Barrons Publishing Dictionary of Computer and Internet Terms 10th

batch processing the noninteractive use of computers. In batch processing, the user gives the computer a “batch” of information, referred to as a job—for example, a program and its input data—and waits for it to be processed as a whole. Batch processing contrasts with interactive processing, in which the user communicates with the computer by means of a terminal while the program is running. The crucial difference is that with batch processing the user must put all of the data into the computer before seeing any of the results, while with interactive processing the user can decide how to handle each item on the basis of the results obtained with earlier items.

baud a unit that measures the speed with which information is transferred. The baud rate is the maximum number of state transitions per second: for instance, a system whose shortest pulses are 1/300 second is operating at 300 baud.

On an RS-232 serial link, the baud rate is equal to the data rate in bits per second (bps). With other kinds of communication, the data rate may be considerably faster than the baud rate. For instance, a 2400-baud modem takes 2400-baud (2400-bps) serial data and encodes it into an audio signal whose true baud rate is 600, near the maximum rate at which a telephone line can transmit pulses. Each pulse carries more than one bit of information. At the receiving end, another modem transforms the signal back into 2400-baud serial data.

Modems connected to telephone lines are still used extensively, but a large amount of Internet traffic is moving to faster BROADBAND services.

bay a place provided for a disk or tape drive in a computer enclosure.

Bayer matrix (or Bayer filter mosaic) the pattern of alternating red, green, and blue filters placed in front of the pixels of a color CMOS or CCD image sensor (Fig. 29), invented by Bryce Bayer (pronounced BY-er) of Kodak.

The idea behind the Bayer matrix is that adjacent points in an image are likely to be the same color, even if they are not the same brightness. Thus, each pixel determines its own brightness, but its color is determined by comparing it to two adjacent pixels with differently colored filters. Green gets twice as many filters as the other two primary colors because it is in the middle of the visible spectrum.

The filters do not block complementary colors completely; they only make each pixel prefer a specific color. The process of decoding colors (DE-BAYERIZATION) is done by the CPU in the digital camera.

Bayes’ theorem the formula

Pr(B|A)Pr(A)

Pr(A|B) =

Pr(B|A)Pr(A) + Pr(B|A′)Pr(A′)

where A and B are two events; Pr(A|B) is the conditional probability that event A happens given that event B happens; Pr(B|A) is the probability

of B given A, and A′ is the event A complement (the event that A does not happen). The vertical line is read as “given that.”

For example, if you have a jar with 2 big red marbles, 4 big blue marbles, 6 small red marbles, and 8 small blue marbles, and you randomly choose one marble, the probability that it will be a big marble given that it is a blue marble will be:

Pr(blue | big)Pr(big)

Pr(big| blue) =

Pr(blue | big)Pr(big) + Pr(blue | small)Pr(small)

In words: “There is a one-third probability that a marble will be big, given that it is blue”; or another way of saying it: “one-third of the blue marbles are big.”

The theorem is named after Rev. Thomas Bayes, English theologian and mathematician (1702–1761).

See BAYESIAN SPAM FILTER.

FIGURE 29. Bayer matrix of red, green, and blue filters

Bayesian spam filter a spam filter that uses BAYES’ RULE to adjust its estimate of the probability that a message is spam. Some words have a high probability of being contained in spam, and other words have a high probability of being contained in legitimate e-mail. These probabilities can be different for different users, so a Bayesian spam filter can adjust

probabilities for different users as users train the system by indicating whether a particular message is spam or not. In general, the Bayesian approach to statistics views probabilities as being conditional on available information, and provides a way to revise them as more information is obtained. However, spammers can try to subvert the filter by including extraneous text in the message that the filter is likely to think is legitimate, or misspelling words that have a high probability of being contained in spam.

bbl chat-room abbreviation for “[I’ll] be back later.”

BBS (bulletin board system) an online message board, especially those that were run on non-networked computers with dial-up modems in the 1980s. See MESSAGE BOARD.

Bcc business abbreviation for “blind copies” (compare CC). In e-mail headers, Bcc: precedes additional addresses to which copies of the message should be sent. Unlike the Cc: header, the Bcc: header is not sent out with the message, so the copies are “blind” (i.e., recipients of the message do not know that copies are being distributed).

Using Bcc:, you can send a message to a large number of people without giving them each other’s e-mail addresses, thus protecting their privacy.

BCD see BINARY-CODED DECIMAL.

BD see BLU-RAY DISC.

bells and whistles elaborate features added to a computer program. The phrase bells and whistles usually connotes that the features are unnecessary and confusing; it originally referred to sound effects devices on theater organs in the silent movie era.

benchmark a computer program used to test the performance of a computer or a piece of software. For example, the speed with which computers do arithmetic is often measured by running a prime-number-finding algorithm called the Sieve of Eratosthenes.

Benchmark results are always somewhat untrustworthy because no single program tests all aspects of a computer’s operation. A particular benchmark may exaggerate a difference between two machines that is unimportant in practice, or it may conceal an important difference. See also MIPS.

Beowulf a type of cluster computing system using machines running LINUX, named for a hero of early English literature. See www.beowulf.org.

