428
+ − 1 \input texinfo @c -*-texinfo-*-
+ − 2 @c %**start of header
+ − 3 @setfilename ../info/standards.info
+ − 4 @settitle GNU Coding Standards
2665
+ − 5 @c In GNU Emacs, this date is automagically updated when this file is
+ − 6 @c saved, but in XEmacs you must set it when synching:
462
+ − 7 @set lastupdate February 21, 2001
428
+ − 8 @c %**end of header
+ − 9
+ − 10 @ifinfo
+ − 11 @format
+ − 12 START-INFO-DIR-ENTRY
+ − 13 * Standards: (standards). GNU coding standards.
+ − 14 END-INFO-DIR-ENTRY
+ − 15 @end format
+ − 16 @end ifinfo
+ − 17
+ − 18 @c @setchapternewpage odd
+ − 19 @setchapternewpage off
+ − 20
462
+ − 21 @c Put everything in one index (arbitrarily chosen to be the concept index).
+ − 22 @syncodeindex fn cp
+ − 23 @syncodeindex ky cp
+ − 24 @syncodeindex pg cp
+ − 25 @syncodeindex vr cp
+ − 26
428
+ − 27 @c This is used by a cross ref in make-stds.texi
+ − 28 @set CODESTD 1
+ − 29 @iftex
+ − 30 @set CHAPTER chapter
+ − 31 @end iftex
+ − 32 @ifinfo
+ − 33 @set CHAPTER node
+ − 34 @end ifinfo
+ − 35
+ − 36 @ifinfo
+ − 37 GNU Coding Standards
462
+ − 38 Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
428
+ − 39
+ − 40 Permission is granted to make and distribute verbatim copies of
+ − 41 this manual provided the copyright notice and this permission notice
+ − 42 are preserved on all copies.
+ − 43
+ − 44 @ignore
+ − 45 Permission is granted to process this file through TeX and print the
+ − 46 results, provided the printed document carries copying permission
+ − 47 notice identical to this one except for the removal of this paragraph
+ − 48 (this paragraph not being relevant to the printed manual).
+ − 49 @end ignore
+ − 50
+ − 51 Permission is granted to copy and distribute modified versions of this
+ − 52 manual under the conditions for verbatim copying, provided that the entire
+ − 53 resulting derived work is distributed under the terms of a permission
+ − 54 notice identical to this one.
+ − 55
+ − 56 Permission is granted to copy and distribute translations of this manual
+ − 57 into another language, under the above conditions for modified versions,
+ − 58 except that this permission notice may be stated in a translation approved
+ − 59 by the Free Software Foundation.
+ − 60 @end ifinfo
+ − 61
+ − 62 @titlepage
+ − 63 @title GNU Coding Standards
+ − 64 @author Richard Stallman
+ − 65 @author last updated @value{lastupdate}
+ − 66 @page
+ − 67
+ − 68 @vskip 0pt plus 1filll
462
+ − 69 Copyright @copyright{} 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
428
+ − 70
+ − 71 Permission is granted to make and distribute verbatim copies of
+ − 72 this manual provided the copyright notice and this permission notice
+ − 73 are preserved on all copies.
+ − 74
+ − 75 Permission is granted to copy and distribute modified versions of this
+ − 76 manual under the conditions for verbatim copying, provided that the entire
+ − 77 resulting derived work is distributed under the terms of a permission
+ − 78 notice identical to this one.
+ − 79
+ − 80 Permission is granted to copy and distribute translations of this manual
+ − 81 into another language, under the above conditions for modified versions,
+ − 82 except that this permission notice may be stated in a translation approved
+ − 83 by the Free Software Foundation.
+ − 84 @end titlepage
+ − 85
+ − 86 @ifinfo
+ − 87 @node Top, Preface, (dir), (dir)
+ − 88 @top Version
+ − 89
+ − 90 Last updated @value{lastupdate}.
+ − 91 @end ifinfo
+ − 92
+ − 93 @menu
+ − 94 * Preface:: About the GNU Coding Standards
+ − 95 * Legal Issues:: Keeping Free Software Free
+ − 96 * Design Advice:: General Program Design
+ − 97 * Program Behavior:: Program Behavior for All Programs
+ − 98 * Writing C:: Making The Best Use of C
+ − 99 * Documentation:: Documenting Programs
+ − 100 * Managing Releases:: The Release Process
+ − 101 * References:: References to Non-Free Software or Documentation
462
+ − 102 * Index::
428
+ − 103 @end menu
+ − 104
+ − 105 @node Preface
+ − 106 @chapter About the GNU Coding Standards
+ − 107
+ − 108 The GNU Coding Standards were written by Richard Stallman and other GNU
+ − 109 Project volunteers. Their purpose is to make the GNU system clean,
+ − 110 consistent, and easy to install. This document can also be read as a
+ − 111 guide to writing portable, robust and reliable programs. It focuses on
+ − 112 programs written in C, but many of the rules and principles are useful
+ − 113 even if you write in another programming language. The rules often
+ − 114 state reasons for writing in a certain way.
+ − 115
+ − 116 Corrections or suggestions for this document should be sent to
462
+ − 117 @email{bug-standards@@gnu.org}. If you make a suggestion, please include a
428
+ − 118 suggested new wording for it; our time is limited. We prefer a context
+ − 119 diff to the @file{standards.texi} or @file{make-stds.texi} files, but if
+ − 120 you don't have those files, please mail your suggestion anyway.
