1ROFF(7) Miscellaneous Information Manual ROFF(7)
2
3
4
6 roff - concepts and history of roff typesetting
7
9 roff is the general name for a set of text formatting programs, known
10 under names like troff, nroff, ditroff, groff, etc. A roff system con‐
11 sists of an extensible text formatting language and a set of programs
12 for printing and converting to other text formats. Unix-like operating
13 systems distribute a roff system as a core package.
14
15 The most common roff system today is the free software implementation
16 GNU roff, groff(1). groff implements the look-and-feel and functional‐
17 ity of its ancestors, with many extensions.
18
19 The ancestry of roff is described in section “History” below. In this
20 document, the term roff always refers to the general class of roff pro‐
21 grams, not to the roff command provided in early Unix systems.
22
23 In spite of its age, roff is in wide use today, for example, the manual
24 pages on Unix systems (man pages), many software books, system documen‐
25 tation, standards, and corporate documents are written in roff. The
26 roff output for text devices is still unmatched, and its graphical out‐
27 put has the same quality as other free type-setting programs and is
28 better than some of the commercial systems.
29
30 roff is used to format Unix manual pages, (or man pages), the standard
31 documentation system on many Unix-derived operating systems.
32
33 This document describes the history of the development of the roff sys‐
34 tem; some usage aspects common to all roff versions, details on the
35 roff pipeline, which is usually hidden behind front-ends like groff(1);
36 a general overview of the formatting language; some tips for editing
37 roff files; and many pointers to further readings.
38
40 Document formatting by computer dates back to the 1960s. The roff sys‐
41 tem itself is intimately connected to the Unix operating system, but
42 its roots go back to the earlier operating systems CTSS and Multics.
43
44 The Predecessor RUNOFF
45 roff's ancestor RUNOFF was written in the MAD language by Jerry Saltzer
46 for the Compatible Time Sharing System (CTSS), a project of the Massa‐
47 chusetts Institute of Technology (MIT), in 1963 and 1964—note that CTSS
48 commands were all uppercase.
49
50 In 1965, MIT's Project MAC teamed with Bell Telephone Laboratories
51 (BTL) and General Electric to begin the Multics system ⟨http://
52 www.multicians.org⟩. A command called runoff was written for Multics
53 in the late 60s in the BCPL language, by Bob Morris, Doug McIlroy, and
54 other members of the Multics team.
55
56 Like its CTSS ancestor, Multics runoff formatted an input file consist‐
57 ing of text and command lines; commands began with a period and were
58 two letters. Output from these commands was to terminal devices such
59 as IBM Selectric terminals. Multics runoff had additional features
60 added, such as the ability to do two-pass formatting; it became the
61 main format for Multics documentation and text processing.
62
63 BCPL and runoff were ported to the GCOS system at Bell Labs when BTL
64 left the development of Multics.
65
66 There is a free archive about historical RUNOFF documents. You can get
67 it anonymously by the shell command
68 $git clone https://github.com/bwarken/RUNOFF_historical.git
69
70 As well, there is a new project for writing a program that can read
71 RUNOFF files , but it does not yet work so far. You can get an early
72 version anonymously by the shell command
73 $git clone https://github.com/bwarken/runoff.git
74
75 The Classical nroff/troff System
76 At BTL, there was a need to drive the Graphic Systems CAT typesetter, a
77 graphical output device from a PDP-11 computer running Unix. As runoff
78 was too limited for this task it was further developed into a more pow‐
79 erful text formatting system by Joseph F. Ossanna, who already pro‐
80 grammed several runoff ports.
81
82 The name runoff was shortened to roff. The greatly enlarged language
83 of Ossanna's version already included all elements of a full roff sys‐
84 tem. All modern roff systems try to implement compatibility to this
85 system. So Joe Ossanna can be called the father of all roff systems.
86
87 This first roff system had three formatter programs.
88
89 troff (typesetter roff) generated a graphical output for the CAT type‐
90 setter as its only device.
91
92 nroff produced text output suitable for terminals and line printers.
93
94 roff was the reimplementation of the former runoff program with its
95 limited features; this program was abandoned in later versions.
96 Today, the name roff is used to refer to a troff/nroff system as
97 a whole.
98
99 Ossanna's first version was written in the PDP-11 assembly language and
100 released in 1973. Brian Kernighan joined the roff development by
101 rewriting it in the C programming language. The C version was released
102 in 1975.
