Pretend you've got a CB radio, with "handles," 10-codes, many channels to pick from, and then let it use text rather than voice transmission.
IRC is "superior" to genuine CB Radio in that it doesn't provide garbled transmissions (beyond spelling mistakes, and messages that are genuinely incoherent!), is not limited by geographic considerations, and provides something of a record that makes it a little easier to follow conversations that have been going on.
It's really quite valuable when they hold a revolution in Moscow, as you can potentially get play-by-play details of what's going on.
Unfortunately, at least 90% of the time it's a simply stupendous timewaster that seems populated by people who can neither spell nor type correctly, but think that they're k00l d00ds.
It also appears that a wide set of infrastructure has grown up surrounding the issue of how to "control" a particular IRC channel against "opponent hackers," which seems to me to defeat the purpose of having a public channel.
And I get the impression that the primary use of it these days is to allow people to "safely" talk dirty whilst thinking they remain anonymous. There seems to be a whole lot of paranoia around about "Internet chat rooms" being a place where sexual predators seek out targets. That is hopefully mostly paranoia, with minimal reality to back it up. But who knows...
A programmable IRC client for Unix written in Perl
An X11 based IRC client
An XML messaging system intended as an "open source" alternative to various proprietary "instant messaging" systems.
Gossip is an instant messaging client for GNOME layered on top of the Jabber protocol.
Gaim is a multi-protocol instant messaging client for Linux, BSD, MacOS X, and Windows. It is compatible with AIM (Oscar and TOC protocols), ICQ, MSN Messenger, Yahoo, IRC, Jabber, Gadu-Gadu, and Zephyr networks.
Kopete - KDE "Instant Messenger"
Silky - SILC Client
The Lightweight Directory Access Protocol (LDAP) is a protocol for accessing online directory services. It was initially used as a front-end to X.500, running directly over TCP, but can also be used with stand-alone and other kinds of directory servers.
Novell has set up software whereby one can query Novell Directory Services via LDAP, providing user and host data;
Lots of systems have implemented LDAP servers to export directory information so that one can construct generic LDAP clients to query directory-like information.
LDAP is intended to be useful for things like user authorization. Combine with Section 2.1 for authentication.
For instance, see NRG4U - The big qmail picture; a system to get qmail to get user account information from an LDAP server.
Notable LDAP applications include:
Email "address books"
User authentication systems such as PAM
Email routing
Some of the Debian folk would like to use LDAP to "publish" information on what packages and versions thereof are installed on a system. That is quite a different use, but seems to fit reasonably well with the intent of LDAP.
A guide to LDAP that tries to answer not just HOWTO, but WHYTO.
OpenLDAP Faq-O-Matic: Directories vs. Relational Database Management Systems
SAMBA PDC LDAP howto - Using LDAP to manage SAMBA authentication
LDAP HOWTO - Guide to using LDAP for authentication via PAM, NSS, DNS, MTAs
Sendmail and LDAP HOWTO - using LDAP for mail address lookups
Sympa - Mail list manager which extracts email addresses from LDAP queries
This document is about building an LDAP server with a PostgreSQL database as backend instead of the standard Berkeley DB.
I have set this up to validate that it functions. It is not enjoyable to configure, and I have not tried to set up such a server as a "production" server, as the complex combination of configuration appears rather fragile.
OpenLDAP Faq-O-Matic: The SQL backend (How do I setup/configure back-sql?)
This proposes using Git to provide backup archival of LDAP data. LDAP data tends to be read a lot more than it gets changed, so this offers a hopefully-cheap long term history.
If you are suffering from spam problems, or other more sinister problems, you may wish to find more information about who's behind a particular Internet host. Here are some relevant tools:
Search VeriSign whois records in the whois section of netsol.com
PIMMIT - an IRIS implementation
IRIS is an XML-based IETF proposal for a sort of successor to WHOIS. PIMMIT is an implementation of this...
PFIR Statement on VeriSign/NSI "Site Finder" and Domain Abuse
The implementation of an " A Record wildcard" for .COM and .NET domains has been somewhat controversial, as it means that any non-existent .COM/.NET domain resolves to a specific "SiteFinder" site.
