Policy issue in the VC manual -- recommending CVS?!?

["Eric S. Raymond" <esr@snark.thyrsus.com>]

The current VC manual recommends RCS for small single-file projects,
which is sensible. But it then recommends GNU Arch or CVS for
distributed projects, which is not,

Recommending GNU arch is dubious considering Arch's poorly-maintained
and poorly-documented state.  If I'm not mistaken, Arch has been
effectively moribund since about 2003.

Recommending CVS is well beyond dubious into outright ridiculous.
It's 2007; the many and painful problems with CVS have been well known
for a decade and and it is anyway *well* past time to be flogging a
file-oriented VCS to anyone on a distributed project.  Anyone who
thinks they want CVS for a new project should get either Subversion or
their head examined.

No recommendations at all would be better than these.  Who decides
what the manual recommends?  If it's "the last person to care", I'm
going to nuke these in a nanosecond.

What I think we ought to be doing is recommending either Subversion or 
any of the modern 3G systems like Mercurial or git or Bzr.
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>

We shall not cease from exploration, and the end of all our exploring will be
to arrive where we started and know the place for the first time.
	-- T.S. Eliot