+ − 121
+ − 122 This release of the GNU Coding Standards was last updated
+ − 123 @value{lastupdate}.
+ − 124
462
+ − 125 @cindex where to obtain @code{standards.texi}
+ − 126 @cindex downloading this manual
+ − 127 If you did not obtain this file directly from the GNU project and
+ − 128 recently, please check for a newer version. You can ftp the GNU Coding
+ − 129 Standards from any GNU FTP host in the directory
+ − 130 @file{/pub/gnu/standards/}. The GNU Coding Standards are available
+ − 131 there in several different formats: @file{standards.text},
+ − 132 @file{standards.texi}, @file{standards.info}, and @file{standards.dvi}.
+ − 133 The GNU Coding Standards are also available on the GNU World Wide Web
+ − 134 server: @uref{http://www.gnu.org/prep/standards_toc.html}.
+ − 135
428
+ − 136 @node Legal Issues
+ − 137 @chapter Keeping Free Software Free
462
+ − 138 @cindex legal aspects
428
+ − 139
+ − 140 This @value{CHAPTER} discusses how you can make sure that GNU software
462
+ − 141 avoids legal difficulties, and other related issues.
428
+ − 142
+ − 143 @menu
+ − 144 * Reading Non-Free Code:: Referring to Proprietary Programs
+ − 145 * Contributions:: Accepting Contributions
462
+ − 146 * Trademarks:: How We Deal with Trademark Issues
428
+ − 147 @end menu
+ − 148
+ − 149 @node Reading Non-Free Code
+ − 150 @section Referring to Proprietary Programs
462
+ − 151 @cindex proprietary programs
+ − 152 @cindex avoiding proprietary code
428
+ − 153
+ − 154 Don't in any circumstances refer to Unix source code for or during
+ − 155 your work on GNU! (Or to any other proprietary programs.)
+ − 156
+ − 157 If you have a vague recollection of the internals of a Unix program,
+ − 158 this does not absolutely mean you can't write an imitation of it, but
+ − 159 do try to organize the imitation internally along different lines,
+ − 160 because this is likely to make the details of the Unix version
+ − 161 irrelevant and dissimilar to your results.
+ − 162
+ − 163 For example, Unix utilities were generally optimized to minimize
+ − 164 memory use; if you go for speed instead, your program will be very
+ − 165 different. You could keep the entire input file in core and scan it
+ − 166 there instead of using stdio. Use a smarter algorithm discovered more
+ − 167 recently than the Unix program. Eliminate use of temporary files. Do
+ − 168 it in one pass instead of two (we did this in the assembler).
+ − 169
+ − 170 Or, on the contrary, emphasize simplicity instead of speed. For some
+ − 171 applications, the speed of today's computers makes simpler algorithms
+ − 172 adequate.
+ − 173
+ − 174 Or go for generality. For example, Unix programs often have static
+ − 175 tables or fixed-size strings, which make for arbitrary limits; use
+ − 176 dynamic allocation instead. Make sure your program handles NULs and
+ − 177 other funny characters in the input files. Add a programming language
+ − 178 for extensibility and write part of the program in that language.
+ − 179
+ − 180 Or turn some parts of the program into independently usable libraries.
+ − 181 Or use a simple garbage collector instead of tracking precisely when
+ − 182 to free memory, or use a new GNU facility such as obstacks.
+ − 183
+ − 184 @node Contributions
+ − 185 @section Accepting Contributions
462
+ − 186 @cindex legal papers
+ − 187 @cindex accepting contributions
428
+ − 188
+ − 189 If the program you are working on is copyrighted by the Free Software
+ − 190 Foundation, then when someone else sends you a piece of code to add to
+ − 191 the program, we need legal papers to use it---just as we asked you to
+ − 192 sign papers initially. @emph{Each} person who makes a nontrivial
+ − 193 contribution to a program must sign some sort of legal papers in order
+ − 194 for us to have clear title to the program; the main author alone is not
+ − 195 enough.
+ − 196
+ − 197 So, before adding in any contributions from other people, please tell
+ − 198 us, so we can arrange to get the papers. Then wait until we tell you
+ − 199 that we have received the signed papers, before you actually use the
+ − 200 contribution.
+ − 201
+ − 202 This applies both before you release the program and afterward. If
+ − 203 you receive diffs to fix a bug, and they make significant changes, we
+ − 204 need legal papers for that change.
+ − 205
+ − 206 This also applies to comments and documentation files. For copyright
+ − 207 law, comments and code are just text. Copyright applies to all kinds of
+ − 208 text, so we need legal papers for all kinds.
+ − 209
+ − 210 We know it is frustrating to ask for legal papers; it's frustrating for
+ − 211 us as well. But if you don't wait, you are going out on a limb---for
+ − 212 example, what if the contributor's employer won't sign a disclaimer?
+ − 213 You might have to take that code out again!
+ − 214
+ − 215 You don't need papers for changes of a few lines here or there, since
+ − 216 they are not significant for copyright purposes. Also, you don't need
+ − 217 papers if all you get from the suggestion is some ideas, not actual code
+ − 218 which you use. For example, if someone send you one implementation, but
+ − 219 you write a different implementation of the same idea, you don't need to
+ − 220 get papers.