Some have felt it necessary to " take measures" against this...
Offers free DNS services as a tertiary DNS server. They support TXT records, which is useful if you're using IPSEC or OpenSWAN/LibreSWAN.
They offer DNS resolution for dynamic IPs, and mail forwarding to to help deal with those irritating ISPs who block port 25.
Offers free (as well as fairly cheap paid) DNS hosting, including Dynamic DNS, which is mighty useful for providing a named "target" for ISP-based connection to systems at home.
Unfortunately, ODS has stopped accepting new registrations, and plans to close down in 2016.
Non-free DNS hosting service.
Non-free DNS hosting service.
Offers a mix of free and paid DNS services.
Free services generally involve using subdomains under one of their pre-provided domains; to use your own domain requires the paid service.
Non-free DNS hosting service.
ez-ipupdate is a small utility for updating your host name for various dynamic DNS services
Software for providing a Dynamic IP DNS service; compatible with ez-ipupdate, usable with a number of DNS hosting services.
Implements an authoritative (e.g. - not a cacheing server) DNS server, supporting DNSSEC, AXFR, IXFR.
There are also some providers of "managed DNS" services, which are rather more expensive, but with considerably more promises offered concerning uptime and such like.
In keeping with RFC 1178 and RFC 2100, here are a variety of naming systems to help inspire the naming of hosts.
To assist the poor administrator at remembering how to spell the names, it is undesirable for names to be cryptic or confusing.
twcsds007 is a pretty terrible name, as it is cryptic.
Naming machines after Unix commands, such as ls, cat, dd would be pretty horrible.
Naming machines "up" and "down" has a likelihood of turning into a "Who's on First?" joke.
Naming hosts after The Three Musketeers may turn out poorly in a user community unfamiliar with French.
I observed this when people were horribly confused at the names of locations of famous-ish Canadian World War I battles, such as Somme and Ypres. (To be fair, I recall hearing that the Canadian troops at the time found pronouncing "Ypres" troublesome!)
As funny as stripper names, or euphemisms for body parts may seem, it's liable to offend someone.
Defined in RFC 2131, Dynamic Host Configuration Protocol indicates a way for hosts to ask a central server what configuration, notably including IP addresses, they should use.
Commonly used is ISC DHCPd.
It is common for routers to include a DHCP server, but I have set up a failover-aware configuration on a couple of my favorite servers at home; see DHCP Failover for instructions as to how to set this up.
There is not universal agreement that ICANN is the ideal body to manage domain names. (That may be putting it mildly!) As a result, there are some folks running an " alternative" namespace resolution scheme. Some of them might be a little kooky...
Where various software including EPP implementation(s) may be found...
Do you know where your ports are?
There are many Unix system services that typically are run as root for the sole reason that they control ports with ID numbers under 1024.
This is a dumb thing. Sendmail has been noted for having security holes due to the complexity of its configuration; the threat is magnified by the fact that it also typically has to run as root. Qmail was specifically written to not run as root, and works very well as a Sendmail replacement.
Moreover, this port requirement is only enforced on Unix systems, which means that it doesn't provide any real authentication e.g. - the fact that a service is provided on port 441 doesn't necessarily mean that the service is "privileged" in any way on its own host and thus even slightly trustable.
In short, services that can be run without root access ought to be.
NEC's Socks Server can be used to set up "secure" data channels between computers. It is most often used to help implement firewalls.
There is a lot of software available for securing communications by cryptographic means.
Quite commonly, really interesting new protocols are released on dates resembling the first of April. Few such protocols are implemented, with RFC 1149 as the one so bizarre that a computer club and a pigeon club got together to make it work...
Sometimes people use wikis...
The Blocks Extensible Exchange Protocol Core (BEEP) - RFC 3080
This memo describes a generic application protocol kernel for connection-oriented, asynchronous interactions called BEEP.
DIME: Direct Internet Message Encapsulation
A lightweight binary message format that may encapsulate multiple application-defined payloads of arbitrary type and size into a single message construct.
Payloads are described by a type, length, and possibly an identifier. URIs and MIME media types are supported as type identifiers, and payloads might include nested DIME messages.
DIME is a message format, not a protocol.