Search billions of records on
Website logo - Click to go to Home page

Jargon used in computing

This document (the Jargon File) is in the public domain, to be freely used, shared, and
modified.  There are (by intention) no legal restraints on what you can do with it, but there
are traditions about its proper use to which many hackers are quite strongly attached.
Please extend the courtesy of proper citation when you quote the File, ideally with a version
number, as it will change and grow over time.


About This Book

This document is a collection of slang terms used by various subcultures of computer hackers.
Though some technical material is included for background and flavor, it is not a technical
dictionary; what we describe here is the language hackers use among themselves for fun,
social communication, and technical debate.

The `hacker culture' is actually a loosely networked collection of subcultures that is
nevertheless conscious of some important shared experiences, shared roots, and shared values.
It has its own myths, heroes, villains, folk epics, in-jokes, taboos, and dreams.  Because
hackers as a group are particularly creative people who define themselves partly by
rejection of `normal' values and working habits, it has unusually rich and conscious
traditions for an intentional culture less than 35 years old.

As usual with slang, the special vocabulary of hackers helps hold their culture together --
- it helps hackers recognize each other's places in the community and expresses shared
values and experiences.  Also as usual, *not* knowing the slang (or using it
inappropriately) defines one as an outsider, a mundane, or (worst of all in hackish
vocabulary) possibly even a {suit}.  All human cultures use slang in this threefold way --
- as a tool of communication, and of inclusion, and of exclusion.

Among hackers, though, slang has a subtler aspect, paralleled perhaps in the slang of jazz
musicians and some kinds of fine artists but hard to detect in most technical or scientific
cultures; parts of it are code for shared states of *consciousness*.  There is a whole
range of altered states and problem-solving mental stances basic to high-level hacking
which don't fit into conventional linguistic reality any better than a Coltrane solo or one
of Maurits Escher's `trompe l'oeil' compositions (Escher is a favourite of hackers), and
hacker slang encodes these subtleties in many unobvious ways.  As a simple example, take
the distinction between a {kluge} and an {elegant} solution, and the differing connotations
attached to each.  The distinction is not only of engineering significance; it reaches right
back into the nature of the generative processes in program design and asserts something
important about two different kinds of relationship between the hacker and the hack.
Hacker slang is unusually rich in implications of this kind, of overtones and undertones
that illuminate the hackish psyche.

But there is more.  Hackers, as a rule, love wordplay and are very conscious and inventive
in their use of language.  These traits seem to be common in young children, but the
conformity-enforcing machine we are pleased to call an educational system bludgeons them
out of most of us before adolescence.  Thus, linguistic invention in most subcultures of
the modern West is a halting and largely unconscious process.  Hackers, by contrast,
regard slang formation and use as a game to be played for conscious pleasure.
Their inventions thus display an almost unique combination of the neotenous enjoyment of
language-play with the discrimination of educated and powerful intelligence.  Further,
the electronic media which knit them together are fluid, `hot' connections, well adapted
to both the dissemination of new slang and the ruthless culling of weak and superannuated
specimens.  The results of this process give us perhaps a uniquely intense and accelerated
view of linguistic evolution in action.

The intensity and consciousness of hackish invention make a compilation of hacker slang a
particularly effective window into the surrounding culture --- and, in fact, this one is
the latest version of an evolving compilation called the `Jargon File', maintained by
hackers themselves for over 15 years.  This one (like its ancestors) is primarily a
lexicon, but also includes `topic entries' which collect background or sidelight
information on hacker culture that would be awkward to try to subsume under individual

Though the format is that of a reference volume, it is intended that the material be
enjoyable to browse.  Even a complete outsider should find at least a chuckle on nearly
every page, and much that is amusingly thought-provoking.  But it is also true that
hackers use humorous wordplay to make strong, sometimes combative statements about what
they feel.  Some of these entries reflect the views of opposing sides in disputes that
have been genuinely passionate; this is deliberate.  We have not tried to moderate or
pretty up these disputes; rather we have attempted to ensure that *everyone's* sacred cows
get gored, impartially.  Compromise is not particularly a hackish virtue, but the honest
presentation of divergent viewpoints is.

The reader with minimal computer background who finds some references incomprehensibly
technical can safely ignore them.  We have not felt it either necessary or desirable to
eliminate all such; they, too, contribute flavor, and one of this document's major
intended audiences --- fledgling hackers already partway inside the culture --- will
benefit from them.

A selection of longer items of hacker folklore and humour is included in appendix A.
The `outside' reader's attention is particularly directed to appendix B, "A Portrait of
J. Random Hacker".  Appendix C is a bibliography of non-technical works which have either
influenced or described the hacker culture.

Because hackerdom is an intentional culture (one each individual must choose by action
to join), one should not be surprised that the line between description and influence
can become more than a little blurred.
Earlier versions of the Jargon File have played a central role in spreading hacker
language and the culture that goes with it to successively larger populations, and we hope
and expect that this one will do likewise.

Of Slang, Jargon, and Techspeak

Linguists usually refer to informal language as `slang' and reserve the term `jargon' for
the technical vocabularies of various occupations. However, the ancestor of this
collection was called the `Jargon File', and hackish slang is traditionally `the jargon'.
When talking about the jargon there is therefore no convenient way to distinguish what a
*linguist* would call hackers' jargon --- the formal vocabulary they learn from textbooks,
technical papers, and manuals.

To make a confused situation worse, the line between hackish slang and the vocabulary of
technical programming and computer science is fuzzy, and shifts over time.  Further, this
vocabulary is shared with a wider technical culture of programmers, many of whom are not
hackers and do not speak or recognize hackish slang.

Accordingly, this lexicon will try to be as precise as the facts of usage permit about
the distinctions among three categories:

   * `slang': informal language from mainstream English or non-technical subcultures
     (bikers, rock fans, surfers, etc.)
   * `jargon': without qualifier, denotes informal `slangy' language peculiar to hackers
      --- the subject of this lexicon
   * `techspeak': the formal technical vocabulary of programming, computer science,
      electronics, and other fields connected to hacking

This terminology will be consistently used throughout the remainder of this lexicon.