103
104 The syntax of the formatting language of the nroff/troff programs was
105 documented in the famous Troff User's Manual [CSTR #54], first pub‐
106 lished in 1976, with further revisions up to 1992 by Brian Kernighan.
107 This document is the specification of the classical troff. All later
108 roff systems tried to establish compatibility with this specification.
109
110 After Ossanna's death in 1977, Kernighan went on with developing troff.
111 In the late 1970s, Kernighan equipped troff with a general interface to
112 support more devices, the intermediate output format, and the postpro‐
113 cessor system. This completed the structure of a roff system as it is
114 still in use today; see section “Using Roff” below. In 1979, these
115 novelties were described in the paper [CSTR #97]. This new troff ver‐
116 sion is the basis for all existing newer troff systems, including
117 groff. On some systems, this device independent troff got a binary of
118 its own, called ditroff(7). All modern troff programs already provide
119 the full ditroff capabilities automatically.
120
121 Availability
122 The source code of both the ancient Unix and classical troff weren't
123 available for two decades. Nowadays, it is accessible again (on-line)
124 for non-commercial use; see SEE ALSO, below.
125
126 groff — free GNU roff
127 The most important free roff project was the GNU implementation of
128 troff, written from scratch by James Clark and put under the GNU Public
129 License ⟨http://www.gnu.org/copyleft⟩. It was called groff (GNU roff).
130 See groff(1) for an overview.
131
132 The groff system is still actively developed. It is compatible to the
133 classical troff, but many extensions were added. It is the first roff
134 system that is available on almost all operating systems — and it is
135 free. This makes groff the de facto roff standard today.
136
137 Free Heirloom roff
138 An alternative is Gunnar Ritter's Heirloom roff project ⟨https://
139 github.com/n-t-roff/heirloom-doctools⟩ project, started in 2005, which
140 provides enhanced versions of the various roff tools found in the Open‐
141 Solaris and Plan 9 operating systems, now available under free li‐
142 censes. You can get this package with the shell command:
143 $ git clone https://github.com/n-t-roff/heirloom-doctools
144
145 Moreover, one finds there the Original Documenter's Workbench Release
146 3.3 ⟨https://github.com/n-t-roff/DWB3.3⟩.
147
149 Most people won't even notice that they are actually using roff. When
150 you read a system manual page (man page) roff is working in the back‐
151 ground. But using roff explicitly isn't difficult either.
152
153 Some roff implementations provide wrapper programs that make it easy to
154 use the roff system on the shell command line. For example, the GNU
155 roff implementation groff(1) provides command-line options to avoid the
156 long command pipes of classical troff; a program grog(1) tries to guess
157 from the document which arguments should be used for a run of groff;
158 people who do not like specifying command-line options should try the
159 groffer(1) program for graphically displaying groff files and man
160 pages.
161
162 The roff Pipe
163 Each roff system consists of preprocessors, roff formatter programs,
164 and a set of device postprocessors. This concept makes heavy use of
165 the piping mechanism, that is, a series of programs is called one after
166 the other, where the output of each program in the queue is taken as
167 the input for the next program.
168
169 cat file | ... | preproc | ... | troff options | postproc
170
171 The preprocessors generate roff code that is fed into a roff formatter
172 (e.g. troff), which in turn generates intermediate output that is fed
173 into a device postprocessor program for printing or final output.
174
175 All of these parts use programming languages of their own; each lan‐
176 guage is totally unrelated to the other parts. Moreover, roff macro
177 packages that were tailored for special purposes can be included.
178
179 Most roff documents use the macros of some package, intermixed with
180 code for one or more preprocessors, spiced with some elements from the
181 plain roff language. The full power of the roff formatting language is
182 seldom needed by users; only programmers of macro packages need to know
183 about the gory details.
184
185 Preprocessors
186 A roff preprocessor is any program that generates output that syntacti‐
187 cally obeys the rules of the roff formatting language. Each preproces‐
188 sor defines a language of its own that is translated into roff code
189 when run through the preprocessor program. Parts written in these lan‐
190 guages may be included within a roff document; they are identified by
191 special roff requests or macros. Each document that is enhanced by
192 preprocessor code must be run through all corresponding preprocessors
193 before it is fed into the actual roff formatter program, for the for‐
194 matter just ignores all alien code. The preprocessor programs extract
195 and transform only the document parts that are determined for them.
196
197 There are a lot of free and commercial roff preprocessors. Some of
198 them aren't available on each system, but there is a small set of pre‐
199 processors that are considered as an integral part of each roff system.
