The Jargon File, Version 2.9.10, 01 Jul 1992. Various

Чтение книги онлайн.

СКАЧАТЬ See {network meltdown}.

      :broken: adj. 1. Not working properly (of programs). 2. Behaving strangely; especially (when used of people) exhibiting extreme depression.

      :broken arrow: [IBM] n. The error code displayed on line 25 of a 3270 terminal (or a PC emulating a 3270) for various kinds of protocol violations and "unexpected" error conditions (including connection to a {down} computer). On a PC, simulated with `->/_', with the two center characters overstruck. In true {luser} fashion, the original documentation of these codes (visible on every 3270 terminal, and necessary for debugging network problems) was confined to an IBM customer engineering manual.

      Note: to appreciate this term fully, it helps to know that `broken arrow' is also military jargon for an accident involving nuclear weapons….

      :broket: /broh'k*t/ or /broh'ket`/ [by analogy with `bracket': a `broken bracket'] n. Either of the characters `<' and `>', when used as paired enclosing delimiters. This word originated as a contraction of the phrase `broken bracket', that is, a bracket that is bent in the middle. (At MIT, and apparently in the {Real World} as well, these are usually called {angle brackets}.)

      :Brooks's Law: prov. "Adding manpower to a late software project makes it later" —- a result of the fact that the advantage from splitting work among N programmers is O(N) (that is, proportional to N), but the complexity and communications cost associated with coordinating and then merging their work is O(N^2) (that is, proportional to the square of N). The quote is from Fred Brooks, a manager of IBM's OS/360 project and author of `The Mythical Man-Month' (Addison-Wesley, 1975, ISBN 0-201-00650-2), an excellent early book on software engineering. The myth in question has been most tersely expressed as "Programmer time is fungible" and Brooks established conclusively that it is not. Hackers have never forgotten his advice; too often, {management} does. See also {creationism}, {second-system effect}.

      :BRS: /B-R-S/ n. Syn. {Big Red Switch}. This abbreviation is fairly common on-line.

      :brute force: adj. Describes a primitive programming style, one in which the programmer relies on the computer's processing power instead of using his or her own intelligence to simplify the problem, often ignoring problems of scale and applying na"ive methods suited to small problems directly to large ones.

      The {canonical} example of a brute-force algorithm is associated with the `traveling salesman problem' (TSP), a classical {NP-}hard problem: Suppose a person is in, say, Boston, and wishes to drive to N other cities. In what order should he or she visit them in order to minimize the distance travelled? The brute-force method is to simply generate all possible routes and compare the distances; while guaranteed to work and simple to implement, this algorithm is clearly very stupid in that it considers even obviously absurd routes (like going from Boston to Houston via San Francisco and New York, in that order). For very small N it works well, but it rapidly becomes absurdly inefficient when N increases (for N = 15, there are already 1,307,674,368,000 possible routes to consider, and for N = 1000 —- well, see {bignum}). See also {NP-}.

      A more simple-minded example of brute-force programming is finding the smallest number in a large list by first using an existing program to sort the list in ascending order, and then picking the first number off the front.

      Whether brute-force programming should be considered stupid or not depends on the context; if the problem isn't too big, the extra CPU time spent on a brute-force solution may cost less than the programmer time it would take to develop a more `intelligent' algorithm. Additionally, a more intelligent algorithm may imply more long-term complexity cost and bug-chasing than are justified by the speed improvement.

      Ken Thompson, co-inventor of UNIX, is reported to have uttered the epigram "When in doubt, use brute force". He probably intended this as a {ha ha only serious}, but the original UNIX kernel's preference for simple, robust, and portable algorithms over {brittle} `smart' ones does seem to have been a significant factor in the success of that OS. Like so many other tradeoffs in software design, the choice between brute force and complex, finely-tuned cleverness is often a difficult one that requires both engineering savvy and delicate esthetic judgment.