+ − 221
+ − 222 The very worst thing is if you forget to tell us about the other
+ − 223 contributor. We could be very embarrassed in court some day as a
+ − 224 result.
+ − 225
+ − 226 We have more detailed advice for maintainers of programs; if you have
+ − 227 reached the stage of actually maintaining a program for GNU (whether
+ − 228 released or not), please ask us for a copy.
+ − 229
462
+ − 230 @node Trademarks
+ − 231 @section Trademarks
+ − 232 @cindex trademarks
+ − 233
+ − 234 Please do not include any trademark acknowledgements in GNU software
+ − 235 packages or documentation.
+ − 236
+ − 237 Trademark acknowledgements are the statements that such-and-such is a
+ − 238 trademark of so-and-so. The GNU Project has no objection to the basic
+ − 239 idea of trademarks, but these acknowledgements feel like kowtowing, so
+ − 240 we don't use them. There is no legal requirement for them.
+ − 241
+ − 242 What is legally required, as regards other people's trademarks, is to
+ − 243 avoid using them in ways which a reader might read as naming or labeling
+ − 244 our own programs or activities. For example, since ``Objective C'' is
+ − 245 (or at least was) a trademark, we made sure to say that we provide a
+ − 246 ``compiler for the Objective C language'' rather than an ``Objective C
+ − 247 compiler''. The latter is meant to be short for the former, but it does
+ − 248 not explicitly state the relationship, so it could be misinterpreted as
+ − 249 using ``Objective C'' as a label for the compiler rather than for the
+ − 250 language.
+ − 251
428
+ − 252 @node Design Advice
+ − 253 @chapter General Program Design
462
+ − 254 @cindex program design
428
+ − 255
+ − 256 This @value{CHAPTER} discusses some of the issues you should take into
+ − 257 account when designing your program.
+ − 258
462
+ − 259 @c Standard or ANSI C
+ − 260 @c
+ − 261 @c In 1989 the American National Standards Institute (ANSI) standardized
+ − 262 @c C as standard X3.159-1989. In December of that year the
+ − 263 @c International Standards Organization ISO adopted the ANSI C standard
+ − 264 @c making minor changes. In 1990 ANSI then re-adopted ISO standard
+ − 265 @c C. This version of C is known as either ANSI C or Standard C.
+ − 266
+ − 267 @c A major revision of the C Standard appeared in 1999.
+ − 268
428
+ − 269 @menu
462
+ − 270 * Source Language:: Which languges to use.
428
+ − 271 * Compatibility:: Compatibility with other implementations
+ − 272 * Using Extensions:: Using non-standard features
462
+ − 273 * Standard C:: Using Standard C features
428
+ − 274 @end menu
+ − 275
462
+ − 276 @node Source Language
+ − 277 @section Which Languages to Use
+ − 278 @cindex programming languges
+ − 279
+ − 280 When you want to use a language that gets compiled and runs at high
+ − 281 speed, the best language to use is C. Using another language is like
+ − 282 using a non-standard feature: it will cause trouble for users. Even if
+ − 283 GCC supports the other language, users may find it inconvenient to have
+ − 284 to install the compiler for that other language in order to build your
+ − 285 program. For example, if you write your program in C++, people will
+ − 286 have to install the GNU C++ compiler in order to compile your program.
+ − 287
+ − 288 C has one other advantage over C++ and other compiled languages: more
+ − 289 people know C, so more people will find it easy to read and modify the
+ − 290 program if it is written in C.
+ − 291
+ − 292 So in general it is much better to use C, rather than the
+ − 293 comparable alternatives.
+ − 294
+ − 295 But there are two exceptions to that conclusion:
+ − 296
+ − 297 @itemize @bullet
+ − 298 @item
+ − 299 It is no problem to use another language to write a tool specifically
+ − 300 intended for use with that language. That is because the only people
+ − 301 who want to build the tool will be those who have installed the other
+ − 302 language anyway.
+ − 303
+ − 304 @item
+ − 305 If an application is of interest only to a narrow part of the community,
+ − 306 then the question of which language it is written in has less effect on
+ − 307 other people, so you may as well please yourself.
+ − 308 @end itemize
+ − 309
+ − 310 Many programs are designed to be extensible: they include an interpreter
+ − 311 for a language that is higher level than C. Often much of the program
+ − 312 is written in that language, too. The Emacs editor pioneered this
+ − 313 technique.
+ − 314
+ − 315 @cindex GUILE
+ − 316 The standard extensibility interpreter for GNU software is GUILE, which
+ − 317 implements the language Scheme (an especially clean and simple dialect
+ − 318 of Lisp). @uref{http://www.gnu.org/software/guile/}. We don't reject
+ − 319 programs written in other ``scripting languages'' such as Perl and
+ − 320 Python, but using GUILE is very important for the overall consistency of
+ − 321 the GNU system.
+ − 322
428
+ − 323 @node Compatibility
+ − 324 @section Compatibility with Other Implementations
462
+ − 325 @cindex compatibility with C and @sc{posix} standards
+ − 326 @cindex @sc{posix} compatibility
428
+ − 327
+ − 328 With occasional exceptions, utility programs and libraries for GNU
+ − 329 should be upward compatible with those in Berkeley Unix, and upward
462
+ − 330 compatible with Standard C if Standard C specifies their
+ − 331 behavior, and upward compatible with @sc{posix} if @sc{posix} specifies
+ − 332 their behavior.