The jargon/techspeak distinction is the delicate one.  A lot of techspeak originated as
jargon, and there is a steady continuing uptake of jargon into techspeak.  On the other
hand, a lot of jargon arises from overgeneralization of techspeak terms (there is more
about this in the "Jargon Construction" section below).

In general, we have considered techspeak any term that communicates primarily by a
denotation well established in textbooks, technical dictionaries, or standards documents.

A few obviously techspeak terms (names of operating systems, languages, or documents) are
listed when they are tied to hacker folklore that isn't covered in formal sources, or
sometimes to convey critical historical background necessary to understand other entries
to which they are cross-referenced.  Some other techspeak senses of jargon words are listed
in order to make the jargon senses clear; where the text does not specify that a straight
technical sense is under discussion, these are marked with `[techspeak]' as an etymology.
Some entries have a primary sense marked this way, with subsequent jargon meanings
explained in terms of it.

We have also tried to indicate (where known) the apparent origins of terms.  The results
are probably the least reliable information in the lexicon, for several reasons.  For one
thing, it is well known that many hackish usages have been independently reinvented
multiple times, even among the more obscure and intricate neologisms.  It often seems
that the generative processes underlying hackish jargon formation have an internal logic
so powerful as to create substantial parallelism across separate cultures and even in
different languages!  For another, the networks tend to propagate innovations so quickly
that `first use' is often impossible to pin down.  And, finally, compendia like this one
alter what they observe by implicitly stamping cultural approval on terms and widening
their use.

Revision History

The original Jargon File was a collection of hacker jargon from technical cultures
including the MIT AI Lab, the Stanford AI lab (SAIL), and others of the old ARPANET
AI/LISP/PDP-10 communities including Bolt, Beranek and Newman (BBN), Carnegie-Mellon
University (CMU), and Worcester Polytechnic Institute (WPI).

The Jargon File (hereafter referred to as `jargon-1' or `the File') was begun by Raphael
Finkel at Stanford in 1975.  From this time until the plug was finally pulled on the SAIL
computer in 1991, the File was named AIWORD.RF[UP,DOC] there.  Some terms in it date back
considerably earlier ({frob} and some senses of {moby}, for instance, go back to the Tech
Model Railroad Club at MIT and are believed to date at least back to the early 1960s).
The revisions of jargon-1 were all unnumbered and may be collectively considered `Version 1'.

In 1976, Mark Crispin, having seen an announcement about the File on the SAIL computer,
{FTP}ed a copy of the File to MIT.  He noticed that it was hardly restricted to `AI words'
and so stored the file on his directory as AI:MRC;SAIL JARGON.

The file was quickly renamed JARGON> (the `>' means numbered with a version number) as a
flurry of enhancements were made by Mark Crispin and Guy L. Steele Jr.  Unfortunately,
amidst all this activity, nobody thought of correcting the term `jargon' to `slang' until
the compendium had already become widely known as the Jargon File.

Raphael Finkel dropped out of active participation shortly thereafter and Don Woods became
the SAIL contact for the File (which was subsequently kept in duplicate at SAIL and MIT,
with periodic resynchronizations).

The File expanded by fits and starts until about 1983; Richard Stallman was prominent
among the contributors, adding many MIT and ITS-related coinages.

A late version of jargon-1, expanded with commentary for the mass market, was edited by Guy
Steele into a book published in 1983 as `The Hacker's Dictionary' (Harper & Row CN 1082,
ISBN 0-06-091082-8).  The other jargon-1 editors (Raphael Finkel, Don Woods, and Mark
Crispin) contributed to the revision, as did Richard M. Stallman and Geoff Goodfellow.
This book (now out of print) is hereafter referred to as `Steele-1983' and those six as
the Steele-1983 coauthors.

Shortly after the publication of Steele-1983, the File effectively stopped growing and
changing.  Originally, this was due to a desire to freeze the file temporarily to
facilitate the production of Steele-1983, but external conditions caused the
`temporary' freeze to become permanent.

The AI Lab culture had been hit hard in the late 1970s by funding cuts and the resulting
administrative decision to use vendor-supported hardware and software instead of homebrew
whenever possible.  At MIT, most AI work had turned to dedicated LISP Machines.  At the
same time, the commercialization of AI technology lured some of the AI Lab's best and
brightest away to startups along the Route 128 strip in Massachusetts and out West in
Silicon Valley.  The startups built LISP machines for MIT; the central MIT-AI computer
became a {TWENEX} system rather than a host for the AI hackers' beloved {ITS}.

The Stanford AI Lab had effectively ceased to exist by 1980, although the SAIL computer
continued as a Computer Science Department resource until 1991.  Stanford became a major
{TWENEX} site, at one point operating more than a dozen TOPS-20 systems; but by the
mid-1980s most of the interesting software work was being done on the emerging BSD UNIX

In April 1983, the PDP-10-centered cultures that had nourished the File were dealt a
death-blow by the cancellation of the Jupiter project at Digital Equipment Corporation.
The File's compilers, already dispersed, moved on to other things.  Steele-1983 was partly
a monument to what its authors thought was a dying tradition; no one involved realized at
the time just how wide its influence was to be.

By the mid-1980s the File's content was dated, but the legend that had grown up around it
never quite died out.  The book, and softcopies obtained off the ARPANET, circulated even
in cultures far removed from MIT and Stanford; the content exerted a strong and continuing
influence on hackish language and humor.  Even as the advent of the microcomputer and
other trends fueled a tremendous expansion of hackerdom, the File (and related materials
such as the AI Koans in Appendix A) came to be seen as a sort of sacred epic, a
hacker-culture Matter of Britain chronicling the heroic exploits of the Knights of the Lab.
The pace of change in hackerdom at large accelerated tremendously --- but the Jargon File,
having passed from living document to icon, remained essentially untouched for seven years.

This revision contains nearly the entire text of a late version of jargon-1 (a few obsolete
PDP-10-related entries were dropped after careful consultation with the editors of
Steele-1983).  It merges in about 80% of the Steele-1983 text, omitting some framing
material and a very few entries introduced in Steele-1983 that are now also obsolete.