      :brute force and ignorance: n. A popular design technique at many software houses —- {brute force} coding unrelieved by any knowledge of how problems have been previously solved in elegant ways. Dogmatic adherence to design methodologies tends to encourage it. Characteristic of early {larval stage} programming; unfortunately, many never outgrow it. Often abbreviated BFI: "Gak, they used a bubble sort! That's strictly from BFI." Compare {bogosity}.

      :BSD: /B-S-D/ n. [abbreviation for `Berkeley System Distribution'] a family of {{UNIX}} versions for the DEC {VAX} and PDP-11 developed by Bill Joy and others at {Berzerkeley} starting around 1980, incorporating paged virtual memory, TCP/IP networking enhancements, and many other features. The BSD versions (4.1, 4.2, and 4.3) and the commercial versions derived from them (SunOS, ULTRIX, and Mt. Xinu) held the technical lead in the UNIX world until AT&T's successful standardization efforts after about 1986, and are still widely popular. See {{UNIX}}, {USG UNIX}.

      :BUAF: // [abbreviation, from the alt.fan.warlord] n. Big Ugly ASCII Font —- a special form of {ASCII art}. Various programs exist for rendering text strings into block, bloob, and pseudo-script fonts in cells between four and six character cells on a side; this is smaller than the letters generated by older {banner} (sense 2) programs. These are sometimes used to render one's name in a {sig block}, and are critically referred to as `BUAF's. See {warlording}.

      :BUAG: // [abbreviation, from the alt.fan.warlord] n. Big Ugly ASCII Graphic. Pejorative term for ugly {ASCII ART}, especially as found in {sig block}s. For some reason, mutations of the head of Bart Simpson are particularly common in the least imaginative {sig block}s. See {warlording}.

      :bubble sort: n. Techspeak for a particular sorting technique in which pairs of adjacent values in the list to be sorted are compared and interchanged if they are out of order; thus, list entries `bubble upward' in the list until they bump into one with a lower sort value. Because it is not very good relative to other methods and is the one typically stumbled on by {na"ive} and untutored programmers, hackers consider it the {canonical} example of a na"ive algorithm. The canonical example of a really *bad* algorithm is {bogo-sort}. A bubble sort might be used out of ignorance, but any use of bogo-sort could issue only from brain damage or willful perversity.

      :bucky bits: /buh'kee bits/ n. 1. obs. The bits produced by the CONTROL and META shift keys on a SAIL keyboard (octal 200 and 400 respectively), resulting in a 9-bit keyboard character set. The MIT AI TV (Knight) keyboards extended this with TOP and separate left and right CONTROL and META keys, resulting in a 12-bit character set; later, LISP Machines added such keys as SUPER, HYPER, and GREEK (see {space-cadet keyboard}). 2. By extension, bits associated with `extra' shift keys on any keyboard, e.g., the ALT on an IBM PC or command and option keys on a Macintosh.

      It is rumored that `bucky bits' were named for Buckminster Fuller during a period when he was consulting at Stanford. Actually, `Bucky' was Niklaus Wirth's nickname when *he* was at Stanford; he first suggested the idea of an EDIT key to set the 8th bit of an otherwise 7-bit ASCII character. This was used in a number of editors written at Stanford or in its environs (TV-EDIT and NLS being the best-known). The term spread to MIT and CMU early and is now in general use. See {double bucky}, {quadruple bucky}.

      :buffer overflow: n. What happens when you try to stuff more data into a buffer (holding area) than it can handle. This may be due to a mismatch in the processing rates of the producing and consuming processes (see {overrun} and {firehose syndrome}), or because the buffer is simply too small to hold all the data that must accumulate before a piece of it can be processed. For example, in a text-processing tool that {crunch}es a line at a time, a short line buffer can result in {lossage} as input from a long line overflows the buffer and trashes data beyond it. Good defensive programming would check for overflow on each character and stop accepting data СКАЧАТЬ