|
TiVo
TLC
TLD
TMC
token ring
toner/toner cartridge
toolbar
top-level domain (TLD)
topology
TopView
TOS
TiVo Because the TiVo recorder records shows based on
your personal preferences, you always have something interesting to watch when you turn on your television. | TiVo
Originally called Teleworld, TiVo was founded in August 1997 by James Barton and Michael Ramsay, two Silicon Graphics engineers. TiVo designs the PVR (personal video recorder; also called a DVR [digital video recorder]), which sits between the user and the TV networks and digitally records shows, much like a VCR. The most amazing aspect of TiVo is that the company did not plan to make money on the hardware. In fact, Tivo doesn't even make the hardware. Hardware is farmed out to Sony, Philips, and Hughes. The first TiVo units were shipped in March 1999. The current cost of a TiVo hardware unit is about $400.
TiVo runs on the Linux operating system. All of the software combined takes up about 24MB with the TiVo software using most of that space. There have been nine major upgrades to the software, and each new upgrade is automatically downloaded to every TiVo when it connects nightly to get an updated schedule. That way, the software is always current. The software is so reliable that TiVo, which is really a PowerPC-based computer, has no on/off switch or reset button. Just as amazing is that the upgraded software runs just fine on the oldest of the hardware units.
TiVo records digitally with MPEG II (Moving Pictures Experts Group, Layer 2) compression using one of four user selected levels of video quality. Currently, maximum recording time is a whopping 60 hours. In order to operate, TiVo needs to know the programming schedule. It gets this information by calling into the TiVo network each night. Currently, schedules are available for the United States and United Kingdom. This service costs $10 per month or $250 for a lifetime subscription.
These calls have raised privacy concerns. The One Privacy Foundation project analyzed the data sent back to TiVo and found that it included a unique identifying number and information on the shows viewed. This would let TiVo build a database of the viewing habits of individual families. TiVo claims that it does not correlate viewing information with customer ID numbers and that it only uses aggregate data.
Currently, TiVo is available in about 4,000 outlets nationally. For the quarter ending September 2001, TiVo had loses of $33.8 million on revenues of only $5.3 million. As of Oct. 31, 2001, TiVo had 280,000 subscribers. Analysts estimate TiVo needs at least 800,000 subscribers to break even. TiVo has approximately 100 employees.
Although the TiVo business model is based mostly on subscription revenues, the company has other sources of revenue. TiVo is able to show banner ads on system menus. TiVo also has the ability to do market research using the TiVo hardware.
Major investors in TiVo include Phillips Electronics and Sony. An additional TiVo investor is AOL Time Warner, which owns a 15% share of the company and plans to offer AOL-TV through TiVo. Finally, DIRECTV owns a 10% share of TiVo. Back to top
TLC
See translation look-aside buffer. Back to top
TLD
See top-level domain. Back to top
TMC
See Thinking Machines Corp. Back to top
token ring In a token ring network, the stations are connected to each other using either a ring format, shown on the left, or a star format, shown above. In a star format, relays (represented by the blue boxes) are used to switch out a nonresponding computer while maintaining the ring. | token ring
The IEEE 802.5 token-ring standard is based on the IBM commercial Token-Ring network developed in the 1970s. Token rings gained broad acceptance due to their backing by IBM, but token-ring networks have never achieved the popularity of Ethernet systems.
In a token-ring network, the stations are connected to one another either in a ring format, where each computer is connected to the next computer, or in a star format, where each computer connects to a central device, using an MSAU (multistation access unit). The second approach is far more common and offers more flexibility.
Regardless of how the stations are connected physically, logically they are connected to each other serially so that station 1 connects to station 2, station 2 connects to station 3, and so on until the last station connects back to the first station. When connected via an MSAU, stations that stop responding can be switched out of the network loop and new stations can be inserted into the network loop as required. Shielded or unshielded twisted-pair cable is used to make the connections. A token-ring network can operate at either 4Mbps (megabits per second) or 16Mbps.
The token-ring technology is based on a small frame, called a token. When no station needs to communicate across the network, a single token is continually passed from station to station around the network. These network-available tokens are tiny 3-byte frames that are passed around the network to indicate that the network is available for sending data. The first byte of the network-available token is a start delimiter, the second byte is an access control byte, and the final byte is an end delimiter. To send data over the network, a station grabs the network-available token and changes the second byte. That changes the token into a start-of-frame sequence. It then appends the data to be sent and passes this information to the next station in the ring. For example, when a station, say station 1, needs to communicate over the network, it must wait until the token reaches it.