200 The classical preprocessors are
201
202 tbl for tables.
203 eqn for mathematical formulae.
204 pic for drawing diagrams.
205 refer for bibliographic references.
206 soelim for including macro files from standard locations.
207 chem for drawing chemical formulæ.
208
209 Other known preprocessors that are not available on all systems include
210
211 grap for constructing graphical elements.
212 grn for including gremlin(1) pictures.
213
214 Formatter Programs
215 A roff formatter is a program that parses documents written in the roff
216 formatting language or uses some of the roff macro packages. It gener‐
217 ates intermediate output, which is intended to be fed into a single de‐
218 vice postprocessor that must be specified by a command-line option to
219 the formatter program. The documents must have been run through all
220 necessary preprocessors before.
221
222 The output produced by a roff formatter is represented in yet another
223 language, the intermediate output format or troff output. This lan‐
224 guage was first specified in [CSTR #97]; its GNU extension is docu‐
225 mented in groff_out(5). The intermediate output language is a kind of
226 assembly language compared to the high-level roff language. The gener‐
227 ated intermediate output is optimized for a special device, but the
228 language is the same for every device.
229
230 The roff formatter is the heart of the roff system. The traditional
231 roff had two formatters, nroff for text devices and troff for graphical
232 devices.
233
234 Often, the name troff is used as a general term to refer to both for‐
235 matters.
236
237 Devices and Postprocessors
238 Devices are hardware interfaces like printers, text or graphical termi‐
239 nals, etc., or software interfaces such as a conversion into a differ‐
240 ent text or graphical format.
241
242 A roff postprocessor is a program that transforms troff output into a
243 form suitable for a special device. The roff postprocessors are like
244 device drivers for the output target.
245
246 For each device there is a postprocessor program that fits the device
247 optimally. The postprocessor parses the generated intermediate output
248 and generates device-specific code that is sent directly to the device.
249
250 The names of the devices and the postprocessor programs are not fixed
251 because they greatly depend on the software and hardware abilities of
252 the actual computer. For example, the classical devices mentioned in
253 [CSTR #54] have greatly changed since the classical times. The old
254 hardware doesn't exist any longer and the old graphical conversions
255 were quite imprecise when compared to their modern counterparts.
256
257 For example, the PostScript device post in classical troff had a reso‐
258 lution of 720 units per inch, while groff's ps device has 72000, a re‐
259 finement of factor 100.
260
261 Today the operating systems provide device drivers for most printer-
262 like hardware, so it isn't necessary to write a special hardware post‐
263 processor for each printer.
264
266 Documents using roff are normal text files decorated by roff formatting
267 elements. The roff formatting language is quite powerful; it is almost
268 a full programming language and provides elements to enlarge the lan‐
269 guage. With these, it became possible to develop macro packages that
270 are tailored for special applications. Such macro packages are much
271 handier than plain roff. So most people will choose a macro package
272 without worrying about the internals of the roff language.
273
274 Macro Packages
275 Macro packages are collections of macros that are suitable to format a
276 special kind of documents in a convenient way. This greatly eases the
277 usage of roff. The macro definitions of a package are kept in a file
278 called name.tmac (classically tmac.name). All tmac files are stored in
279 one or more directories at standardized positions. Details on the nam‐
280 ing of macro packages and their placement is found in groff_tmac(5).
281
282 A macro package that is to be used in a document can be announced to
283 the formatter by the command-line option -m, see troff(1), or it can be
284 specified within a document using the file inclusion requests of the
285 roff language, see groff(7).
286
287 Famous classical macro packages are man for traditional man pages, mdoc
288 for BSD-style manual pages; the macro sets for books, articles, and
289 letters are me (probably from the first name of its creator Eric All‐
290 man), ms (from Manuscript Macros), and mm (from Memorandum Macros).
291
292 The roff Formatting Language
293 The classical roff formatting language is documented in the Troff
294 User's Manual [CSTR #54]. The roff language is a full programming lan‐
295 guage providing requests, definition of macros, escape sequences,
296 string variables, number or size registers, and flow controls.
297
298 Requests are the predefined basic formatting commands similar to the
299 commands at the shell prompt. The user can define request-like ele‐
300 ments using predefined roff elements. These are then called macros. A
301 document writer will not note any difference in usage for requests or
302 macros; both are written on a line on their own starting with a dot.
303
304 Escape sequences are roff elements starting with a backslash ‘\’. They
305 can be inserted anywhere, also in the midst of text in a line. They
306 are used to implement various features, including the insertion of non-
307 ASCII characters with \(, font changes with \f, in-line comments with
308 \", the escaping of special control characters like \\, and many other
309 features.
