E-mail is sent and web pages are found through the use of domain names. Statistically, there are now 168 million domain names, with 77 million of them ending with those 3 letters .com. For this to work, your PC connects to a local ISP (Internet Service Provider)'s network. You then have access to his Domain Name System Server - DNS Server for short - software on a computer that translates a host name you send it e.g. www.google.com into a corresponding IP address (72.14.207.99). This Internet Protocol address specifies first the network, and second the host computer (similar to the way a phone number works). If the DNS server doesn't currently know the host name, it then endeavours to connect (within a second or so) to an authoritative DNS Server that does.

Ultimately, that server is one of 900 or so worldwide servers appointed by the Internet's 280 TLD authorities - Top-Level Domains. E.g. if the domain name ends in .au, then the Australian auDA is the authority for that TLD zone. If it ends in .cn, then the China Internet Network Information Center is the authority. If the name ends in .com or .net (and with no 2 letter country suffix) then the US-based Verisign Global Registry Services is the authority for those 2 TLD zones. And so on. Your request is successful when firstly the name is found, and secondly acknowledged by the nameservers on the network which actually hosts the pages, files & emails sent to that domain. At this point your DNS server caches ("stores") that name and IP address for subsequent requests, for perhaps 48 hours or so. After that, it empties the name & IP address from its cache. This cache minimizes requests made on the authoritative DNS Servers, but also ensures it won't be out of date for more than 48 hours or so on any domain. And of course this only happens if the domain changes hosts.

And if a TLD server is unknown, it can be found in a small official root zone file maintained by Verisign under contract to US. Dept of Commerce & published by 12 Root Name Server operators. Your DNS Server caches the TLD address also for 48 hours or so. And to further reduce Internet traffic, a PC caches a copy of the page in a folder on its hard drive known as "Temporary Internet Files", and normally only downloads a fresh copy if the "Last Update" date on the page is newer than the date in the cache. A proxy server similarly caches a copy of the page for every PC on its network. For more reading, click here.

Now, if looking up details for one of the "open" .au domains, you, as an individual, can go to Ausregistry. This database provides the IP addresses of name servers for all the domains within the five "open" 2nd level domain (2LD) space i.e. ones ending in .com.au, .org.au, .net.au, .id.au or .asn.au (asn association). Note - since 2002 when it was appointed, AusRegistry has never dealt directly with the public in registering domains. Commercial registrars carry out this task , thus preventing potential "conflict of interest" situations within AusRegistry. Then, with regard to "closed" government 2LDs, Netregistry is the registrar for .gov.au and Education .au for .edu.au.

Some Background: On Oct 25th 2001, auDA (a Government endorsed body) became the authorized Domain Administrator for the .au TLD. They then appointed AusRegistry in July 2002 on a 4 year term and this was renewed in 2005 for a further 4 year term starting July 1st, 2006. Prior to auDA and AusRegistry, Robert Elz at Melbourne University administered the whole .au TLD at no charge from March 1986. In 1996 he delegated an exclusive commercial license for the .com.au 2LD to a startup company, Melbourne IT, who made a good profit over the next five years. Click here for details.

Example - Stephen Williamson Computing Services
So, by going to AusRegistry, we learn that host computer(s) for the domain swcs.com.au can be found by accessing IP Addresses 202.3.100.8 and 202.3.100.10 — ns1.webmetrix.net.au and ns2.webmetrix.net.au — name servers managed by Greg Smith at Webmetrix. WebMetrix is a company that has Fast Ethernet links on the AAPT backbone running directly between Brisbane, Sydney and Melbourne and is also peered with many local Internet Service Providers. Next, we learn that the commercial registrar for swcs.com.au is Melbourne IT in Melbourne (who acts on behalf of the owner of the name i.e. Stephen Williamson Computing Services Pty Ltd).

This means that swcs.com.au is currently hosted on the Webmetrix network, who provide a virtual Web hosting service capable of hosting numerous domains transparently. Thousands of different domains might in fact share the same processor (with pages being published in different folders). If the Internet traffic grows too heavy on the Webmetrix network, the swcs domain may in the future require its own, dedicated network & name servers. (This situation has not yet been reached).
Click here to look up the IP address for any other domain.

