Re: Some cvs questions

[Larry Jones]

Anne McCaffrey writes:
>
> MIME-Version: 1.0

Please do not send MIME and/or HTML encrypted messages to the list.
Plain text only, PLEASE!

> What I want to know is how cvs is being invoked from the client side?I
> am sure it is not just setting the CVS_SERVER variable on the client and
> .rhosts file on the server.(I just followed the cvs manual while setting
> up cvs)What program on the server is actually issuing the command to run
> the cvs command that the client runs?And where does rsh come in? In some
> of our windows clients, rsh is not installed but cvs works fine.How is
> this possible if cvs exclusively uses rsh for remote communication?

CVS uses rsh -- either the external command (for :ext:) or a built-in
rsh client (for :server:) -- to run a command on a remote system. 
There's an rsh daemon (usually called rshd) that runs on the server that
authenticates the client (that's where .rhosts comes in) and then, if
the remote user is valid, actually executes the command on behalf on the
client.  In CVS's case, the command that is run is a CVS command built
from $CVS_SERVER and other stuff.  Systems with the built-in client
don't need an external rsh command to communicate with the server's
rshd, which is why it works on Windows without rsh.

CVS can also use its own protocol (for :pserver:, :gserver:, and
:kserver:) that doesn't use rsh and rshd but rather requires CVS to be
run as a server on the server system, typically by a server manager such
as inetd or xinetd.

-Larry Jones

Another casualty of applied metaphysics. -- Hobbes