428
+ − 333
+ − 334 When these standards conflict, it is useful to offer compatibility
+ − 335 modes for each of them.
+ − 336
462
+ − 337 @cindex options for compatibility
+ − 338 Standard C and @sc{posix} prohibit many kinds of extensions. Feel
+ − 339 free to make the extensions anyway, and include a @samp{--ansi},
428
+ − 340 @samp{--posix}, or @samp{--compatible} option to turn them off.
+ − 341 However, if the extension has a significant chance of breaking any real
462
+ − 342 programs or scripts, then it is not really upward compatible. So you
+ − 343 should try to redesign its interface to make it upward compatible.
+ − 344
+ − 345 @cindex @code{POSIXLY_CORRECT}, environment variable
428
+ − 346 Many GNU programs suppress extensions that conflict with @sc{posix} if the
+ − 347 environment variable @code{POSIXLY_CORRECT} is defined (even if it is
+ − 348 defined with a null value). Please make your program recognize this
+ − 349 variable if appropriate.
+ − 350
+ − 351 When a feature is used only by users (not by programs or command
+ − 352 files), and it is done poorly in Unix, feel free to replace it
+ − 353 completely with something totally different and better. (For example,
+ − 354 @code{vi} is replaced with Emacs.) But it is nice to offer a compatible
+ − 355 feature as well. (There is a free @code{vi} clone, so we offer it.)
+ − 356
462
+ − 357 Additional useful features are welcome regardless of whether
+ − 358 there is any precedent for them.
428
+ − 359
+ − 360 @node Using Extensions
+ − 361 @section Using Non-standard Features
462
+ − 362 @cindex non-standard extensions
428
+ − 363
+ − 364 Many GNU facilities that already exist support a number of convenient
+ − 365 extensions over the comparable Unix facilities. Whether to use these
+ − 366 extensions in implementing your program is a difficult question.
+ − 367
+ − 368 On the one hand, using the extensions can make a cleaner program.
+ − 369 On the other hand, people will not be able to build the program
+ − 370 unless the other GNU tools are available. This might cause the
+ − 371 program to work on fewer kinds of machines.
+ − 372
+ − 373 With some extensions, it might be easy to provide both alternatives.
+ − 374 For example, you can define functions with a ``keyword'' @code{INLINE}
+ − 375 and define that as a macro to expand into either @code{inline} or
+ − 376 nothing, depending on the compiler.
+ − 377
+ − 378 In general, perhaps it is best not to use the extensions if you can
+ − 379 straightforwardly do without them, but to use the extensions if they
+ − 380 are a big improvement.
+ − 381
+ − 382 An exception to this rule are the large, established programs (such as
462
+ − 383 Emacs) which run on a great variety of systems. Using GNU extensions in
+ − 384 such programs would make many users unhappy, so we don't do that.
+ − 385
+ − 386 Another exception is for programs that are used as part of compilation:
+ − 387 anything that must be compiled with other compilers in order to
+ − 388 bootstrap the GNU compilation facilities. If these require the GNU
+ − 389 compiler, then no one can compile them without having them installed
+ − 390 already. That would be extremely troublesome in certain cases.
+ − 391
+ − 392 @node Standard C
+ − 393 @section Standard C and Pre-Standard C
+ − 394 @cindex @sc{ansi} C standard
+ − 395
+ − 396 1989 Standard C is widespread enough now that it is ok to use its
+ − 397 features in new programs. There is one exception: do not ever use the
+ − 398 ``trigraph'' feature of Standard C.
+ − 399
+ − 400 1999 Standard C is not widespread yet, so please do not require its
+ − 401 features in programs. It is ok to use its features if they are present.
+ − 402
+ − 403 However, it is easy to support pre-standard compilers in most programs,
+ − 404 so if you know how to do that, feel free. If a program you are
+ − 405 maintaining has such support, you should try to keep it working.
+ − 406
+ − 407 @cindex function prototypes
+ − 408 To support pre-standard C, instead of writing function definitions in
+ − 409 standard prototype form,
428
+ − 410
+ − 411 @example
+ − 412 int
+ − 413 foo (int x, int y)
+ − 414 @dots{}
+ − 415 @end example
+ − 416
+ − 417 @noindent
462
+ − 418 write the definition in pre-standard style like this,
428
+ − 419
+ − 420 @example
+ − 421 int
+ − 422 foo (x, y)
+ − 423 int x, y;
+ − 424 @dots{}
+ − 425 @end example
+ − 426
+ − 427 @noindent
+ − 428 and use a separate declaration to specify the argument prototype:
+ − 429
+ − 430 @example
+ − 431 int foo (int, int);
+ − 432 @end example
+ − 433
+ − 434 You need such a declaration anyway, in a header file, to get the benefit
462
+ − 435 of prototypes in all the files where the function is called. And once
+ − 436 you have the declaration, you normally lose nothing by writing the
+ − 437 function definition in the pre-standard style.
428
+ − 438
+ − 439 This technique does not work for integer types narrower than @code{int}.
+ − 440 If you think of an argument as being of a type narrower than @code{int},
+ − 441 declare it as @code{int} instead.