310
311 Strings are variables that can store a string. A string is stored by
312 the .ds request. The stored string can be retrieved later by the \*
313 escape sequence.
314
315 Registers store numbers and sizes. A register can be set with the re‐
316 quest .nr and its value can be retrieved by the escape sequence \n.
317
319 Manual pages (man pages) take the section number as a file name exten‐
320 sion, e.g., the filename for this document is roff.7, i.e., it is kept
321 in section 7 of the man pages.
322
323 The classical macro packages take the package name as an extension,
324 e.g. file.me for a document using the me macro package, file.mm for mm,
325 file.ms for ms, file.pic for pic files, etc.
326
327 But there is no general naming scheme for roff documents, though
328 file.tr for troff file is seen now and then. Maybe there should be a
329 standardization for the filename extensions of roff files.
330
331 File name extensions can be very handy in conjunction with the less(1)
332 pager. It provides the possibility to feed all input into a command-
333 line pipe that is specified in the shell environment variable LESSOPEN.
334 This process is not well documented, so here an example:
335
336 LESSOPEN='|lesspipe %s'
337
338 where lesspipe is either a system supplied command or a shell script of
339 your own.
340
341 More details for file name extensions can be found at groff_file‐
342 names(5).
343
345 All roff formatters provide automated line breaks and horizontal and
346 vertical spacing. In order to not disturb this, the following tips can
347 be helpful.
348
349 • Never include empty or blank lines in a roff document. Instead,
350 use the empty request (a line consisting of a dot only) or a
351 line comment .\" if a structuring element is needed.
352
353 • Never start a line with whitespace because this can lead to un‐
354 expected behavior. Indented paragraphs can be constructed in a
355 controlled way by roff requests.
356
357 • Start each sentence on a line of its own, for the spacing after
358 a dot is handled differently depending on whether it terminates
359 an abbreviation or a sentence. To distinguish both cases, do a
360 line break after each sentence.
361
362 • To additionally use the auto-fill mode in Emacs, it is best to
363 insert an empty roff request (a line consisting of a dot only)
364 after each sentence.
365
366 The following example shows judicious line breaking in a roff input
367 file.
368
369 This is an example of a
370 .I roff
371 document that you can type into your text editor.
372 .
373 This is the next sentence in the same paragraph.
374 .
375 This is a longer sentence stretching over several input lines;
376 abbreviations like cf. are easily identified because the dot is
377 not followed by a line break.
378 .
379 In the output, this sentence continues the same paragraph.
380
381 Editing with Emacs
382 The best program for editing a roff document is Emacs (or XEmacs); see
383 emacs(1). It provides an nroff mode that is suitable for all kinds of
384 roff dialects. This mode can be activated by the following methods.
385
386 When editing a file within Emacs the mode can be changed by typing ‘M-x
387 nroff-mode’, where M-x means to hold down the Meta key (or Alt) and
388 press the x key at the same time.
389
390 But it is also possible to have the mode automatically selected when
391 the file is loaded into the editor.
392
393 • The most general method is to include the following 3 comment
394 lines at the end of the file.
395
396 .\" Local Variables:
397 .\" mode: nroff
398 .\" End:
399
400 • There is a set of file name extensions, e.g. the man pages that
401 trigger the automatic activation of the nroff mode.
402
403 • Theoretically, it is possible to write the sequence
404
405 .\" -*- nroff -*-
406
407 as the first line of a file to have it started in nroff mode
408 when loaded. Unfortunately, some applications such as the man
409 program are confused by this; so this is deprecated.
410
411 Editing with Vim
412 Besides Emacs, some other editors provide nroff style files too, e.g.
413 vim(1), an extension of the vi(1) program. Vim's highlighting can be
414 made to recognize roff files by setting the filetype option in a Vim
415 modeline. For this feature to work, your copy of vim must be built
416 with support for, and configured to enable, several features; consult
417 the editor's online help topics “auto-setting”, “filetype”, and “syn‐
418 tax”. Then put the following at the end of your roff files, after any
419 Emacs configuration:
420
421 .\" vim: set filetype=groff:
422
423 Replace “groff” in the above with “nroff” if you want highlighing that
424 does not recognize many of the GNU extensions to roff, such as request,
425 register, and string names longer than two characters.
426
428 This document was written by Bernd Warken ⟨groff-bernd.warken-72@
429 web.de⟩.