This new version casts a wider net than the old Jargon File; its aim is to cover not just
AI or PDP-10 hacker culture but all the technical computing cultures wherein the true
hacker-nature is manifested.  More than half of the entries now derive from {USENET} and
represent jargon now current in the C and UNIX communities, but special efforts have been
made to collect jargon from other cultures including IBM PC programmers, Amiga fans, Mac
enthusiasts, and even the IBM mainframe world.

Eric S. Raymond maintains the new File with assistance from Guy L. Steele Jr.; these are
the persons primarily reflected in the File's editorial `we', though we take pleasure in
acknowledging the special contribution of the other coauthors of Steele-1983.

Some snapshot of this on-line version will become the main text of a `New Hacker's
Dictionary', to be published by MIT Press possibly as early as Summer 1991.
The maintainers are committed to updating the on-line version of the Jargon File through
and beyond paper publication, and will continue to make it available to archives and
public-access sites as a trust of the hacker community.

Here is a chronology of the recent on-line revisions:

Version 2.1.1, Jun 12 1990: the Jargon File comes alive again after a seven-year hiatus.
Reorganization and massive additions were by Eric S. Raymond, approved by Guy Steele.
Many items of UNIX, C, USENET, and microcomputer-based jargon were added at that time
(as well as The Untimely Demise of Mabel The Monkey).  Some obsolete usages (mostly PDP-10
derived) were moved to Appendix B.

Version 2.1.5, Nov 28 1990: changes and additions by ESR in response to numerous USENET
submissions and comment from the First Edition co-authors.  The bibliography (Appendix C)
was also appended.  This version had 6028 lines, 46946 words, 307510 characters, and 866

Version 2.2.1, Dec 15 1990: most of the contents of the 1983 paper edition edited by Guy
Steele was merged in.  Many more USENET submissions added, including the International
Style and the material on Commonwealth Hackish.  This version had 9394 lines, 75954 words,
490501 characters, and 1046 entries.

Version 2.3.1, Jan 03 1991: the great format change --- case is no longer smashed in
lexicon keys and cross-references.  A very few entries from jargon-1 which were basically
straight techspeak were deleted; this enabled the rest of Appendix B to be merged back
into main text and the appendix replaced with the Portrait of J. Random Hacker.  More
USENET submissions were added.  This version had 10728 lines, 85070 words, 558261
characters, and 1138 entries.

Version 2.4.1, Jan 14 1991: the Story of Mel and many more USENET submissions merged in.
More material on hackish writing habits added. Numerous typo fixes.  This version had
12362 lines, 97819 words, 642899 characters, and 1239 entries.

Version 2.5.1, Jan 29 1991: many new entries merged in.  Discussion of inclusion styles
added.  This version had 14145 lines, 111904 words, 734285 characters, and 1425 entries.

Version 2.6.1, Feb 13 1991: second great format change; no more <> around headwords or
references.  Merged in results of serious copy-editing passes by Guy Steele, Mark Brader.
Still more entries added.  This version had 15011 lines, 118277 words, 774942 characters,
and 1485 entries.

Version 2.7.1, Mar 01 1991: new section on slang/jargon/techspeak added. Results of Guy's
second edit pass merged in.  This version had 16087 lines, 126885 words, 831872 characters,
and 1533 entries.

Version 2.8.1, Mar 22 1991: material from the TMRC Dictionary and MRC's editing pass
merged in.  This version had 17154 lines, 135647 words, 888333 characters, and 1602 entries.

Version 2.9.1, Jun 05 1991: last network release before book.  This version had 18610
lines, 146262 words, 957178 characters, and 1670 entries.

Version 2.9.2, Jun 21 1991: corresponds to reproduction copy for book. This version had
18911 lines, 1478291 words, 973269 characters, and 1697 entries.

Version numbering: Version numbers should be read as major.minor.revision.
Major version 1 is reserved for the `old' (ITS) Jargon File, jargon-1.
Major version 2 encompasses revisions by ESR (Eric S. Raymond) with assistance from GLS
(Guy L.  Steele, Jr.). Someday, the next maintainer will take over and spawn `version 3'.
Usually later versions will either completely supersede or incorporate earlier versions,
so there is generally no point in keeping old versions around.

Our thanks to the coauthors of Steele-1983 for oversight and assistance, and to the
hundreds of USENETters (too many to name here) who contributed entries and encouragement.
More thanks go to several of the old-timers on the USENET group alt.folklore.computers,
who contributed much useful commentary and many corrections and valuable historical
perspective: Joseph M. Newcomer <>, Bernie Cosell, Earl Boebert, and
Joe Morris.

We were fortunate enough to have the aid of some accomplished linguists. David Stampe and
Charles Hoequist contributed valuable criticism; Joe Keane helped us improve the
pronunciation guides.

A few bits of this text quote previous works.  We are indebted to Brian A. LaMacchia for
obtaining permission for us to use material from the `TMRC Dictionary'; also, Don Libes
contributed some appropriate material from his excellent book `Life With UNIX'.

We thank Per Lindberg, author of the remarkable Swedish-language 'zine `Hackerbladet',
for bringing `FOO!'  comics to our attention and smuggling one of the IBM hacker
underground's own baby jargon files out to us.  Thanks also to Maarten Litmaath for
generously allowing the inclusion of the ASCII pronunciation guide he formerly
maintained. Finally, Mark Brader and George V. Reilly submitted many thoughtful comments
and did yeoman service in catching typos and minor usage bobbles, and Eric Tiedemann
contributed sage advice on rhetoric, amphigory, and philosophunculism.

