Security, DOS, etc, firewalling...
We have already some attempts to solve this notes.
tcp_wrappers is one (I see it as temporary solution);
rlinetd; xinetd.
So we can see that people agree with that "plain" inetd is
not perfect. But that inetd replacements are not perfect too
(at least as I can see now) -- for example, different config file
is a great disadvantage, as many programs expect to see inetd.conf,
and, for example, xinetd does not have ability to deal with
RPC services.
And as I can see, there is no "inetd maintainer" (netkit).
I'm, for example, can "enhance" inetd "as I think right" (ala "Right Thing (tm)").
But is that will also right for others?
Plans include:
integration with tcp_wrappers;
allow dest-based connections (e.g. localhost.telnet instead of *.telnet)
upper limit on concurrent connections/load
Both xinetd and rlinetd (interesting one, but only for Linux, but I use Linux also)
have that ability.
This also can be of interest to other people in (at least Linux) community.
--
To unsubscribe:
mail -s unsubscribe [EMAIL PROTECTED] < /dev/null