On Fri, 5 Jun 1998, David McDonald wrote:
> http://www.nmt.edu/~shipman/z/nom/homepage.html . Follow the links 'The Bird
> Banding Lab (BBL) system' and 'Rules for forming BBL codes'. He strongly
> recommends a six letter code to minimise collisions. This makes sense for a
> large list such as Australia's national list.
I've set up a web page which allows experimentation with coding systems
for bird names. You indicate the letter positions you wish to use
in constructing name. It gives you the codes for all the birds on
Christidis and Boles's checklist and summarises any collisions. The URL is:
http://www.cse.unsw.edu.au/cgi-bin/birdcode.cgi
It not hard to come up with 6 letter systems with few collisions.
4 or 5 letter codes for Christidis and Boles are more problematic.
Robert Berry wrote:
> a 5 letter key would suffice for all the insects in the universe
Maybe not. 26^5 = 11881376. There is interesting speculation about
extreme tropical rainforest arthropod diversity stemming from the work
of Terry Erwin. One estimate suggests there may be 8,000,000 species
of rainforest canopy beetles alone.
Andrew Taylor
|