Your PC now uses this IP address to form packet headers for routing data to & from the host

It began in 1969, when the Advanced Research Projects Agency at US Dept of Defense, working with Leonard Kleinrock at UCLA, built the world's first packet-switching network - the ARPANET. Using four Honeywell computers (as routers) & the Telnet protocol, Bolt Beranek and Newman connected 3 university mainframes at 56 Kbps:- first an SDS at UCLA, then an IBM at UCSB & DEC at UU to an SDS at Stanford Research Institute near San Francisco. Email came in 1971, and the file transfer protocol (FTP) in 1973. By then, 40 mainframe hosts were connected, including satellite links to Norway & London. That year Robert Kahn & Vinton Cerf started work on an Internetwork Transmission Control Protocol using set port numbers for specific uses. IP addresses and port numbers thus form "sockets", connecting clients to services. Request for Comments (RFCs) for these TCP/IP standards and issuing IP address blocks to assignees was overseen by Jon Postel, until his death in 1998, director at Information Sciences Institute at the University of Southern California in LA. Meanwhile from 1972, every host was being sent regular updates of HOSTS.TXT — a readable list of all host names with name-to-address mapping — by the Defense Dept's Network Information Center (NIC) at Stanford.

With IPv4 in 1980, the National Science Foundation created a core network for institutions without access to the ARPANET. Three Computer Science depts — Wisconsin-Madison, Delaware, Purdue initially joined. Vinton Cerf came up with a plan for an inter-network connection between this CSNET and the ARPANET.

Meanwhile, at the physical level, the Ethernet standard — i.e. a multipoint data communication system with collision detection — had been patented in 1977 by Robert Metcalfe at the Xerox Corporation, operating with a data rate of 3 Mbps. Success with that project attracted early attention and led in 1980 to the joint development of the 10-Mbps Ethernet Version 1.0 specification by the three-company consortium: Digital Equipment Corporation, Intel Corporation, and Xerox Corporation. The company Robert formed to market this was 3-Com. The draft standard was approved by a IEEE 802.3 working group in 1983, becoming an official international standard two years later (ANSI/IEEE Std. 802.3-1985).

Back to the software. On January 1st 1983, the Defense Communications Agency at Stanford split off the military network — MILNET — from their research based ARPANET network, while mandating TCP/IP protocols on every host. And so TCP/IP networking became a standard part of the Unix operating system, with Berkeley Unix — using software written by Bill Joy at UC Berkeley — bringing out the first general release that August. And thus the ARPANET and CSNET grew, adding gateways for any network worldwide — regardless of that network's internal protocols — allowing these original core networks to remain intact. Accordingly defense contractors' networks, the university network known as BITNET, later large private networks - CompuServe in 1989, and AOL in 1994, all of whom used different protocols, established gateways. Many date the true arrival of the Internet as 1983. In that same year, the University of Wisconsin established the first Name Server — a directory service that looked up host names when sending email on the CSNET.

In September 1984, taking this to the next logical step, Stanford replaced the HOSTS.TXT file with the Domain Name System, establishing the first of the Top Level Domains — .mil .gov .org .edu and .com. In 1985, with 100 autonomous networks now connected, registration within these TLDs commenced. Click Here for a list of the first domains. Country code TLDs were then appointed shortly afterwards. Stanford now administered:
1. A small centralized Root Zone file providing authoritative IP addresses for each country code TLD.
2. Name Registries providing IP addresses of name servers for each domain in the generic TLDs - ie .mil .gov .org .net .com
with Jon Postel administering the .edu TLD & the .us TLD.

In 1987, the first email from China to Germany occurred, via the CSNET. In early 1988 as 200 networks came online, IANA - the Internet Assigned Numbers Authority under Jon Postel received a formal 10 year funding contract at the Institute from the Defense Dept. In 1989 with 500 networks online, the very first commercial ISP The World provided indirect access via dial-up UUCP accounts. This first downloaded a file from the Internet to the ISP, then the ISP used the Unix to Unix Copy command to copy the file to a paying customer's computer. In 1990 our history of internet search engines starts with Archie designed to index files in "archives". In 1991, with 4000 networks now online, domain name registration entered the public arena - free of charge, as administration of the Network Information Centre moved from Stanford to a small, but capable private-sector firm Network Solutions.

World Wide Web

In Feb 1991, Tim Berners-Lee, programming a NeXTcube workstation in the European Laboratory for Particle Physics, held a Seminar for a World-Wide Web and with the help of Paul Kunz at Stanford, demonstrated in France on Jan 15, 1992 the first web browser and offsite server. The server downloaded file(s) using a new protocol he called HTTP - Hyper Text Transfer Protocol, forming a page that was formatted by the browser using a Markup Language called HTML. This allowed Hyper Text on a page to link to another page on a computer running his web service anywhere in the world, by including in the markup language the linked page's URL - Uniform Resource Locator - its domain name, path & filename. Thus the user could upload a request for that new page by simply selecting the link. But the truly standout feature was that the user could also type in and upload search phrase(s) to a database hosted on that remote computer, a feature that Google was to later utilize on their database servers with devastating effect. But back in this year, Jean Polly now published the phrase 'Surfing the INTERNET'.