+ − 442
+ − 443 There are a few special cases where this technique is hard to use. For
+ − 444 example, if a function argument needs to hold the system type
+ − 445 @code{dev_t}, you run into trouble, because @code{dev_t} is shorter than
+ − 446 @code{int} on some machines; but you cannot use @code{int} instead,
+ − 447 because @code{dev_t} is wider than @code{int} on some machines. There
462
+ − 448 is no type you can safely use on all machines in a non-standard
+ − 449 definition. The only way to support non-standard C and pass such an
428
+ − 450 argument is to check the width of @code{dev_t} using Autoconf and choose
+ − 451 the argument type accordingly. This may not be worth the trouble.
+ − 452
462
+ − 453 In order to support pre-standard compilers that do not recognize
+ − 454 prototypes, you may want to use a preprocessor macro like this:
+ − 455
+ − 456 @example
+ − 457 /* Declare the prototype for a general external function. */
+ − 458 #if defined (__STDC__) || defined (WINDOWSNT)
+ − 459 #define P_(proto) proto
+ − 460 #else
+ − 461 #define P_(proto) ()
+ − 462 #endif
+ − 463 @end example
428
+ − 464
+ − 465 @node Program Behavior
+ − 466 @chapter Program Behavior for All Programs
+ − 467
462
+ − 468 This @value{CHAPTER} describes conventions for writing robust
+ − 469 software. It also describes general standards for error messages, the
+ − 470 command line interface, and how libraries should behave.
428
+ − 471
+ − 472 @menu
+ − 473 * Semantics:: Writing robust programs
+ − 474 * Libraries:: Library behavior
+ − 475 * Errors:: Formatting error messages
462
+ − 476 * User Interfaces:: Standards about interfaces generally
+ − 477 * Graphical Interfaces:: Standards for graphical interfaces
+ − 478 * Command-Line Interfaces:: Standards for command line interfaces
+ − 479 * Option Table:: Table of long options
428
+ − 480 * Memory Usage:: When and how to care about memory needs
462
+ − 481 * File Usage:: Which files to use, and where
428
+ − 482 @end menu
+ − 483
+ − 484 @node Semantics
+ − 485 @section Writing Robust Programs
+ − 486
462
+ − 487 @cindex arbitrary limits on data
428
+ − 488 Avoid arbitrary limits on the length or number of @emph{any} data
+ − 489 structure, including file names, lines, files, and symbols, by allocating
+ − 490 all data structures dynamically. In most Unix utilities, ``long lines
+ − 491 are silently truncated''. This is not acceptable in a GNU utility.
+ − 492
462
+ − 493 @cindex @code{NUL} characters
428
+ − 494 Utilities reading files should not drop NUL characters, or any other
+ − 495 nonprinting characters @emph{including those with codes above 0177}.
+ − 496 The only sensible exceptions would be utilities specifically intended
+ − 497 for interface to certain types of terminals or printers
+ − 498 that can't handle those characters.
+ − 499 Whenever possible, try to make programs work properly with
+ − 500 sequences of bytes that represent multibyte characters, using encodings
+ − 501 such as UTF-8 and others.
+ − 502
462
+ − 503 @cindex error messages
428
+ − 504 Check every system call for an error return, unless you know you wish to
+ − 505 ignore errors. Include the system error text (from @code{perror} or
+ − 506 equivalent) in @emph{every} error message resulting from a failing
+ − 507 system call, as well as the name of the file if any and the name of the
+ − 508 utility. Just ``cannot open foo.c'' or ``stat failed'' is not
+ − 509 sufficient.
+ − 510
462
+ − 511 @cindex @code{malloc} return value
+ − 512 @cindex memory allocation failure
428
+ − 513 Check every call to @code{malloc} or @code{realloc} to see if it
+ − 514 returned zero. Check @code{realloc} even if you are making the block
+ − 515 smaller; in a system that rounds block sizes to a power of 2,
+ − 516 @code{realloc} may get a different block if you ask for less space.
+ − 517
+ − 518 In Unix, @code{realloc} can destroy the storage block if it returns
+ − 519 zero. GNU @code{realloc} does not have this bug: if it fails, the
+ − 520 original block is unchanged. Feel free to assume the bug is fixed. If
+ − 521 you wish to run your program on Unix, and wish to avoid lossage in this
+ − 522 case, you can use the GNU @code{malloc}.
+ − 523
+ − 524 You must expect @code{free} to alter the contents of the block that was
+ − 525 freed. Anything you want to fetch from the block, you must fetch before
+ − 526 calling @code{free}.
+ − 527
+ − 528 If @code{malloc} fails in a noninteractive program, make that a fatal
+ − 529 error. In an interactive program (one that reads commands from the
+ − 530 user), it is better to abort the command and return to the command
+ − 531 reader loop. This allows the user to kill other processes to free up
+ − 532 virtual memory, and then try the command again.
+ − 533
462
+ − 534 @cindex command-line arguments, decoding
428
+ − 535 Use @code{getopt_long} to decode arguments, unless the argument syntax
+ − 536 makes this unreasonable.
+ − 537
+ − 538 When static storage is to be written in during program execution, use
+ − 539 explicit C code to initialize it. Reserve C initialized declarations
+ − 540 for data that will not be changed.
+ − 541 @c ADR: why?
