In my other Web pages, I use terms such as server and domain, which I need to define. Here are simplified definitions that might not be as complete or accurate as an experienced computer professional might desire.
To simplify browsing, these definitions appear in a window separate from the pages where the terms are used. You can leave this window open and return to the window that sent you here without having to reload either. However, after this window has opened on one definition, the user who then selects another term will have to select this window to bring it in front.
- backbone
- The worldwide network of Internet systems that connect one ISP to another. Different companies or government agencies — the backbone providers — operate different portions of the network.
- bot
- See crawler
- client
- Software that provides a user with the interfaces — input, menus, displays — to use a server. E-mail and Web-browsing software on a PC are clients.
- crawler
- An automated application (also known as a bot, a shortened form of robot) that seeks and visits Web pages, not to view them but to collect information from them. Some crawlers analyze text content to index Web pages for search engines (e.g., Google, Yahoo). Some crawlers seek E-mail addresses embedded in Web pages in order to build mailing lists for spammers. The Internet Archive uses a crawler to collect complete Web pages for its archive.
- domain
- The name of a server or other entity accessible through the Internet. Examples of domains are www.nwsla.noaa.gov and www.vcnet.com.
- DNS
- Domain name server, a server that translates domain names into IP addresses and vice versa. A DNS looks in a database table — the DNS table (of course) — to perform this task. There are many DNSs scattered across the Internet. While they do not all have identical tables, DNSs are constantly updating their tables as domains move from one host to another.
When a client requests a server to access a domain by name, the server starts by accessing a local DNS. If the domain name does not appear in that DNS's table, the DNS in turn accesses a farther DNS. This continues until either the name is found or the chain of DNSs is exhausted. Obviously, most DNSs contain www.yahoo.com in their tables. Less well-known domains are carried in only a few tables. All domains are carried in the tables at the root DNSs at the ends of the chains.
- host
- The computer on which a client or server operates (is hosted). This is sometimes called the server's platform.
- IP address
- The numeric address of a domain on a host. The address is used to route requests and data over the Internet. It is in the form of four 3-digit numbers separated by periods; leading zeros in the numbers are omitted. Many domains have multiple IP addresses to allow multiple connections at the same time. The IP addresses for www.iswest.com range from 207.178.128.0 to 207.178.255.255.
At home, I have a dial-up modem for connecting to the Internet; and I connect to a new IP address each time. This is a dynamic IP address. At work, I had a dedicated ethernet connection, which gave me a static, unchanging IP address. Often, DSL and cable modem connections are static; but some ISPs assign a new IP address each time a computer with such a connection reboots.
If a domain moves from one host to another, a new IP address is assigned to the domain because IP addresses are associated with a particular host's connection to the Internet. Thus, the old IP address becomes available for reassignment to another domain on the old host.
In addition to domains, other connections to the Internet have IP addresses, including your own computer, which has the IP address
38.103.63.16
- ISP
- Internet Service Provider, the company through which you connect to the Internet. See my Looking For a New Internet Service Provider.
- open relay
- An E-mail server configured to accept and send messages from individuals who have not logged in (i.e., from users who are not on the server's user list). Since open relays appear to be the sources of relayed messages (not merely an intermediary), they are often used by spammers to hide the real source. The configuration that permits an E-mail server to be an open relay generally indicates an error by an ISP's (or other server operator) system administrators.
- platform
- See host.
- POP
-
- Point of presence, the place where your phone line, T1 line, TV cable, et cetera connects to the Internet or where different ISPs connect to the backbone. When using a dial-up modem, the phone number you dial is at the POP.
There are also dial-up pseudo-POPs, which are very similar to call-forwarding (and might even use that capability). If you dial a pseudo-POP, you actually connect to a different phone number. Thus, when I worked in Los Angeles County, I could call my ISP on a local phone number. I would then connect to a POP in Ventura County more than 40 miles away. If I were to dial the actual POP, it would not be a local call.