Meanwhile in Amsterdam, Holland, Réseaux IP Européens (RIPE) in 1992 received the first allocation of a block of IP addresses from IANA that enabled it to become a regional IP registry, independently responsible for the physical location of each network.
And in Australia, Geoff Huston from AARNet "applied to IANA for a large block of addresses on behalf of the Australian network community .... because allocations from the US were taking weeks .... The address space allocated in 1993 was large enough for over 4 million individual host addresses .... The Asia-Pacific Network Information Centre (APNIC) then started as an experimental project in late 1993 (in Tokyo), based on volunteer labour and donated facilities from a number of countries. It evolved into an independent IP address registry ... that operates out of Brisbane" - R.Clarke's Internet in Australia

Back in the U.S.

In 1993, the Dept of Defense ceased all funding of non-military registrations. The US National Science Foundation, who had been providing Internet funding to the universities took over. Their physical network, the NSFNET, had been built in 1986 with high speed links around the country, taking over backbone duties from the much slower ARPANET (which had then been decommissioned in June 1990). Now, working with IANA, the NSF set up the Internet Network Information Center - InterNIC as the new Internet governing body. AT&T were appointed for directory, database & (later) information services, with Network Solutions providing registration services.

This year, students and staff working at the NSF-supported National Center for Supercomputing Applications (NCSA) at the University of Illinois, launched Mosaic, a free web browser with versions for Unix, Windows & Mac. As a graphical browser, it provided home users a much richer experience than running a command line shell. Initial costs for these dial-up SLIP / PPP connections were $US175 per month. But competition between ISPs & new technology meant that over the next two years prices plummeted rapidly. Although Mosaic was a fairly basic program by today's standards, its graphical nature introduced huge numbers of "unskilled" users to the web. The .com domain space was now receiving a massive number of new registrations. By the end of the year there were 20,000 separate networks, involving over 2 million host computers and 20 million individual users.

In 1994 the immensely successful web directory Yahoo was created. Southwest Airlines offerred the first e-tickets for passengers. At the start of 1995, with 30,000 web sites now online, Netscape browser was released. It cost non-academic users just $US39 and gained up to 80%+ market share.

On April 30, 1995 the NSFNET was decommissioned. In its place, a number of large private networks (or Tier 1 networks) progressively formed a new Internet backbone via Internet Exchange Points (IXPs). These required mutual peering agreements - i.e. allowing network traffic exchange to occur without cost.

By mid 1995 there were 50,000 autonomous networks online. Figures were now doubling yearly. In August, Microsoft released Internet Explorer 1.0 free in a Windows 95 add-on pack. In December, Netscape added support for Javascript as a brand new web language. Then the first search engine to allow natural language queries — Alta Vista was released. And InterNIC imposed their first yearly registration fee of $50 per domain name.

In 1996, the first cable modems lifted download speeds on the Internet from 56 kbps over telephone wire to 1.5mbps over fibre optic cable (i.e. over 25 fold). The massively popular online chat room and games-playing company — America Online (or AOL), became an Internet Service Provider. In 1997 AOL bought out Compuserv, a private company which in 1969 had been the first major commercial online service company in the US. And this same year DSL (Digital Subscriber Technology) which employed high frequency signalling over telephone wire was retooled for Internet access, offerring initial download speeds of 768kbps. In 1998, the movie "You've Got Mail" was released with Tom Hanks and Meg Ryan and featuring AOL as their ISP.

ICANN

Foreign Characters in Domain Names
The Domain Name System service had been originally designed to only support 37 ASCII characters i.e. the 26 letters "a - z", the 10 digits "0 - 9", and the "-" character. Although domain names could be generated in English using upper case or lower case characters, the system itself was case-insensitive — it always ignored the case used when resolving the IP address of the host. Then, in 2003 a system was released to allow domain names to contain foreign characters. A special syntax called Punycode was developed to employ the prefix "xn--" in the domain label and encode any foreign characters within the label from Unicode into a unique ASCII address — e.g. www.中.com would thus find itself encoded as www.xn--fiq.com. In 2008, with a view to also having TLDs in foreign scripts, ICANN released 11 new TLDs using the word for "test" in 11 different languages, and letting DNS Servers confirm that they were able to handle a .test TLD in each of those scripts. Click here to view the results. Click here for the latest news on TLDs and the new ones coming.

Stephen Williamson Computing Services 202.3.100.8 and 202.3.100.10
Now that we have some background, let's learn more about IP address allocation in Australia.
By clicking on www.iana.org, or ipindex.homelinux.net we learn that 202.0.0.0 - 202.255.255.255 i.e. 16 million addresses were allocated to APNIC Asia-Pacific Network Information Centre.
And by clicking on APNIC we learn that IP Addresses 202.3.96 - 202.3.111 (which is 4000 addresses) were allocated to WebMetrix in Brisbane.