When station 1 obtains the token, it changes 1 bit in the token. This changes the token to a start-of-frame sequence for a data frame. The station then adds the remainder of the fields needed to construct a data frame, including the address of the station to receive the data, say station 4. It then transmits the packet to the next station in the ring.
Each station, in turn, examines the packet of data to see if the data is addressed to it. If not, that station passes the packet on to the next station in the ring. Once the packet reaches its destination, the receiving station reads the packet. However, station 4 cannot free up the network. Rather, it converts the data packet into an acknowledgement packet. The destination station (station 4) then changes the destination address to the original sending station (station 1) and passes the packet on to the next station in the ring.
As before, each station, in turn, examines the acknowledgement packet to see if that packet is addressed to it. If not, that station passes the acknowledgement packet to the next station in the ring. Once the acknowledgement packet reaches station 1, the loop is complete. Station 1 changes the acknowledgement packet to a token and sends that token on to the next station, thus freeing up the network for use by another station. Because only one packet or token is in circulation at any given time, it is impossible for data packets to collide with one another on the network.
When lightly loaded, a token-ring network is somewhat inefficient because a station must wait on a token to come around before transmitting data. However, when the network is highly loaded, which is when efficiencies really matter, a token-ring network's natural round-robin approach allocates bandwidth in an efficient and fair manner. This is the principal advantage of a token-ring network. This efficiency at high loads makes a token-ring network the ideal approach where delays must be predictable and robust network operation is critical, such as real-time factory automation.
The principal disadvantage of a token-ring network is the need for token maintenance. Because the loss of a token prevents further utilization of the network, the network must make sure that the token is not lost. Having two tokens active at once can also disrupt network operations. For these reasons, one station in the ring must act as a monitor to make sure that exactly one token is active on the network at any given time. This is called the active monitor. Each time the network starts up, the stations on the network negotiate to decide which one will become the monitor. The computer with the highest MAC (media control access) address, a unique hardware address associated with each computer, wins this negotiation.
Another problem faced by a token-ring network is the loss of one of the stations connected to the network. Because data is handed off station-to-station serially, the loss of a single station can stop network operations. This is avoided by a process called beaconing. Every few seconds, the monitoring station sends out a command telling every station on the network to announce itself. If a downstream station fails to receive an announcement from an upstream station, the downstream station informs the network that the upstream station is no longer active. The network then automatically reconfigures itself to skip the inactive station. Back to top
toner/toner cartridge
Toner is the ink that lets a laser printer print. However, toner is not anything like typical ink. First, it is not a liquid but a fine powder. Second, it contains plastic and rust. To understand toner, you must first understand how a laser printer prints.
Printing in a laser printer starts with applying an electrical charge to a rotating drum. A laser beam is then used to remove the charge at those locations that are to receive the toner. After the laser writes the image, the rotating drum is coated with toner. This method is called black write. It is the most common approach. There is also a white write method where the laser marks off the white nonprinting areas rather than the black printing areas.
If the toner were liquid, it would wet the entire drum, but because it is a powder, it does not naturally bind to the drum. To cause it to bind, the toner is given the same charge as the rotating drum. That way, the charged part of the drum repels the toner so the toner only binds to the part of the drum where the laser has removed the charge.
With the toner magnetically attached to the rotating drum in a pattern that matches the image to be printed, the drum rotates over a sheet of paper. The paper has been given a strong magnetic charge that is opposite of the charge of the toner. This lets the paper pull the toner off the rotating drum. Because the paper is moving at the same speed as the drum, this transfers the image to the paper.
At this point, the paper simply has toner clinging to it via a magnetic charge. You could easily brush the toner off. To keep the toner on the paper permanently, the paper passes through the fuser. The fuser is a pair of heated rollers that heat the paper and melt the toner. The fuser rollers are coated with Teflon to keep the toner from sticking to them rather than the paper.
The toner itself is a fine, electrically charged powder with the consistency of talcum powder. Toner particles are generally 5 to 15 microns (millionth of a meter) wide. The main ingredients are pigment and plastic. In addition to pigment and plastic, toner has a high iron oxide (rust) content, typically around 40%.