How Jargon Works

Jargon Construction

There are some standard methods of jargonification that became established quite early
(i.e., before 1970), spreading from such sources as the Tech Model Railroad Club, the
PDP-1 SPACEWAR hackers, and John McCarthy's original crew of LISPers.  These include the

Verb doubling ------------- A standard construction in English is to double a verb and use
it as an exclamation, such as "Bang, bang!" or "Quack, quack!".  Most of these are names
for noises.  Hackers also double verbs as a concise, sometimes sarcastic comment on what
the implied subject does.  Also, a doubled verb is often used to terminate a conversation,
in the process remarking on the current state of affairs or what the speaker intends to do
next.  Typical examples involve {win}, {lose}, {hack}, {flame}, {barf}, {chomp}:

     "The disk heads just crashed."  "Lose, lose."
     "Mostly he talked about his latest crock.  Flame, flame."
     "Boy, what a bagbiter!  Chomp, chomp!"

Some verb-doubled constructions have special meanings not immediately obvious from the
verb.  These have their own listings in the lexicon.

Soundalike slang ---------------- Hackers will often make rhymes or puns in order to
convert an ordinary word or phrase into something more interesting.  It is considered
particularly {flavorful} if the phrase is bent so as to include some other jargon word;
thus the computer hobbyist magazine `Dr. Dobb's Journal' is almost always referred to
among hackers as `Dr. Frob's Journal' or simply `Dr. Frob's'.  Terms of this kind that
have been in fairly wide use include names for newspapers:

     Boston Herald => Horrid (or Harried)
     Boston Globe =^gt; Boston Glob
     Houston (or San Francisco) Chronicle => the Crocknicle (or the Comical)
     New York Times => New York Slime

However, terms like these are often made up on the spur of the moment.
Standard examples include:

     Data General => Dirty Genitals
     IBM 360 => IBM Three-Sickly
     Government Property --- Do Not Duplicate (on keys) =>
     Government Duplicity --- Do Not Propagate
     for historical reasons => for hysterical raisins
     Margaret Jacks Hall (the CS building at Stanford) => Marginal Hacks Hall

This is not really similar to the Cockney rhyming slang it has been compared to in the
past, because Cockney substitutions are opaque whereas hacker punning jargon is
intentionally transparent.

The `-P' convention ------------------- Turning a word into a question by appending the
syllable `P'; from the LISP convention of appending the letter `P' to denote a predicate
(a boolean-valued function).  The question should expect a yes/no answer, though it
needn't.  (See {T} and {NIL}.)

     At dinnertime:
           Q: "Foodp?"
           A: "Yeah, I'm pretty hungry." or "T!"

     At any time:
           Q: "State-of-the-world-P?"
           A: (Straight) "I'm about to go home."
           A: (Humorous) "Yes, the world has a state."

     On the phone to Florida:
           Q: "State-p Florida?"
           A: "Been reading JARGON.TXT again, eh?"

[One of the best of these is a {Gosperism}.  Once, when we were at a Chinese restaurant,
Bill Gosper wanted to know whether someone would like to share with him a two-person-sized
bowl of soup.  His inquiry was: "Split-p soup?" --- GLS]

Overgeneralization ------------------ A very conspicuous feature of jargon is the
frequency with which techspeak items such as names of program tools, command language
primitives, and even assembler opcodes are applied to contexts outside of computing
wherever hackers find amusing analogies to them.  Thus (to cite one of the best-known
examples) UNIX hackers often {grep} for things rather than searching for them.
Many of the lexicon entries are generalizations of exactly this kind.

Hackers enjoy overgeneralization on the grammatical level as well.  Many hackers love to
take various words and add the wrong endings to them to make nouns and verbs, often by
extending a standard rule to nonuniform cases (or vice versa).  For example, because

     porous => porosity
     generous => generosity

hackers happily generalize:

     mysterious => mysteriosity
     ferrous => ferrosity
     obvious => obviosity
     dubious => dubiosity

Also, note that all nouns can be verbed.  E.g.: "All nouns can be verbed",
"I'll mouse it up", "Hang on while I clipboard it over", "I'm grepping the files".
English as a whole is already heading in this direction (towards pure-positional grammar
like Chinese); hackers are simply a bit ahead of the curve.

However, note that hackers avoid the unimaginative verb-making techniques characteristic
of marketroids, bean-counters, and the Pentagon; a hacker would never, for example,
`productize', `prioritize', or `securitize' things.  Hackers have a strong aversion to
bureaucratic bafflegab and regard those who use it with contempt.

Similarly, all verbs can be nouned.  This is only a slight overgeneralization in modern
English; in hackish, however, it is good form to mark them in some standard nonstandard
way.  Thus:

     win => winnitude, winnage
     disgust => disgustitude
     hack => hackification

Further, note the prevalence of certain kinds of nonstandard plural forms.  Some of these
go back quite a ways; the TMRC Dictionary noted that the defined plural of `caboose' is
`cabeese', and includes an entry which implies that the plural of `mouse' is {meeces}.
On a similarly Anglo-Saxon note, almost anything ending in `x' may form plurals in `-xen'
(see {VAXen} and {boxen} in the main text).  Even words ending in phonetic /k/ alone are
sometimes treated this way; e.g., `soxen' for a bunch of socks.  Other funny plurals are
`frobbotzim' for the plural of `frobbozz' (see {frobnitz}) and `Unices' and `Tenices'
(rather than `Unixes' and `Tenexes'; see {UNIX}, {TENEX} in main text).  But note that
`Unixen' and `Tenexen' are never used; it has been suggested that this is because `-ix'
and `-ex' are Latin singular endings that attract a Latinate plural.  Finally, it has
been suggested to general approval that the plural of `mongoose' ought to be `polygoose'.

The pattern here, as with other hackish grammatical quirks, is generalization of an
inflectional rule that in English is either an import or a fossil (such as the Hebrew
plural ending `-im', or the Anglo-Saxon plural suffix `-en') to cases where it isn't
normally considered to apply.

This is not `poor grammar', as hackers are generally quite well aware of what they are
doing when they distort the language.  It is grammatical creativity, a form of playfulness.
It is done not to impress but to amuse, and never at the expense of clarity.

Spoken inarticulations ---------------------- Words such as `mumble', `sigh', and `groan'
are spoken in places where their referent might more naturally be used.  It has been
suggested that this usage derives from the impossibility of representing such noises on
a comm link or in email (interestingly, the same sorts of constructions have been showing
up with increasing frequency in comic strips).  Another expression sometimes heard is
"Complain!", meaning "I have a complaint!"