APNIC is a nonprofit organization based in Milton, Brisbane, since 1998, having started as a pilot project in Tokyo in late 1993. Today the majority of its members are Internet Service Providers (ISPs) in the Asia-Pacific region. Naturally, China is involved. In Australia, Telstra and Optus are two national ISPs. In 1999, Optus introduced Cable modems offering high speed connections using on/off light pulses transmitted over fibre optic (television) cable. In 2000, Telstra introduced ADSL modems providing broadband (high-frequency) signals over copper (telephone) wire. With both cable and ADSL, both companies allocate one permanent or "static" IP address that is linked to the MAC hardware address of the network adapter on the customer's PC or on their router. For customers with slower Dial-up modems that utilize the voice-frequency range over telephone wire, they both have a pool of IP addresses which get temporarily / "dynamically" linked to the customer's telephone number as required.

Click here for an interesting article on commercial peering in Australia - the establishment of the so called "Gang of Four" in 1998 - Telstra, Optus, Connect (AAPT), and Ozemail (now iiNet).

Now, to summarize. IP addresses are used to deliver packets of data across a network and have what is termed end-to-end significance. This means that the source and destination IP address remains constant as the packet traverses a network. Each time a packet travels through a router, the router will reference its routing table to see if it can match the network number of the destination IP address with an entry in its routing table. If a match is found, the packet is forwarded to the next hop router for the destination network in question (note that a router does not necessarily know the complete path from source to destination — it just knows the MAC hardware address of the next hop router to go to). If a match is not found, one of two things happens. The packet may be forwarded to the router defined as the default gateway, or the packet may be dropped by the router. (In the language of TCP/IP, a gateway is a router.) When packets are forwarded to a default router, it is in the belief that the default router has more network information in its routing table and will therefore be able to route the packet correctly on to its final destination. IP101: All about IP Addresses. To view an example of these router "hops", click here.

Note, also, that only a few thousand characters are generally involved in each transfer packet of data. If many transfers are necessary to transfer all the information, the program on the sender's machine needs to ensure that each packet's arrival is successfully acknowledged. This is an important point: in packet switching, the sender, not the network, is responsible for each transfer. After an initial connection is established, packets can be simply resent if that acknowledgement is not received.

Click here to download and save a program that will look up any IP address or Host from the command line prompt. Click here for the documentation.

HTML - Hyper Text Markup Language:

You specify markup commands in HTML by enclosing them within < and > characters.
E.g. <a href="http://www.swcs.com.au/about us.htm" target="_blank"> Load SWCS Page</a>
This is a markup command and text.

• <a href means to anchor a hyper reference (or start drawing an underline under the words Load SWCS Page indicating a linked section of text).
• </a> means to stop drawing the underline for the link.
• http:// is the access method - it means to use the hyper text transfer protocol built into the computer's web browser program. It sends a request to the web server (on the Internet) that hosts the www.swcs.com.au domain, tells it to go to the aboutus.htm file, download it to the client computer, then the browser will display it as a web page.
• target="_blank" means the target for displaying this page is a new, blank page (on the client computer). Otherwise the page will simply replace the current page and you will need to press the BACK button to return to it.

Other Examples:

•  
  <img src="steveandyve2.jpg" align=left> will load the jpg file (in this example it is stored in the same folder as the web page) and align it on the left so that the text that follows will flow around it (on the right). If the align command is omitted, the text will start underneath it (instead).
  

• <H1>(or <H2> or <H3>) cause the Heading characters that follow to be large and bold (1 is highest, 5 is lowest). Close with </H1>(or </H2> or </H3>)
• Normally all text wraps within the window. <br> breaks or starts a new line. <p> starts a new paragraph.
• <b> starts bold. </b> closes bold. <i> starts italic. </i> closes italic.
• Multiple spaces are ignored. Enter &nbsp; to enter a non-breakable space (not to be ignored, and not to be used as a place to wrap text).
• <ul> Start indenting an unordered list. Use <li> for each point (a dot as you see in these points). Use </ul> to end.
• <ol> Start indenting an ordered list. Use <li> for each point (a sequential number). Use </ol> to end.
• The individual characters used to specify markup commands require their own group of characters when displaying themselves on a page. E.g. &lt; displays a < character. &gt; displays a > character. And &amp; displays an & character.

Most of these examples can be seen on this page that you are viewing. To see the text file that is the source of this page, right click on the mouse, then click View Source.

References:

1. Gilster, Paul (1995). The New Internet Navigator. (3rd ed.) John Wiley & Sons, Inc.
2. Roger Clarke's Brief History of the Internet in Australia - 2001, 2004
   
3. Goralski, Walter (2002). Juniper and Cisco Routing. John Wiley - The Internet and the Router - excerpt
   
4. History of Computing (with photo links) and the Internet - 2005
   
5. The brief History of Internet 1957-1996 by webhostingsearch.com
   
6. History of the Internet - Wikipedia - 2008