Graphs with n edges

[Frank Ellermann]

Dr. Arne Ring wrote:

> How should we proceed with this ?

AFAIK there's yet no "dupe/dead" check except from a lookup:
A046901 is referenced by A002905 and A066951, adding "dead"
to A046951 also requires corrections of A002905 and A066951.

Here's how I would do this (in a mail to Neil), simple form:

~~~ cut ~~~
%I A046091
%S A046091 1,1,1,3,5,12,30,79,227,709,2318
%N A046091 Connected planar graphs with n edges.
%K A046091 dupe
%C A046091 This is an erroneous dupe of A002905.
%Y A046091 A002905,A066951.

%I A002905 M2486 N0985
%S A002905 1,1,1,3,5,12,30,79,227,710,2322,8071,
%H A002905 E. W. Weisstein, <a href="http://mathworld.wolfram.com/PlanarConnectedGraph.html">Planar Conected Graphs</a> (MathWorld).
%H A002905 E. W. Weisstein, <a href="http://mathworld.wolfram.com/PythagoreanTriple.html">Pythagorean Triple</a> (MathWorld).
%N A002905 Connected graphs with n edges.
%Y A002905 Cf. A000664.

Reference to A046091 deleted, links from A046091 copied and
titles corrected.

%I A066951
%S A066951 1,1,3,5,12,28
%e A066951 Up to five edges, every planar graph can be drawn with edges of length 1, so up to this point the sequence a\
  grees with A002905 (connected planar graphs with n edges) [except for the fact that that sequence begins with no edge\
  s]. For six edges, the only graphs that cannot be drawn with edges of length 1 are K_4 and K_{3,2}. According to A002\
  905 there are 30 connected planar graphs with 6 edges so the sixth term is 28.
%Y A066951 Cf. A003055, A070860, A002905.

Reference to A002905 added.  A046091 replaced by A002905 twice.

~~~ end ~~~

In fact I would simply edit the complete sequences and resubmit
them as EDITED, because the huge example line %e A066951 in the
form shown above (split) cannot be joined in my excuse for a
mail reader (i.e. netscape 3, this line would be too long).

But then any delay in this correction was good:  There are two
links from A046091 to MathWorld, and so I considered to copy
these links from A046091 to A002905.  Checking this I found a
reverse link from MathWorld to A046091, so we shouldn't simply
delete A046091.

The 2nd link in MathWorld also mentions, that A046089 - A046091
are a triple, so it's obviously not okay to delete it now, but
keep it as a "dupe" until the MathWorld sections are updated.
I've now copied the links to A002905 in the example (see above).

Please check this again, I have no idea about the background,
I only know how to manipulate databases in a consistent way ;-)

                              Bye, Frank