Re: Whats the point?

From: The Fungi (
Date: 06/28/02

On Fri, Jun 28, 2002 at 12:27:47PM -0700, Gerald Florence wrote:
> It's simple programming science.  I forget the formula, but there's a
> direct relationship between the number of modules and how they
> relate to each other in testing.  One module is one test.  Two
> modules is one connection.  Three is three connections.  Four
> becomes 6 connections.  Five is 10.  As you can see, the number
> of modules grow quickly once you get past 4 modules.

This is the triangular sequence (though it is known by other names
as well). It is described by applying the transformation n(n+1)/2
against the set of ordinals. The geometric figures represented in
your diagram would be point, segment, triangle, tetrahedron,
hypertetrahedron, supertetrahedron, superdupertetrahedron (sometimes
mathematicians choose silly names), ad nauseum. Not sure what those
applied dupes in compsci call their theory, but there's the math.

OBCIRCLE: My idea for a forked codebase is to use well-maintained
patchsets. What I've started implementing with my circlemud packages
for Debian is a system by which I maintain a set of up-to-date diffs
for the different add-on stuff and have it patched in at
compilation. Then as the codebase or a patch gets updated by its
official maintainer(s) all I have to do is replace the appropriate
files in the package and tweak a few of the diffs downstream to
patch cleanly in sequence. It doesn't go quite as smoothly as all
that, but it's not as insane as trying to merge Circle, DG or Oasis
version updates into an already heavily modified codebase.
{ IRL(Jeremy_Stanley); PGP(9E8DFF2E4F5995F8FEADDC5829ABF7441FB84657);
SMTP(; IRC(; ICQ(114362511);
AIM(dreadazathoth); YAHOO(crawlingchaoslabs); FINGER(;
MUD(; WWW(; }

   | FAQ: |
   | Archives: |
   | Newbie List:   |

This archive was generated by hypermail 2b30 : 06/25/03 PDT