stupid newbie question about source directory structure

["Eric Smith" <eric-ecd at brouhaha dot com>]

Even after reading the eCos Component Writer's Guide, I'm still
confused about the utility of package versioning.  It seems to me
that having a version as the name of a directory in the path causes
more problems and confusion that it solves.  For instance, it makes
applying patches difficult.  I've resorted to doing this:

for f in `find . -name v2_0b1`; do echo ln -s $f `dirname $f`/current; done

Am I overlooking some clever tricks to avoid the need for this?

On other operating systems I've used, it was certainly possible to select
between versions of components, but you could only install one at a time
(unless you wanted to manage symlinks yourself), and I never had any
problem with that.  Is the eCos way really that much of a win?

Sorry, I know this is a newbie question and that it's rude for me to
wander in here and criticize how things are set up.  I'm sure there
must be some value to this arrangement that I can't see because I'm
too accustomed to doing things a different way.

Thanks,
Eric