+ − 542
+ − 543 Try to avoid low-level interfaces to obscure Unix data structures (such
+ − 544 as file directories, utmp, or the layout of kernel memory), since these
+ − 545 are less likely to work compatibly. If you need to find all the files
+ − 546 in a directory, use @code{readdir} or some other high-level interface.
+ − 547 These are supported compatibly by GNU.
+ − 548
462
+ − 549 @cindex signal handling
428
+ − 550 The preferred signal handling facilities are the BSD variant of
+ − 551 @code{signal}, and the @sc{posix} @code{sigaction} function; the
+ − 552 alternative USG @code{signal} interface is an inferior design.
+ − 553
+ − 554 Nowadays, using the @sc{posix} signal functions may be the easiest way
+ − 555 to make a program portable. If you use @code{signal}, then on GNU/Linux
+ − 556 systems running GNU libc version 1, you should include
+ − 557 @file{bsd/signal.h} instead of @file{signal.h}, so as to get BSD
+ − 558 behavior. It is up to you whether to support systems where
+ − 559 @code{signal} has only the USG behavior, or give up on them.
+ − 560
462
+ − 561 @cindex impossible conditions
428
+ − 562 In error checks that detect ``impossible'' conditions, just abort.
+ − 563 There is usually no point in printing any message. These checks
+ − 564 indicate the existence of bugs. Whoever wants to fix the bugs will have
+ − 565 to read the source code and run a debugger. So explain the problem with
+ − 566 comments in the source. The relevant data will be in variables, which
+ − 567 are easy to examine with the debugger, so there is no point moving them
+ − 568 elsewhere.
+ − 569
+ − 570 Do not use a count of errors as the exit status for a program.
+ − 571 @emph{That does not work}, because exit status values are limited to 8
+ − 572 bits (0 through 255). A single run of the program might have 256
+ − 573 errors; if you try to return 256 as the exit status, the parent process
+ − 574 will see 0 as the status, and it will appear that the program succeeded.
+ − 575
462
+ − 576 @cindex temporary files
+ − 577 @cindex @code{TMPDIR} environment variable
428
+ − 578 If you make temporary files, check the @code{TMPDIR} environment
+ − 579 variable; if that variable is defined, use the specified directory
+ − 580 instead of @file{/tmp}.
+ − 581
462
+ − 582 In addition, be aware that there is a possible security problem when
+ − 583 creating temporary files in world-writable directories. In C, you can
+ − 584 avoid this problem by creating temporary files in this manner:
+ − 585
+ − 586 @example
+ − 587 fd = open(filename, O_WRONLY | O_CREAT | O_EXCL, 0600);
+ − 588 @end example
+ − 589
+ − 590 @noindent
+ − 591 or by using the @code{mkstemps} function from libiberty.
+ − 592
+ − 593 In bash, use @code{set -C} to avoid this problem.
+ − 594
428
+ − 595 @node Libraries
+ − 596 @section Library Behavior
462
+ − 597 @cindex libraries
428
+ − 598
+ − 599 Try to make library functions reentrant. If they need to do dynamic
+ − 600 storage allocation, at least try to avoid any nonreentrancy aside from
+ − 601 that of @code{malloc} itself.
+ − 602
+ − 603 Here are certain name conventions for libraries, to avoid name
+ − 604 conflicts.
+ − 605
+ − 606 Choose a name prefix for the library, more than two characters long.
+ − 607 All external function and variable names should start with this
+ − 608 prefix. In addition, there should only be one of these in any given
+ − 609 library member. This usually means putting each one in a separate
+ − 610 source file.
+ − 611
+ − 612 An exception can be made when two external symbols are always used
+ − 613 together, so that no reasonable program could use one without the
+ − 614 other; then they can both go in the same file.
+ − 615
+ − 616 External symbols that are not documented entry points for the user
462
+ − 617 should have names beginning with @samp{_}. The @samp{_} should be
+ − 618 followed by the chosen name prefix for the library, to prevent
+ − 619 collisions with other libraries. These can go in the same files with
+ − 620 user entry points if you like.
428
+ − 621
+ − 622 Static functions and variables can be used as you like and need not
+ − 623 fit any naming convention.
+ − 624
+ − 625 @node Errors
+ − 626 @section Formatting Error Messages
462
+ − 627 @cindex formatting error messages
+ − 628 @cindex error messages, formatting
428
+ − 629
+ − 630 Error messages from compilers should look like this:
+ − 631
+ − 632 @example
+ − 633 @var{source-file-name}:@var{lineno}: @var{message}
+ − 634 @end example
+ − 635
+ − 636 @noindent
+ − 637 If you want to mention the column number, use this format:
+ − 638
+ − 639 @example
+ − 640 @var{source-file-name}:@var{lineno}:@var{column}: @var{message}
+ − 641 @end example
+ − 642
+ − 643 @noindent
+ − 644 Line numbers should start from 1 at the beginning of the file, and
+ − 645 column numbers should start from 1 at the beginning of the line. (Both
+ − 646 of these conventions are chosen for compatibility.) Calculate column
+ − 647 numbers assuming that space and all ASCII printing characters have
+ − 648 equal width, and assuming tab stops every 8 columns.
+ − 649
+ − 650 Error messages from other noninteractive programs should look like this:
+ − 651
+ − 652 @example
+ − 653 @var{program}:@var{source-file-name}:@var{lineno}: @var{message}
+ − 654 @end example
+ − 655
+ − 656 @noindent
+ − 657 when there is an appropriate source file, or like this:
+ − 658
+ − 659 @example
+ − 660 @var{program}: @var{message}
+ − 661 @end example
+ − 662
+ − 663 @noindent
+ − 664 when there is no relevant source file.
