[seqfan] Re: A187781 - Number of noncongruent polygonal regions in a regular n-gon with all diagonals drawn

[brad klee]

> I recently undertook to port... Food for thought?

Mathematica has a similar potential for accidental failures as OEIS,
which requires a wrong programmer, and one or more wrong reviewers.

An egoistic line is drawn between bugs and what I call "wrong by 
design". In the second case, it's possible to become committed at 
the executive level to making the type of mistakes that would get 
an earthly person a "B" for their grad. school homework assignment. 

RotationMatrix is the specific example I'm thinking of. It's advertised 
to work for complex spanning vectors, but the source code doesn't 
(other than the dead branch with my ignored bug fix) make any 
reference to the Pauli matrices??

(more info here: https://community.wolfram.com/groups/-/m/t/2753373)

At OEIS there is a "dead" keyword, perhaps useful for preserving 
backwards compatibility? Will A187781 be marked dead or bug fixed? 

I don't know, it's not my place. I think that question goes to Neil 
or Joerg. Their names are on the history as editors and/or reviewers.

As for Mathematica CAS... It usually works for me, and the issue 
I had with canonicalization doesn't bother me that much. When 
there are many different paths to the same answer, there's a 
law of diminishing returns for canonicalization, and different 
users have different expectations. Not much to do about that.

If you feel like your complaint is being ignored, you're welcome 
to take it up with me, or more likely with Bill Gosper. Bill has 
a very high level view of how "The Plan" is unfolding, and who 
needs to be held accountable when it goes off the rails. 

All the best,

--Brad