- Post Office Protocol, the format of incoming E-mail messages held at a mail server so that your E-mail client can access them. Actually, this protocol not only defines the format of the messages on the server but also defines the way your E-mail client communicates with the server to download the messages. Some ISPs do not use this protocol, requiring subscribers to use only the ISP's own proprietary E-mail clients.
- server
- Software that performs a task as requested by a client. In general, a server performs tasks such as searching a database and retrieving E-mail — tasks that require major computer power — while a client performs data-processing activities such as collecting inputs to send to a server and generating displays from data retrieved by a server. Users are very aware of clients but not necessarily of their supporting servers.
Servers generally operate in one of two modes:
- A server may operate continuously, waiting for a request for service from any client that can access it. In this mode, a server may actually respond to several clients simultaneously.
- A server may be started at the request of a client, perform its requested task, and then terminate. In this mode, a server supports only one client, each client starting its own copy of the server software.
Most Internet servers operate in the former mode.
Because many hosts host only one server, the term server is often used for the hardware; this can only cause confusion when the same platform is a host to multiple copies of a server or different servers and might even host some clients. (Good system design, however, generally avoids having clients and servers hosted on the same platform.)
- site certificate
- A file used to establish a secure, authenticated connection between a user's computer and a Web site. A site certificates is used by a bank or other business to assure its customers that their Web site is indeed authentic. It also provides the means to encrypt and decrypt data going back and forth between the user's computer and the Web server.
A site certificate is issued by a certificate authority (CA) for a particular domain. A CA has a root certificate that is used to encrypt part of a site certificate, thereby signing the latter. Sometimes, there might be an intermediate certificate that was signed by a root certificate and that, in turn, signs the site certificate.
For all this to work, the site certificate is installed on the Web server along with any intermediate certificates; and the root certificate is installed in a database contained within the user's Web browser. Most browsers come with a large repertoire of root certificates. Also, for this to work, the Web pages are addressed beginning with https instead of http, the s indicating secure. The URL beginning with https must have the domain that agrees with the domain in the site certificate.
- sniffing
- See the discussion of this term on my "Professional" Web Developers.
- spoof
- Use a fake UA string to make a Web server act as if a browser being used is different from the actual browser (e.g., as if Opera were being used when actually Firefox is being used). Spoofing may be necessary when trying to access a Web site that is incorrectly sniffing UA strings.
- UA
- User agent. The client used on the Internet, usually a browser or E-mail application. The brand and version of a user agent is usually indicated by the UA string, an identifier sent by the user agent to a server when requesting a response (e.g., a Web page, an E-mail message).
At the time this page was last updated, my browser was SeaMonkey 1.1.8, which had the UA string
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8
This means:
- Mozilla/5.0
compatibility with Mozilla products at version 5.0
- Windows
I am using a Windows operating system
- U
strong security for accessing secure Web sites
- Windows NT 5.1
I am using Windows XP (which is derived from Windows NT) version 5.1
- en-US
my browser is localized to use U.S. English (not U.K. English, which is en-GB)
- rv:1.8.1.12
the version of the Gecko rendering engine used by SeaMonkey to format Web pages for display
- Gecko/20080201
the Gecko engine was created on 1 February 2008
- SeaMonkey/1.1.8
the brand-name and version of my browser
Your UA string is
CCBot/1.0 (+http://www.commoncrawl.org/bot.html)
- URL
- Uniform resource locator. The address of a Web page or other Internet entity. The URL for this page is http://www.rossde.com/internet/intr_gloss.html.
- The URL begins with a type or protocol indicator (e.g., http for hypertext transfer protocol, the standardized interface for Web pages).
- Next is the domain: www.rossde.com.
- Finally, there is the location at that domain: /internet/intr_gloss.html. Here, the location involves both a path /internet and a file /intr_gloss.html. In some cases (e.g., my home page at http://www.rossde.com/index.html), the file is at the root of the domain and has no path.
Technically, the term URI (uniform resource identifier) has replaced URL. However, URL remains in common usage. While the definition of URI is more generalized than URL, the difference is mostly in technical details.
