error-prone feature?

Well, that is in all good texts on R, together with the 
solution: drop=FALSE.  See ?"[" for the on-line details.

`S Programming' says

  S programmers have often overlooked these rules, which can result
  in puzzling or incorrect behaviour when just one observation or
  variable meets some selection criterion.

I suggest you get yourself a copy and read it.

  while looking why the cclust(cclust) doesn't work for 1-dimensional data,
  I've found unpleasant behavior in semantics of R. Indeed:

    is.matrix(matrix(cbind(c(1,2,3,4)),ncol=2)[1:2,])  == TRUE

  but:

    is.matrix(matrix(c(1,2))[1:2,]) == FALSE

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595