+ − 665
+ − 666 If you want to mention the column number, use this format:
+ − 667
+ − 668 @example
+ − 669 @var{program}:@var{source-file-name}:@var{lineno}:@var{column}: @var{message}
+ − 670 @end example
+ − 671
+ − 672 In an interactive program (one that is reading commands from a
+ − 673 terminal), it is better not to include the program name in an error
+ − 674 message. The place to indicate which program is running is in the
+ − 675 prompt or with the screen layout. (When the same program runs with
+ − 676 input from a source other than a terminal, it is not interactive and
+ − 677 would do best to print error messages using the noninteractive style.)
+ − 678
+ − 679 The string @var{message} should not begin with a capital letter when
+ − 680 it follows a program name and/or file name. Also, it should not end
+ − 681 with a period.
+ − 682
+ − 683 Error messages from interactive programs, and other messages such as
+ − 684 usage messages, should start with a capital letter. But they should not
+ − 685 end with a period.
+ − 686
+ − 687 @node User Interfaces
462
+ − 688 @section Standards for Interfaces Generally
+ − 689
+ − 690 @cindex program name and its behavior
+ − 691 @cindex behavior, dependent on program's name
428
+ − 692 Please don't make the behavior of a utility depend on the name used
+ − 693 to invoke it. It is useful sometimes to make a link to a utility
+ − 694 with a different name, and that should not change what it does.
+ − 695
+ − 696 Instead, use a run time option or a compilation switch or both
+ − 697 to select among the alternate behaviors.
+ − 698
462
+ − 699 @cindex output device and program's behavior
428
+ − 700 Likewise, please don't make the behavior of the program depend on the
+ − 701 type of output device it is used with. Device independence is an
+ − 702 important principle of the system's design; do not compromise it merely
+ − 703 to save someone from typing an option now and then. (Variation in error
+ − 704 message syntax when using a terminal is ok, because that is a side issue
+ − 705 that people do not depend on.)
+ − 706
+ − 707 If you think one behavior is most useful when the output is to a
+ − 708 terminal, and another is most useful when the output is a file or a
+ − 709 pipe, then it is usually best to make the default behavior the one that
+ − 710 is useful with output to a terminal, and have an option for the other
+ − 711 behavior.
+ − 712
+ − 713 Compatibility requires certain programs to depend on the type of output
+ − 714 device. It would be disastrous if @code{ls} or @code{sh} did not do so
+ − 715 in the way all users expect. In some of these cases, we supplement the
+ − 716 program with a preferred alternate version that does not depend on the
+ − 717 output device type. For example, we provide a @code{dir} program much
+ − 718 like @code{ls} except that its default output format is always
+ − 719 multi-column format.
+ − 720
462
+ − 721 @node Graphical Interfaces
+ − 722 @section Standards for Graphical Interfaces
+ − 723 @cindex graphical user interface
+ − 724
+ − 725 @cindex gtk
+ − 726 When you write a program that provides a graphical user interface,
+ − 727 please make it work with X Windows and the GTK toolkit unless the
+ − 728 functionality specifically requires some alternative (for example,
+ − 729 ``displaying jpeg images while in console mode'').
+ − 730
+ − 731 In addition, please provide a command-line interface to control the
+ − 732 functionality. (In many cases, the graphical user interface can be a
+ − 733 separate program which invokes the command-line program.) This is
+ − 734 so that the same jobs can be done from scripts.
+ − 735
+ − 736 @cindex corba
+ − 737 @cindex gnome
+ − 738 Please also consider providing a CORBA interface (for use from GNOME), a
+ − 739 library interface (for use from C), and perhaps a keyboard-driven
+ − 740 console interface (for use by users from console mode). Once you are
+ − 741 doing the work to provide the functionality and the graphical interface,
+ − 742 these won't be much extra work.
+ − 743
+ − 744 @node Command-Line Interfaces
+ − 745 @section Standards for Command Line Interfaces
+ − 746 @cindex command-line interface
+ − 747
+ − 748 @findex getopt
428
+ − 749 It is a good idea to follow the @sc{posix} guidelines for the
+ − 750 command-line options of a program. The easiest way to do this is to use
+ − 751 @code{getopt} to parse them. Note that the GNU version of @code{getopt}
+ − 752 will normally permit options anywhere among the arguments unless the
+ − 753 special argument @samp{--} is used. This is not what @sc{posix}
+ − 754 specifies; it is a GNU extension.
+ − 755
462
+ − 756 @cindex long-named options
428
+ − 757 Please define long-named options that are equivalent to the
+ − 758 single-letter Unix-style options. We hope to make GNU more user
+ − 759 friendly this way. This is easy to do with the GNU function
+ − 760 @code{getopt_long}.
+ − 761
+ − 762 One of the advantages of long-named options is that they can be
+ − 763 consistent from program to program. For example, users should be able
+ − 764 to expect the ``verbose'' option of any GNU program which has one, to be
+ − 765 spelled precisely @samp{--verbose}. To achieve this uniformity, look at
+ − 766 the table of common long-option names when you choose the option names
+ − 767 for your program (@pxref{Option Table}).