The pigment colors the mixture and the plastic melts in the fuser to bind the pigment to the paper. Binding the pigment using melted plastic means that a laser printer can print onto almost any type of paper. It also means that the printing will tightly bind to the paper and is not easily smudged if it gets wet or is handled. The iron oxide helps the toner hold its magnetic charge.
In addition to pigment, plastic, and iron oxide, several other ingredients are used. The two most common are sand and wax. Finely ground sand or silica is added to prevent clumping. Wax is added to help distribute the toner as it melts.
To make toner, these ingredients are combined according to the recipe for a particular printer. Recipes differ mainly because different fusers heat to different temperatures, and the toner must match that temperature. The toner ingredients are heated to a consistency of hot taffy. This mixture is then cooled into sheets. These sheets are then ground into pellets, and the pellets are milled into the fine powder found in printers. The finer the powder, the sharper the images, and the higher the printing resolution, the finer the toner must be.
The laser printer stores the toner in an area called the toner hopper. The toner is transferred from the toner hopper to the rotating drum using a roller called a developer roller. To reduce maintenance, most laser printers combine the toner hopper, developer roller, and drum assembly into one replaceable unit called a toner cartridge. Typically, you can print several thousand pages using a single toner cartridge. Each time you replace the toner cartridge, not only is the toner replenished, most of the parts that are likely to fail are also replaced, giving your laser printer a long life span. Back to top
toolbar The Eudora toolbar gives you quick access to some
e-mail functions, such as sending your e-mail and
checking to see if you have mail. | toolbar
Complex programs, such as Microsoft Word or Excel, have literally hundreds of menu options. Programmers try to arrange menus so they make sense, but the programmers must also make sure that none of the menus get too long. In order to do this they often create menus inside of menus, called nested menus.
The result is a complex menu system that has some actions buried three or more levels deep. When an action is used infrequently or has the potential to be dangerous, it may not be a bad thing to make it hard to find. However, when an action needs to be performed frequently, such as saving a data file or printing, the harder it is to reach, the more it slows down the user. Not only that, hiding actions layers deep in the menu system makes the program harder to learn.
Toolbars solve the problem of nested menus, making it quick and easy for users to reach common menu options. Each menu option on the toolbar is represented by an icon. These icons are arranged in a row, generally near the top of the program just below the menus. Clicking an icon will generally perform the appropriate action without the user doing anything else. For example, clicking the Print icon will usually print the current data file without the user filling in a dialog box.
To make the icons easier to understand, most toolbars will show a brief text description of an icon if you pause the mouse pointer over an icon without clicking. These explanations usually appear below the mouse cursor. Some programs let you customize their toolbars.
Toolbars are so popular with users that many programs have several different toolbars. For example, Microsoft Word has up to 16 different toolbars you can display, depending on the version, and the Standard and Formatting toolbars are turned on by default. Back to top
top-level domain (TLD)
The Internet was originally designed on the premise that any computer connected to the Internet could be identified by its IP (Internet Protocol) address. (The IP address for Smartcomputing.com is 12.39.144.5, and this will work as a URL [uniform resource locator] in most browsers.) Although computers have no problem dealing with IP addresses, humans have a great deal of difficulty remembering them. Additionally, if a Web site changes host or computer, its IP address also changes.
These problems resulted in the development of the friendlier DNS (Domain Name System). With DNS, Web sites have easy-to-remember names, such as Smartcomputing.com. When the user enters the name of a Web site, DNS translates that name into an IP address for the computer to use.
Using DNS to translate site names into IP addresses requires that the names be constructed in a systematic method. The first part of the name is the portion that comes after the last dot. This portion is called the top-level domain. In the URL http://www.smartcomputing.com, for example, the TLD is .com.
In the United States, the TLD normally indicates the type of organization that owns the Web site—for example, .com for commercial businesses; .edu for U.S. educational institutions; .gov for government agencies or departments; .org for nonprofit organizations; .mil for the U.S. military branches, agencies, or departments; and .net for network providers, especially ISPs (Internet service providers). Recently approved TLDs include .biz for businesses, .info for informational sites, .name for personal sites, .pro for professional individuals (such as doctors and lawyers), .museum for authentic museums, .coop for business cooperatives, and .aero for the aviation industry. In other countries, the TLD is a two-digit code that represents the country of origin, .af for Afghanistan or .tv for Tuvalu, for example.
