Back in the old days (i.e. a few months ago) a litmus test for MO questions was:
Is this of interest to research mathematicians?
These days, the litmus test has changed to:
Is this of interest to research mathematics?
There are many reasons for the change, not the least of which is the existence of our cousin site math.SE. This is an example of a question which might pass the first test but not the second. What do we think of this gradual change? Are there still people who use the old litmus test?
]]>I would also lump http://mathoverflow.net/questions/38652/collaboration-graph in with this.
But let me make it clear that I also find the genealogy graph (and collaboration graph) fascinating - as anyone who visited Sheffield a couple of years ago will know! Indeed, from my graph: http://www.math.ntnu.no/~stacey/HowDidIDoThat/Random/sheffield.html, it's clear that there are several who do not go back to Gauss or Euler. Some, it's true, may be due to a lack of knowledge but some are confirmed.
I did a similar graph with the department collaborations. That is more reliable since one can use MathSciNet as a definitive source of information. I haven't written up all the details, but of course anyone who actually wants to produce such a graph is welcome to contact me.
]]>