Of the five listed constructions, verb doubling, peculiar noun formations, and (especially)
spoken inarticulations have become quite general; but punning jargon is still largely
confined to MIT and other large universities, and the `-P' convention is found only where
LISPers flourish.

Semantically, one rich source of jargon constructions is the hackish tendency to
anthropomorphize hardware and software.  This isn't done in a naive way; hackers don't
personalize their stuff in the sense of feeling empathy with it, nor do they mystically
believe that the things they work on every day are `alive'.  What *is* common is to hear
hardware or software talked about as though it has homunculi talking to each other inside
it, with intentions and desires.  Thus, one hears "The protocol handler got confused", or
that programs "are trying" to do things, or one may say of a routine that "its goal in life
is to X". One even hears explanations like "...  and its poor little brain couldn't
understand X, and it died."  Sometimes modelling things this way actually seems to make
them easier to understand, perhaps because it's instinctively natural to think of anything
with a really complex behavioral repertoire as `like a person' rather than `like a thing'.

Finally, note that many words in hacker jargon have to be understood as members of sets of
comparatives.  This is especially true of the adjectives and nouns used to describe the
beauty and functional quality of code.  Here is an approximately correct spectrum:

     monstrosity  brain-damage  screw  bug  lose  misfeature
     crock  kluge  hack  win  feature  elegance  perfection

The last is spoken of as a mythical absolute, approximated but never actually attained.
Another similar scale is used for describing the reliability of software:

     broken  flaky  dodgy  fragile  brittle
     solid  robust  bulletproof  armor-plated

Note, however, that `dodgy' is primarily Commonwealth hackish (it is rare in the U.S.) and
may change places with `flaky' for some speakers.

Coinages for describing {lossage} seem to call forth the very finest in hackish linguistic
inventiveness; it has been truly said that hackers have even more words for equipment
failures than Yiddish has for obnoxious people.

Hacker Writing Style

We've already seen that hackers often coin jargon by overgeneralizing grammatical rules.
This is one aspect of a more general fondness for form-versus-content language jokes that
shows up particularly in hackish writing.  One correspondent reports that he consistently
misspells `wrong' as `worng'.  Others have been known to criticize glitches in Jargon
File drafts by observing (in the mode of Douglas Hofstadter) "This sentence no verb", or
"Bad speling", or "Incorrectspa cing." Similarly, intentional spoonerisms are often made
of phrases relating to confusion or things that are confusing; `dain bramage' for `brain
damage' is perhaps the most common (similarly, a hacker would be likely to write "Excuse
me, I'm cixelsyd today", rather than "I'm dyslexic today").  This sort of thing is quite
common and is enjoyed by all concerned.

Hackers tend to use quotes as balanced delimiters like parentheses, much to the dismay
of American editors.  Thus, if "Jim is going" is a phrase, and so are "Bill runs" and
"Spock groks", then hackers generally prefer to write: "Jim is going", "Bill runs", and
"Spock groks".  This is incorrect according to standard American usage (which would put
the continuation commas and the final period inside the string quotes); however, it is
counter-intuitive to hackers to mutilate literal strings with characters that don't belong
in them.  Given the sorts of examples that can come up in discussions of programming,
American-style quoting can even be grossly misleading.  When communicating command lines
or small pieces of code, extra characters can be a real pain in the neck.

Consider, for example, a sentence in a {vi} tutorial that looks like this:

     Then delete a line from the file by typing "dd".

Standard usage would make this

     Then delete a line from the file by typing "dd."

but that would be very bad -- because the reader would be prone to type the string d-d-dot,
and it happens that in `vi(1)' dot repeats the last command accepted.  The net result would
be to delete *two* lines!

The Jargon File follows hackish usage throughout.

Interestingly, a similar style is now preferred practice in Great Britain, though the older
style (which became established for typographical reasons having to do with the aesthetics
of comma and quotes in typeset text) is still accepted there.  `Hart's Rules' and the
`Oxford Dictionary for Writers and Editors' call the hacker-like style `new' or `logical'

Another hacker quirk is a tendency to distinguish between `scare' quotes and `speech'
quotes; that is, to use British-style single quotes for marking and reserve American-style
double quotes for actual reports of speech or text included from elsewhere.
Interestingly, some authorities describe this as correct general usage, but mainstream
American English has gone to using double-quotes indiscriminately enough that hacker usage
appears marked [and, in fact, I thought this was a personal quirk of mine until I checked
with USENET --- ESR].  One further permutation that is definitely *not* standard is a
hackish tendency to do marking quotes by using apostrophes in pairs; that is, 'like this'.
This is modelled on string and character literal syntax in some programming languages
(reinforced by the fact that many character-only terminals display the apostrophe in
typewriter style, as a vertical single quote).

One quirk that shows up frequently in the {email} style of UNIX hackers in particular is
a tendency for some things that are normally all-lowercase (including usernames and the
names of commands and C routines) to remain uncapitalized even when they occur at the
beginning of sentences.  It is clear that, for many hackers, the case of such identifiers
becomes a part of their internal representation (the `spelling') and cannot be overridden
without mental effort (an appropriate reflex because UNIX and C both distinguish cases and
confusing them can lead to {lossage}).  A way of escaping this dilemma is simply to avoid
using these constructions at the beginning of sentences.

There seems to be a meta-rule behind these nonstandard hackerisms to the effect that
precision of expression is more important than conformance to traditional rules; where the
latter create ambiguity or lose information they can be discarded without a second thought.
It is notable in this respect that other hackish inventions (for example, in vocabulary)
also tend to carry very precise shades of meaning even when constructed to appear slangy and
loose.  In fact, to a hacker, the contrast between `loose' form and `tight' content in
jargon is a substantial part of its humour!

