-Date: Thu, 13 Nov 2003 19:11:13 GMT
Organization: Comcast Online
Date: Thu, 13 Nov 2003 19:11:13 GMT
Xref: archiver1.google.com comp.lang.ada:2456
Date: 2003-11-13T19:11:13+00:00
List-Id: <comp.lang.ada>

>> have been comfortable with (2S + B)/(N * ln(N)) where N is the total
  Given an N package system, with "goodness" G, and the opportunity to
split one package, A, into two, A1 and A2:  Everyplace that 'with'ed A
now needs to 'with' either A1 or A2 or both.  If the number of bodies that
will need both (ie, experience a net increase in 'with's), plus twice the
number of specs that will need both, is less than G*ln(N), you should do
the split.
  So in a 150 package system with goodness 5, if a package can be split
with no more than 8 specs that will need to 'with' *both* A1 and A2, and
no more than 9 such bodies, doing the split will make the system "better".
  Does a similar metric - limiting the context necessary for some
piece of code - apply to the size of a procedure or its number of
variables, or the number of first-level components of a record, etc?