On Tue, Nov 13, 2001 at 08:04:41AM -0500, Sam Hartman wrote: > >>>>> "Francesco" == Francesco Lovergine <[EMAIL PROTECTED]> writes: > > Francesco> All this programs provide a RADIUS server for > Francesco> acct/auth. Generally the daemon is named radiusd and > Francesco> uses standard portsi 1645/1646, but this is not a major > Francesco> constraint. This has the same goal of a vp like httpd: > Francesco> they all provide (almost) the same services, but not > Francesco> necessarily the same commands, files or so on. > > Francesco> -- Francesco P. Lovergine > > > Except in the case of httpd, on a Debian system, I can guarantee that > in the default configuration files I stick in /var/www will be served. > > How would this VP be useful? >
Not configuration files, only web pages are in /var/www, and not necessarily all functionalities could be present (cgi, server-side-include, etc.) for files there (in fact a new httpd-with-cgi is proposed). And not all other features (eg. ip/named virtual hosting). So, also in the case of httpd: How is that VP useful :) ? Or in the case of telnet-server: what are your certainties? That it provides a telnet protocol server on port 23/tcp, and that's all. In the case of a radius-server you could be sure that basic RADIUS protocols (as specified in a couple of rfcs) are provided. So, if telnet-server exists as a VP, why not a radius-server? It's better than the current status anyway, i.e. conflicting binaries (and ports) are provided and none can install a couple of radius-servers AFTER downloaded them due to this... IMHO a VP is the only clean way to fix this status of things. cheers -- Francesco P. Lovergine