Berners-Lee, Tim developer of the hypertext concepts that became the World Wide Web, while he was at the European physics laboratory CERN in the early 1990s.

best practices highly recommended procedures for maintaining computer security or performing other management tasks.

beta testing the second stage of testing a new software product that is almost ready for market. Beta testing is carried out by volunteers in a wide variety of settings like those in which the finished product will be used. Contrast ALPHA TESTING; GAMMA TESTING.

Bézier spline a curve that connects two points smoothly and is further defined by two more points that it does not pass through. Most draw programs represent curves as Bézier splines. For an illustration, see SPLINE. (Bézier is pronounced, roughly, “bay-zee-ay.”)

A Bézier spline can be thought of as a gradual transition from one line to another. Call the four control points that define the curve P1, P2, P3, and P4. Then the curve starts out heading from P1 toward P2. But it curves around so that by the time it gets to P4 it is approaching from the direction of P3.

To plot a Bézier spline, let (x1,y1), (x2,y2), (x3,y3), and (x4,y4) by the coordinates of P1 . . . P4 respectively, and let t range from 0 to 1. Then

compute x and y from t as follows:

a = –t3 + 3t2 – 3t + 1 b = 3t3 – 6t2 + 3t

c = –3t3 + 3t2

3

x = ax1 + bx2 + cx3 + t3x4 y = ay1 + by2 + cy3 + t y4

Notice that as t goes from 0 to 1, a drops from 1 to 0, and t3 rises from 0 to 1. This computation was first described by Pierre Bézier in 1970.

FIGURE 30. Bézier spline

BHO (Browser Help Object) a software component that can be added to Internet Explorer to add new functions. However, BHOs have also been used as SPYWARE.

bid an offer to buy something for a particular price. See AUCTION; EBAY;

RESERVE PRICE.

bidirectional able to transmit data in two directions. See PARALLEL PORT.

big-endian a system of memory addressing in which numbers that occupy more than one byte in memory are stored “big end first,” with the uppermost 8 bits at the lowest address.

For example, the 16-digit binary number 1010111010110110 occupies two 8-bit bytes in memory. On a big-endian computer such as the Macintosh, the upper byte, 10101110, is stored at the first address and the lower byte, 10110110, is stored at the next higher address. On a lit- tle-endian machine, the order is reversed. Contrast LITTLE-ENDIAN.

The terms big-endian and little-endian are from Gulliver’s Travels; they originally referred to the parties in a dispute over which end of a boiled egg should be broken first.

bilingual

1.using more than one language.

2.(describing FIREWIRE cables) having a Firewire 1394b 9-pin connector on one end and a Firewire 1394a 4- or 6-pin connector on the other.

binary addition the process of calculating a sum of numbers expressed in base-2 form. It is one of the basic arithmetic operations performed by computers, and understanding it is the key to understanding how machines do arithmetic.

If each of the numbers to be added has only one digit, addition is simple:

0 + 0 = 00

0 + 1 = 01

1 + 0 = 01

1 + 1 = 10

A circuit that implements this function is called a half adder and can be made out of standard logic gates as shown in Figure 31. Notice that there are two digits of output; the higher digit is called the carry bit because it is carried to the next column when adding multi-digit numbers.

To add numbers with more than one digit, we proceed one digit at a time, starting at the right, just as with pencil-and-paper arithmetic. Each step of the addition can have three inputs: one digit from each of the numbers to be added, plus a digit carried from the previous column to the right. Accordingly, for each column except the rightmost, we need a circuit called a full adder, which takes three one-digit inputs. Its output still has only two digits because the largest value that can be obtained is 1 + 1 + 1 = 11. A full adder can be built out of two half adders (Figure 32).

To add two 16-bit binary numbers, a computer needs 15 full adders and one half adder, with the carry output of each adder connected to an input of the adder to its left. All of this circuitry is part of the CPU; when the CPU receives an add instruction, it sends the contents of two registers to the inputs of the set of adders, and then stores the output in a register.

binary-coded decimal (BCD) a way of representing numbers by means of codes for the decimal digits. For example, consider the number 65. In binary, 65 is 01000001, and that is how most computers represent it. But some computer programs might represent it as the code for 6 followed by the code for 5 (i.e., 0110 0101).

The advantage of BCD shows up when a number has a fractional part, such as 0.1. There is no way to convert 0.1 into binary exactly; it would

1

require an infinite number of digits, just like converting 3 into decimal

(see ROUNDING ERROR). But in BCD, 0.1 is represented by the code for 1 immediately after the point, and no accuracy is lost.

BCD arithmetic is considerably slower and takes more memory than binary arithmetic. It is used primarily in financial work and other situations where rounding errors are intolerable. Pocket calculators use BCD.

FIGURE 31. Binary addition: half adder

FIGURE 32. Binary addition: full adder

binary file a file containing bits or bytes that do not necessarily represent printable text. The term binary file usually denotes any file that is not a text file, such as executable machine language code. Crucially, special software is required to print a binary file or view it on the screen.