Hackers have also developed a number of punctuation and emphasis conventions adapted to
single-font all-ASCII communications links, and these are occasionally carried over into
written documents even when normal means of font changes, underlining, and the like are

One of these is that TEXT IN ALL CAPS IS INTERPRETED AS `LOUD', and this becomes such an
ingrained synesthetic reflex that a person who goes to caps-lock while in {talk mode} may
be asked to "stop shouting, please, you're hurting my ears!".

Also, it is common to use bracketing with unusual characters to signify emphasis.  The
asterisk is most common, as in "What the *hell*?" even though this interferes with the
common use of the asterisk suffix as a footnote mark.  The underscore is also common,
suggesting underlining (this is particularly common with book titles; for example, "It
is often alleged that Joe Haldeman wrote _The_Forever_War_ as a rebuttal to Robert
Heinlein's earlier novel of the future military, _Starship_Troopers_.").  Other forms
exemplified by "=hell=", "\hell/", or "/hell/" are occasionally seen (it's claimed that in
the last example the first slash pushes the letters over to the right to keep them from
falling down, and the second keeps them from falling over).  Finally, words may also be
emphasized line of the text.

There is a semantic difference between *emphasis like this* (which emphasizes the phrase
as a whole), and *emphasis* *like* *this* (which suggests the writer speaking very slowly
and distinctly, as if to a very young child or a mentally impaired person).  Bracketing a
word with the `*' character may also indicate that the writer wishes readers to consider
that an action is taking place or that a sound is being made.
Examples: *bang*, *hic*, *ring*, *grin*, *kick*, *stomp*, *mumble*.

There is also an accepted convention for `writing under erasure'; the text

     Be nice to this fool^H^H^H^Hgentleman, he's in from corporate HQ.

would be read as "Be nice to this fool, I mean this gentleman...".  This comes from the
fact that the digraph ^H is often used as a print representation for a backspace.  It
parallels (and may have been influenced by) the ironic use of `slashouts' in science-fiction

In a formula, `*' signifies multiplication but two asterisks in a row are a shorthand for
exponentiation (this derives from FORTRAN).  Thus, one might write 2 ** 8 = 256.

Another notation for exponentiation one sees more frequently uses the caret (^, ASCII
1011110); one might write instead `2^8 = 256'.  This goes all the way back to Algol-60,
which used the archaic ASCII `up-arrow' that later became the caret; this was picked up
by Kemeny and Kurtz's original BASIC, which in turn influenced the design of the `bc(1)'
and `dc(1)' UNIX tools, which have probably done most to reinforce the convention on USENET.
The notation is mildly confusing to C programmers, because `^' means logical {XOR} in C.
Despite this, it was favoured 3:1 over ** in a late-1990 snapshot of USENET.  It is used
consistently in this text.

In on-line exchanges, hackers tend to use decimal forms or improper fractions (`3.5' or
`7/2') rather than `typewriter style' mixed fractions (`3-1/2').  The major motive here is
probably that the former are more readable in a monospaced font, together with a desire to
avoid the risk that the latter might be read as `three minus one-half'.  The decimal form
is definitely preferred for fractions with a terminating decimal representation; there may
be some cultural influence here from the high status of scientific notation.

Another on-line convention, used especially for very large or very small numbers, is taken
from C (which derived it from FORTRAN).  This is a form of `scientific notation' using `e'
to replace `*10^'; for example, one year is about 3e7 seconds long.

The tilde (~) is commonly used in a quantifying sense of `approximately'; that is, `~50'
means `about fifty'.

On USENET and in the {MUD} world, common C boolean, logical, and relational operators
such as `|', `&', `||', `&&', `!', `==', `!=', `>', and `<', `>=', and `=<' are often
combined with English.  The Pascal not-equals, `<>', is also recognized, and occasionally
one sees `/=' for not-equals (from Ada, Common Lisp, and Fortran 90).  The use of prefix
`!' as a loose synonym for `not-' or `no-' is particularly common; thus, `!clue' is read
`no-clue' or `clueless'.

Another habit is that of using angle-bracket enclosure to genericize a term; this derives
from conventions used in {BNF}.  Uses like the following are common:

     So this <ethnic> walks into a bar one day, and...

Hackers also mix letters and numbers more freely than in mainstream usage.  In particular,
it is good hackish style to write a digit sequence where you intend the reader to
understand the text string that names that number in English.  So, hackers prefer to
write `1970s' rather than `nineteen-seventies' or `1970's' (the latter looks like a

It should also be noted that hackers exhibit much less reluctance to use multiply nested
parentheses than is normal in English.  Part of this is almost certainly due to influence
from LISP (which uses deeply nested parentheses (like this (see?)) in its syntax a lot),
but it has also been suggested that a more basic hacker trait of enjoying playing with
complexity and pushing systems to their limits is in operation.

One area where hackish conventions for on-line writing are still in some flux is the
marking of included material from earlier messages --- what would be called `block
quotations' in ordinary English.  From the usual typographic convention employed for
these (smaller font at an extra indent), there derived the notation of included text
being indented by one ASCII TAB (0001001) character, which under UNIX and many other
environments gives the appearance of an 8-space indent.

Early mail and netnews readers had no facility for including messages this way, so people
had to paste in copy manually.  BSD `Mail(1)' was the first message agent to support
inclusion, and early USENETters emulated its style.  But the TAB character tended to push
included text too far to the right (especially in multiply nested inclusions), leading to
ugly wraparounds.  After a brief period of confusion (during which an inclusion leader
consisting of three or four spaces became established in EMACS and a few mailers), the
use of leading `>' or `> ' became standard, perhaps owing to its use in `ed(1)' to display
tabs (alternatively, it may derive from the `>' that some early UNIX mailers used to quote
lines starting with "From" in text, so they wouldn't look like the beginnings of new
message headers).  Inclusions within inclusions keep their `>' leaders, so the `nesting
level' of a quotation is visually apparent.

A few other idiosyncratic quoting styles survive because they are automatically generated.
One particularly ugly one looks like this:

     /* Written hh:mm pm  Mmm dd, yyyy by user@site in <group> */
     /* ---------- "Article subject, chopped to 35 ch" ---------- */
        <quoted text>
     /* End of text from local:group */

It is generated by an elderly, variant news-reading system called `notesfiles'.  The
overall trend, however, is definitely away from such verbosity.

