Regarding the other issues pointed out in Mark's mail, here's a quick
analysis:

- Error 2, ",statistics": This seems to be due to the ,statistics
  meta-command expecting the result of invoking `repl-gc-stats' (which
  is based on `gc-stats') to carry several fields that are not present
  with current Guile.  For example, the ,statistics code references
  these fields in the gc-stats alist: `cells-allocated',
  `cell-heap-size', `cells-marked', `cells-swept', etc.  None of these
  are present anymore:

  scheme@(guile-user)> (gc-stats)
  $1 = ((gc-time-taken . 0) (heap-size . 12034048) (heap-free-size . 167936) (heap-total-allocated . 12935072) (protected-objects . 95) (gc-times . 7))

  So the solution would be to significantly cut down the implementation
  of ,statistics, to make use only of the fields that are currently
  present.  I assume that none of the now-missing fields can be
  reinstated, since BDW-GC doesn't provide that information, right?  If
  this is correct, I can give this bug a whirl, hopefully providing a
  fix by Wednesday.

- Errors 3-11: These are not really "errors" in my view, but rather
  suboptimal presentation of the doc-strings associated with the
  meta-commands by ,help.  Probably the docstrings should be changed not
  to duplicate the summary (which is the first line) in the rest of the
  docstring; another possible way to fix this would be to not show theq
  summary if there's more documentation after the first line.  Thoughts?

[0] http://lists.gnu.org/archive/html/bug-guile/2011-01/msg00070.html

Regards, Rotty
-- 
Andreas Rottmann -- <http://rotty.yi.org/>