I added the remark that finite index subgroups of a profinite group are not generally open, and gave the usual example involving non-principal ultrafilters. I also briefly discussed when a group is isomorphic to its own profinite completion.

]]>Stone Spaces chapter VI, section 2.9 has a general result in this direction.

]]>I vaguely recall that it’s not entirely trivial that pro-(finite algebras) are the same as internal algebras in profinite sets – there’s some restriction on the signature and axiomatisation, I think.

]]>moved the statement that a profinite group is a group internal to profinite sets from the “Examples”-section to “Definition”-section

]]>