The practice of including text from the parent article when posting a followup helped solve
what had been a major nuisance on USENET: the fact that articles do not arrive at different
sites in the same order. Careless posters used to post articles that would begin with, or
even consist entirely of, "No, that's wrong" or "I agree" or the like.  It was hard to see
who was responding to what.  Consequently, around 1984, new news-posting software evolved a
facility to automatically include the text of a previous article, marked with "> " or
whatever the poster chose.  The poster was expected to delete all but the relevant lines.
The result has been that, now, careless posters post articles containing the *entire* text
of a preceding article, *followed* only by "No, that's wrong" or "I agree".

Many people feel that this cure is worse than the original disease, and there soon appeared
newsreader software designed to let the reader skip over included text if desired.  Today,
some posting software rejects articles containing too high a proportion of lines beginning
with `>' -- but this too has led to undesirable workarounds, such as the deliberate
inclusion of zero-content filler lines which aren't quoted and thus pull the message below
the rejection threshold.

Because the default mailers supplied with UNIX and other operating systems haven't evolved
as quickly as human usage, the older conventions using a leading TAB or three or four
spaces are still alive; however,>-inclusion is now clearly the prevalent form in both
netnews and mail.

In 1991 practice is still evolving, and disputes over the `correct' inclusion style
occasionally lead to {holy wars}.  One variant style reported uses the citation character
`|' in place of `>' for extended quotations where original variations in indentation are
being retained.
One also sees different styles of quoting a number of authors in the same message: one
(deprecated because it loses information) uses a leader of `> ' for everyone, another
(the most common) is `>>>> ', `>>> ', etc. (or `>>>> ', `>>> ', etc., depending on
line length and nesting depth) reflecting the original order of messages, and yet another
is to use a different citation leader for each author, say `> ', `: ', `| ', `} '
(preserving nesting so that the inclusion order of messages is still apparent, or tagging
the inclusions with authors' names).  Yet *another* style is to use each poster's initials
(or login name) as a citation leader for that poster.  Occasionally one sees a `# ' leader
used for quotations from authoritative sources such as standards documents; the intended
allusion is to the root prompt (the special UNIX command prompt issued when one is running
as the privileged super-user).

Finally, it is worth mentioning that many studies of on-line communication have shown that
electronic links have a de-inhibiting effect on people.  Deprived of the body-language cues
through which emotional state is expressed, people tend to forget everything about other
parties except what is presented over that ASCII link.  This has both good and bad effects.
The good one is that it encourages honesty and tends to break down hierarchical authority
relationships; the bad is that it may encourage depersonalization and gratuitous rudeness.
Perhaps in response to this, experienced netters often display a sort of conscious formal
politesse in their writing that has passed out of fashion in other spoken and written
media (for example, the phrase "Well said, sir!" is not uncommon).

Many introverted hackers who are next to inarticulate in person communicate with
considerable fluency over the net, perhaps precisely because they can forget on an
unconscious level that they are dealing with people and thus don't feel stressed and
anxious as they would face to face.

Though it is considered gauche to publicly criticize posters for poor spelling or grammar,
the network places a premium on literacy and clarity of expression.  It may well be that
future historians of literature will see in it a revival of the great tradition of personal
letters as art.

Hacker Speech Style

Hackish speech generally features extremely precise diction, careful word choice, a
relatively large working vocabulary, and relatively little use of contractions or street
slang.  Dry humour, irony, puns, and a mildly flippant attitude are highly valued --- but
an underlying seriousness and intelligence are essential.  One should use just enough jargon
to communicate precisely and identify oneself as a member of the culture; overuse of jargon
or a breathless, excessively gung-ho attitude is considered tacky and the mark of a loser.

This speech style is a variety of the precisionist English normally spoken by scientists,
design engineers, and academics in technical fields.  In contrast with the methods of
jargon construction, it is fairly constant throughout hackerdom.

It has been observed that many hackers are confused by negative questions --- or, at least,
that the people to whom they are talking are often confused by the sense of their answers.
The problem is that they have done so much programming that distinguishes between

     if (going) {


     if (!going) {

that when they parse the question "Aren't you going?" it seems to be asking the opposite
question from "Are you going?", and so merits an answer in the opposite sense.  This
confuses English-speaking non-hackers because they were taught to answer as though the
negative part weren't there.  In some other languages (including Russian, Chinese, and
Japanese) the hackish interpretation is standard and the problem wouldn't arise.  Hackers
often find themselves wishing for a word like French `si' or German `doch' with which one
could unambiguously answer `yes' to a negative question.

For similar reasons, English-speaking hackers almost never use double negatives, even if
they live in a region where colloquial usage allows them.  The thought of uttering something
that logically ought to be an affirmative knowing it will be misparsed as a negative tends
to disturb them.

International Style

Although the Jargon File remains primarily a lexicon of hacker usage in American English,
we have made some effort to get input from abroad. Though the hacker-speak of other
languages often uses translations of jargon from English (often as transmitted to them by
earlier Jargon File versions!), the local variations are interesting, and knowledge of them
may be of some use to travelling hackers.

There are some references herein to `Commonwealth English'.  These are intended to describe
some variations in hacker usage as reported in the English spoken in Great Britain and the
Commonwealth (Canada, Australia, India, etc. --- though Canada is heavily influenced by
American usage). There is also an entry on {{Commonwealth Hackish}} reporting some general
phonetic and vocabulary differences from U.S. hackish.

Hackers in Western Europe and (especially) Scandinavia are reported to often use a mixture
of English and their native languages for technical conversation.  Occasionally they develop
idioms in their English usage that are influenced by their native-language styles.  Some of
these are reported here.

A few notes on hackish usages in Russian have been added where they are parallel with
English idioms and thus comprehensible to English-speakers.