Contrast TEXT FILE.

binary multiplication a basic operation in computer arithmetic. For singledigit numbers, the binary multiplication table is very simple and is the same as the Boolean AND operation (see AND GATE):

0 × 0 = 0

0 × 1 = 0

1 × 0 = 0

1 × 1 = 1

For numbers with more than one digit, the computer does something very similar to what we do to decimal numbers in pencil-and-paper arithmetic. To find 13 × 21 in decimal, we proceed like this:

21 × 13

63

21

273

First, find 3 × 21. Then, find 10 × 21, and add these two results together to get the final product. Note that the product has more digits than either of the two original numbers.

You can follow the same procedure to multiply two binary numbers:

10101 × 01101

10101

00000

10101

10101

00000

100010001

Notice that each of the partial products is either zero or a copy of 10101 shifted leftward some number of digits. The partial products that are zero can, of course, be skipped. Accordingly, in order to multiply in binary, the computer simply starts with 0 in the accumulator and works through the second number to be multiplied (01101 in the example), checking whether each digit of it is 1 or 0. Where it finds a 0, it does nothing; where it finds a 1, it adds to the accumulator a copy of the first number, shifted leftward the appropriate number of places.

binary number a number expressed in binary (base-2) notation, a system that uses only two digits, 0 and 1. Binary numbers are well suited for use by computers, since many electrical devices have two distinct states: on and off. Writing numbers in binary requires more digits than writing numbers in decimal, so binary numbers are cumbersome for people to use. Each digit of a binary number represents a power of 2. The rightmost digit is the 1’s digit, the next digit leftward is the 2’s digit, then the 4’s digit, and so on:

Table 4 shows examples of numbers written in binary and decimal form.

See also DECIMAL NUMBER; HEXADECIMAL NUMBER; OCTAL.

binary search a method for locating a particular item from a list of items in alphabetical or numerical order. Suppose you need to find the location of a particular word in a list of alphabetized words. To execute a binary search, look first at the word that is at the exact middle of the list. If the word you’re looking for comes before the midpoint word, you know that it must be in the first half of the list (if it is in the list at all). Otherwise, it must be in the second half. Once you have determined which half of the list to search, use the same method to determine which quarter, then which eighth, and so on. At most, a binary search will take about N steps if the list contains about 2N items.

binary subtraction a basic operation in computer arithmetic. The easiest way to subtract two binary numbers is to make one of the numbers negative and then add them. Circuits for doing binary addition are readily constructed with logic gates (see BINARY ADDITION). The negative counterpart of a binary number is called its 2-complement.

Suppose that we have a number x, represented as a binary number with k digits. The 2-complement of x (written as x) is

x = 2k – x

Then, to find the difference a – x we can compute

a – x = a + x – 2k

This is easier than it looks, for two reasons. First, subtracting 2k is trivial, because 2k is a binary number of the form 1000, 100000, and so on, with k +1 digits. So all we have to do is discard the leftmost digit to get our k-digit answer.

Second, finding the 2-complement of x is easy: just invert all the digits of x (changing 0’s to 1’s and 1’s to 0’s) and then add 1. See INVERTER.

bind to associate symbols with data, or to associate one piece of data with another, in several different ways, among them:

1.to give a variable a value; to INITIALIZE it.

2.to associate a network protocol with a particular Ethernet port or the like. See PROTOCOL.

3.to map an XML document onto a set of variables or objects in Java or another programming language.

4.to put together the pages of a book.

binding see BIND (all definitions).

biometrics measurable physical characteristics of the human body, used to identify an individual for security purposes. They include fingerprints, the distinctive appearance of faces and eyes, and the distinctive sound quality of one’s voice. There are computer input devices to read these characteristics.

BIOS (Basic Input Output System) a set of procedures stored on a ROM chip inside PC-compatible computers. These routines handle all inputoutput functions, including screen graphics, so that programs do not have to manipulate the hardware directly. This is important because if the hardware is changed (e.g., by installing a newer kind of video adapter), the BIOS can be changed to match it, and there is no need to change the application programs.

The BIOS is not re-entrant and is therefore not easily usable by multitasking programs. Windows programs do not call the BIOS; instead, they use procedures provided by the operating system.

BIOS enumerator the BIOS routine that tells a PLUG AND PLAY system what hardware is installed.

bipolar transistor a semiconductor device formed by sandwiching a thin layer of P- or N-type semiconductor between two layers of the opposite type of semiconductor. (See TRANSISTOR.) The other general type of transistor is the field-effect transistor (FET).

bis Latin for “a second time,” used to denote revised CCITT and ITU-T standards. See CCITT; ITU-T.

BIST (built-in self test) a feature included in newer integrated circuits and other electronic equipment. An electronic device that has BIST can test itself thoroughly whenever it is turned on. See INTEGRATED CIRCUIT.

bit a shorthand term for binary digit. There are only two possible binary digits: 0 and 1. (See BINARY NUMBER.) Bits are represented in computers by two-state devices, such as flip-flops. A computer memory is a collection of devices that can store bits.

A byte is the number of bits (usually 8) that stand for one character. Memory is usually measured in units of kilobytes or megabytes. See

MEMORY.