[r-t] Unclassified (Was: ?CCBR meeting - Methcom proposals)

[John Harrison]

In article <alpine.LRH.2.02.1606030144180.26826 at sphinx.mythic-beasts.com>,
   Richard Smith <richard at ex-parrot.com> wrote:

> The key, I think, is to get away from the idea that all methods should
> fall into a well-defined class.  Once we accept that a method might be
> unclassified, we can tighten up the definition of the classes to
> reflect their normal use.

That sounds a more flexible approach, and more consistent with the ability
of the classification system to evolve in the light of developing trends or
concepts.  

'Unclassified' should mean 'not yet classified', ie does not fit into any
current classification.  Thus what now looks like a lone exception or three
might later be absorbed into a new class if lots of other methods emerge
that share with it what by then is recognised as a useful distinction.

-- 
John Harrison
Website http://jaharrison.me.uk