The TLD .com has become very crowded. Virtually all of the easy-to-remember names, words, and even phrases have already been taken. To overcome this, the nonprofit organization that oversees the infrastructure for Internet addresses, ICANN (Internet Corporation for Assigned Names and Numbers), plans on slowly introducing new TLDs. The slow introduction is planned to reduce cybersquatting, the registering of a site name solely for the purpose of selling it to someone else. Back to top
topology
In a network, the topology of the network refers to the geometrical arrangement and connections of network nodes. The physical topology is the actual layout of the cabling or other transmission media used to connect the computers and other peripherals (nodes) on the network. The logical topology is the scheme used on the network to let devices send and receive information over the transmission media without interference from one another. The physical and logical topologies do not have to be the same.
There are three main topologies: linear bus, ring, and star. However, networks often employ combinations of these rather than using just one of them exclusively.
In a linear bus topology, a single cable supports the entire network. This cable is sometimes called a backbone. Nodes are usually connected directly to the bus by a T-connector or drop connector. There is a terminating resistor at each end of the cable. The signal travels up and down the cable. The nodes on the bus passively listen for a signal addressed to them. The failure of a node will not cause the network to fail since the nodes are not actively involved in passing the signals. The linear bus topology is sometimes called a daisy chain topology.
In a ring topology, the nodes are connected together to form a circle. Each node has a connection to its upstream neighbor to receive data and its downstream neighbor to pass along data. The data only travels in one direction along the network. A ring topology usually uses token passing. The ring topology is almost always implemented only as the logical topology of the network. The physical topology is usually a star topology with a central hub simulating the physical ring topology.
In a star topology, each node has its own cable running between it and a central device called a hub. The signal runs bi-directionally along this cable since the node uses it to both send and receive data. The hub connects the nodes together internally.
There are two additional, and less common, topologies: multi-connected and mesh. In a multi-connected topology, the nodes are connected to one another using direct links (called point-to-point links) in an arbitrary fashion with each node connected to a minimum of two additional nodes and possibly many more. If each node is only connected to two other nodes, you have a ring topology. The additional connections improve reliability and reduce congestion but make routing much more complex since there are multiple paths to each node.
In a mesh topology, each node is connected to every other node. This maximizes the number of connections giving the greatest fault tolerance from all the redundant links. However, mesh topology also maximizes routing complexity because of the redundant links. Due to the expense of all the network cards and cabling required, mesh topology is rarely used. Mesh topology is sometimes called fully connected topology.
In a parallel processing architecture, topology refers to the interconnection between processors rather than the interconnection between nodes on a network. Back to top
TopView
In 1985, the typical PC had an 8088 or 80286 processor with between 256KB and 640KB of RAM. The OS (operating system) was DOS, and Microsoft was not yet shipping any version of Windows.
DOS could only run one program at a time. If you were writing in your word processor and needed to check some numbers in your spreadsheet, you had to save your document, close your word processor, open your spreadsheet program, load your worksheet, check your numbers, close your spreadsheet, open your word processor, and reload your document.
To overcome the problem of only being able to run one program at a time, in February 1985, IBM introduced TopView. TopView was a character-based multitasking environment. At this time, most computers did not have mice. The few with a mouse had one that had three buttons rather than the two-button mice we use today. TopView used this now-defunct third mouse button for its operations.
TopView had a number of problems that doomed it to failure. First, its memory requirements were out of line with the computers of that day. Second, it was slow, and in 1985, computers were too slow to accommodate it. (TopView was so slow that its nickname was TopHeavy.) Third, it was not 100% compatible with DOS, and many programs needed to be adapted to run under TopView.
TopView was a major failure in the marketplace and is often found on lists of the top 10 blunders of the PC industry. However, it would have a lasting impact on IBM. Microsoft contacted IBM several times regarding a program they were developing called Windows. IBM was not interested because, at the time, they were working on TopView. After the release of Windows, IBM would scrap TopView in June 1987 when IBM and Microsoft began cooperating on the development of OS/2. Back to top
TOS
See terms of service. Back to top
|
Email This
Print This