430
432 There is a lot of documentation on roff. The original papers on clas‐
433 sical troff are still available, and all aspects of groff are docu‐
434 mented in great detail.
435
436 Internet sites
437 History of Unix Manpages
438 The history page ⟨http://manpages.bsd.lv/history.html⟩ of the
439 mdocml project provides an overview of roff development up to
440 date, with links to original documentation and comments of the
441 original authors.
442
443 troff.org
444 The historical troff site ⟨http://www.troff.org⟩ provides an
445 overview and pointers to the historical aspects of roff.
446
447 Multics
448 The Multics site ⟨http://www.multicians.org⟩ contains a lot of
449 information on the MIT projects, CTSS, Multics, early Unix, in‐
450 cluding runoff; especially useful are a glossary and the many
451 links to ancient documents.
452
453 Unix Archive
454 The Ancient Unixes Archive ⟨http://www.tuhs.org/Archive/⟩ pro‐
455 vides the source code and some binaries of the ancient Unixes
456 (including the source code of troff and its documentation) that
457 were made public by Caldera since 2001, e.g. of the famous Unix
458 version 7 for PDP-11 at the Unix V7 site ⟨http://www.tuhs.org/
459 Archive/PDP-11/Trees/V7⟩.
460
461 Developers at AT&T Bell Labs
462 Bell Labs Computing and Mathematical Sciences Research ⟨http://
463 www.bell-labs.com/⟩ provides a search facility for tracking in‐
464 formation on the early developers.
465
466 Plan 9 The Plan 9 operating system ⟨http://plan9.bell-labs.com⟩ by AT&T
467 Bell Labs.
468
469 runoff Jerry Saltzer's home page ⟨http://web.mit.edu/Saltzer/www/
470 publications/pubs.html⟩ stores some documents using the ancient
471 RUNOFF formatting language.
472
473 CSTR Papers
474 The Bell Labs (now Alcatel) CSTR site ⟨https://
475 www.alcatel-lucent.com/bell-labs-journals⟩ stores the original
476 troff manuals (CSTR #54, #97, #114, #116, #122) and famous his‐
477 torical documents on programming.
478
479 GNU roff
480 The groff web site ⟨http://www.gnu.org/software/groff⟩ provides
481 the free roff implementation groff, the actual standard roff.
482
483 Historical roff Documentation
484 Many classical troff documents are still available on-line. The two
485 main manuals of the troff language are
486
487 [CSTR #54]
488 J. F. Ossanna, Nroff/Troff User's Manual ⟨http://www.troff.org/
489 54.pdf⟩; Bell Labs, 1976; revised by Brian Kernighan, 1992.
490
491 [CSTR #97]
492 Brian Kernighan, A Typesetter-independent TROFF ⟨http://
493 cm.bell-labs.com/cm/cs/cstr/97.ps.gz⟩, Bell Labs, 1981, revised
494 March 1982.
495
496 The “little language” roff papers are
497
498 [CSTR #114]
499 Jon L. Bentley and Brian W. Kernighan, GRAP – A Language for
500 Typesetting Graphs ⟨http://cm.bell-labs.com/cm/cs/cstr/
501 114.ps.gz⟩; Bell Labs, August 1984.
502
503 [CSTR #116]
504 Brian W. Kernighan, PIC – A Graphics Language for Typesetting
505 ⟨http://cm.bell-labs.com/cm/cs/cstr/116.ps.gz⟩; Bell Labs, De‐
506 cember 1984.
507
508 [CSTR #122]
509 J. L. Bentley, L. W. Jelinski, and B. W. Kernighan, CHEM – A
510 Program for Typesetting Chemical Structure Diagrams, Computers
511 and Chemistry ⟨http://cm.bell-labs.com/cm/cs/cstr/122.ps.gz⟩;
512 Bell Labs, April 1986.
513
514 You can get an archive with most classical roff documentation as rea‐
515 sonable PDF files at github using the shell command
516 $ git clone https://github.com/bwarken/roff_classical.git
517
518 Manual Pages
519 Due to its complex structure, a full roff system has many man pages,
520 each describing a single aspect of roff. Unfortunately, there is no
521 general naming scheme for the documentation among the different roff
522 implementations.
523
524 In groff, the man page groff(1) contains a survey of all documentation
525 available in groff.
526
527 On other systems, you are on your own, but troff(1) might be a good
528 starting point.
529
530
531
532groff 1.22.4 20 January 2022 ROFF(7)