I've taken this example from the quoted source, fr. Anonymus. In my opinion it shows that it should be possible (sorry for the lost alignment):
# /robots.txt for
http://www.fict.org/
# comments to
webmaster@fict.org
User-agent: unhipbot
Disallow: /
User-agent: webcrawler
User-agent: excite
Disallow:
User-agent: *
Disallow: /org/plans.html
Allow: /org/
Allow: /serv
Allow: /~mak
Disallow: /
The following matrix shows which robots are allowed to access URLs:
unhipbot webcrawler-excite other
http://www.fict.org/ No Yes No
http://www.fict.org/index.html No Yes No
http://www.fict.org/robots.txt Yes Yes Yes
http://www.fict.org/server.html No Yes Yes
http://www.fict.org/services/fast.html No Yes Yes
http://www.fict.org/services/slow.html No Yes Yes
http://www.fict.org/orgo.gif No Yes No
http://www.fict.org/org/about.html No Yes Yes
http://www.fict.org/org/plans.html No Yes No
http://www.fict.org/%7Ejim/jim.html No Yes No
http://www.fict.org/%7Emak/mak.html No Yes Yes