How to Use the Lexicon

Pronunciation Guide

Pronunciation keys are provided in the jargon listings for all entries that are neither
dictionary words pronounced as in standard English nor obvious compounds thereof.  Slashes
bracket phonetic pronunciations, which are to be interpreted using the following conventions:

  1. Syllables are hyphen-separated, except that an accent or back-accent follows each
     accented syllable (the back-accent marks a secondary accent in some words of four or
     more syllables).

  2. Consonants are pronounced as in American English.  The letter `g' is always hard (as in
     "got" rather than "giant"); `ch' is soft ("church" rather than "chemist").  The letter
     `j' is the sound that occurs twice in "judge".  The letter `s' is always as in "pass",
     never a z sound.  The digraph `kh' is the guttural of "loch" or "l'chaim".

  3. Uppercase letters are pronounced as their English letter names; thus (for example)
     /H-L-L/ is equivalent to /aitch el el/.  /Z/ may be pronounced /zee/ or /zed/
     depending on your local dialect.

  4. Vowels are represented as follows:

            back, that
            far, mark
            flaw, caught
            bake, rain
            less, men
            easy, ski
            their, software
            trip, hit
            life, sky
            father, palm
            flow, sew
            loot, through
            more, door
            out, how
            boy, coin
            but, some
            put, foot
            yet, young
            few, chew
            /oo/ with optional fronting as in `news' (/nooz/ or /nyooz/)

A /*/ is used for the `schwa' sound of unstressed or occluded vowels (the one that is often
written with an upside-down `e').  The schwa vowel is omitted in syllables containing
vocalic r, l, m or n; that is, `kitten' and `color' would be rendered /kit'n/ and /kuhl'r/,
not /kit'*n/ and /kuhl'*r/.

Entries with a pronunciation of `//' are written-only usages.  (No, UNIX weenies, this does
*not* mean `pronounce like previous pronunciation'!)

Other Lexicon Conventions

Entries are sorted in case-blind ASCII collation order (rather than the letter-by-letter
order ignoring interword spacing common in mainstream dictionaries), except that all entries
beginning with nonalphabetic characters are sorted after Z.  The case-blindness is a feature,
not a bug.

In pure ASCII renderings of the Jargon File, you will see {} used to bracket words which
themselves have entries in the File.  This isn't done all the time for every such word, but
it is done everywhere that a reminder seems useful that the term has a jargon meaning and
one might wish to refer to its entry.

In this all-ASCII version, headwords for topic entries are distinguished from those for
ordinary entries by being followed by "::" rather than ":"; similarly, references are
surrounded by "{{" and "}}" rather than "{" and "}".

Defining instances of terms and phrases appear in `slanted type'.  A defining instance is
one which occurs near to or as part of an explanation of it.
Prefix * is used as linguists do; to mark examples of incorrect usage.

We follow the `logical' quoting convention described in the Writing Style section above.
In addition, we reserve double quotes for actual excerpts of text or (sometimes invented)
speech.  Scare quotes (which mark a word being used in a nonstandard way), and philosopher's
quotes (which turn an utterance into the string of letters or words that name it) are both
rendered with single quotes.

References such as `malloc(3)' and `patch(1)' are to UNIX facilities (some of which, such as
`patch(1)', are actually freeware distributed over USENET).  The UNIX manuals use `foo(n)'
to refer to item foo in section (n) of the manual, where n=1 is utilities, n=2 is system
calls, n=3 is C library routines, n=6 is games, and n=8 (where present) is system
administration utilities.  Sections 4, 5, and 7 of the manuals have changed roles frequently
and in any case are not referred to in any of the entries.

Various abbreviations used frequently in the lexicon are summarized here:

     synonym (or synonymous with)
     verb (may be transitive or intransitive)
     intransitive verb
     transitive verb

Where alternate spellings or pronunciations are given, alt. separates two possibilities with
nearly equal distribution, while var. prefixes one that is markedly less common than the

Where a term can be attributed to a particular subculture or is known to have originated
there, we have tried to so indicate.  Here is a list of abbreviations used in etymologies:

     University of California at Berkeley
     the university in England (*not* the city in Massachusetts where MIT happens to be
     Bolt, Beranek & Newman
     Carnegie-Mellon University
     Commodore Business Machines
     The Digital Equipment Corporation
     The Fairchild Instruments Palo Alto development group
     See the {Fidonet} entry
     International Business Machines
     Massachusetts Institute of Technology; esp. the legendary MIT AI Lab culture of roughly
     1971 to 1983 and its feeder groups, including the Tech Model Railroad Club
     New York University
     The Oxford English Dictionary
     Purdue University
     Stanford Artificial Intelligence Laboratory (at Stan-ford University)
    >From Syst`eme International, the name for the standard conventions of metric
     nomenclature used in the sciences
     Stanford University
     Sun Microsystems
     Some MITisms go back as far as the Tech Model Railroad Club (TMRC) at MIT c. 1960.
     Material marked TMRC is from `An Abridged Dictionary of the TMRC Language', originally
     compiled by Pete Samson in 1959
     University of California at Los Angeles
     the United Kingdom (England, Wales, Scotland, Northern Ireland)
     See the {USENET} entry
     Worcester Polytechnic Institute, site of a very active community of PDP-10 hackers
     during the 1970s
     XEROX's Palo Alto Research Center, site of much pioneering research in user interface
     design and networking
     Yale University

Some other etymology abbreviations such as {UNIX} and {PDP-10} refer to technical cultures
surrounding specific operating systems, processors, or other environments.  The fact that a
term is labelled with any one of these abbreviations does not necessarily mean its use is
confined to that culture.  In particular, many terms labelled `MIT' and `Stanford' are in
quite general use.  We have tried to give some indication of the distribution of speakers
in the usage notes; however, a number of factors mentioned in the introduction conspire to
make these indications less definite than might be desirable.

A few new definitions attached to entries are marked [proposed]. These are usually
generalizations suggested by editors or USENET respondents in the process of commenting on
previous definitions of those entries.  These are *not* represented as established jargon.

The Jargon Lexicon

JGC Logo Valid HTML5 Logo HTML5 Logo Valid CSS3 Logo JGC Logo
Copyright logo
This page (jargon1.html) was last modified on Sunday 27/01/2013