+ − 768
+ − 769 It is usually a good idea for file names given as ordinary arguments to
+ − 770 be input files only; any output files would be specified using options
+ − 771 (preferably @samp{-o} or @samp{--output}). Even if you allow an output
+ − 772 file name as an ordinary argument for compatibility, try to provide an
+ − 773 option as another way to specify it. This will lead to more consistency
+ − 774 among GNU utilities, and fewer idiosyncracies for users to remember.
+ − 775
462
+ − 776 @cindex standard command-line options
428
+ − 777 All programs should support two standard options: @samp{--version}
+ − 778 and @samp{--help}.
+ − 779
+ − 780 @table @code
462
+ − 781 @cindex @samp{--version} option
428
+ − 782 @item --version
+ − 783 This option should direct the program to print information about its name,
+ − 784 version, origin and legal status, all on standard output, and then exit
+ − 785 successfully. Other options and arguments should be ignored once this
+ − 786 is seen, and the program should not perform its normal function.
+ − 787
462
+ − 788 @cindex canonical name of a program
+ − 789 @cindex program's canonical name
428
+ − 790 The first line is meant to be easy for a program to parse; the version
+ − 791 number proper starts after the last space. In addition, it contains
+ − 792 the canonical name for this program, in this format:
+ − 793
+ − 794 @example
+ − 795 GNU Emacs 19.30
+ − 796 @end example
+ − 797
+ − 798 @noindent
+ − 799 The program's name should be a constant string; @emph{don't} compute it
+ − 800 from @code{argv[0]}. The idea is to state the standard or canonical
+ − 801 name for the program, not its file name. There are other ways to find
+ − 802 out the precise file name where a command is found in @code{PATH}.
+ − 803
+ − 804 If the program is a subsidiary part of a larger package, mention the
+ − 805 package name in parentheses, like this:
+ − 806
+ − 807 @example
+ − 808 emacsserver (GNU Emacs) 19.30
+ − 809 @end example
+ − 810
+ − 811 @noindent
+ − 812 If the package has a version number which is different from this
+ − 813 program's version number, you can mention the package version number
+ − 814 just before the close-parenthesis.
+ − 815
+ − 816 If you @strong{need} to mention the version numbers of libraries which
+ − 817 are distributed separately from the package which contains this program,
+ − 818 you can do so by printing an additional line of version info for each
+ − 819 library you want to mention. Use the same format for these lines as for
+ − 820 the first line.
+ − 821
+ − 822 Please do not mention all of the libraries that the program uses ``just
+ − 823 for completeness''---that would produce a lot of unhelpful clutter.
+ − 824 Please mention library version numbers only if you find in practice that
+ − 825 they are very important to you in debugging.
+ − 826
+ − 827 The following line, after the version number line or lines, should be a
+ − 828 copyright notice. If more than one copyright notice is called for, put
+ − 829 each on a separate line.
+ − 830
+ − 831 Next should follow a brief statement that the program is free software,
+ − 832 and that users are free to copy and change it on certain conditions. If
+ − 833 the program is covered by the GNU GPL, say so here. Also mention that
+ − 834 there is no warranty, to the extent permitted by law.
+ − 835
+ − 836 It is ok to finish the output with a list of the major authors of the
+ − 837 program, as a way of giving credit.
+ − 838
+ − 839 Here's an example of output that follows these rules:
+ − 840
+ − 841 @smallexample
+ − 842 GNU Emacs 19.34.5
+ − 843 Copyright (C) 1996 Free Software Foundation, Inc.
+ − 844 GNU Emacs comes with NO WARRANTY,
+ − 845 to the extent permitted by law.
+ − 846 You may redistribute copies of GNU Emacs
+ − 847 under the terms of the GNU General Public License.
+ − 848 For more information about these matters,
+ − 849 see the files named COPYING.
+ − 850 @end smallexample
+ − 851
+ − 852 You should adapt this to your program, of course, filling in the proper
+ − 853 year, copyright holder, name of program, and the references to
+ − 854 distribution terms, and changing the rest of the wording as necessary.
+ − 855
+ − 856 This copyright notice only needs to mention the most recent year in
+ − 857 which changes were made---there's no need to list the years for previous
+ − 858 versions' changes. You don't have to mention the name of the program in
+ − 859 these notices, if that is inconvenient, since it appeared in the first
+ − 860 line.
+ − 861
462
+ − 862 @cindex @samp{--help} option
428
+ − 863 @item --help
+ − 864 This option should output brief documentation for how to invoke the
+ − 865 program, on standard output, then exit successfully. Other options and
+ − 866 arguments should be ignored once this is seen, and the program should
+ − 867 not perform its normal function.
+ − 868
462
+ − 869 @cindex address for bug reports
+ − 870 @cindex bug reports
428
+ − 871 Near the end of the @samp{--help} option's output there should be a line
+ − 872 that says where to mail bug reports. It should have this format:
+ − 873
+ − 874 @example
+ − 875 Report bugs to @var{mailing-address}.
+ − 876 @end example
+ − 877 @end table
+ − 878
+ − 879 @node Option Table
+ − 880 @section Table of Long Options
462
+ − 881 @cindex long option names
+ − 882 @cindex table of long options
428
+ − 883
+ − 884 Here is a table of long options used by GNU programs. It is surely
+ − 885 incomplete, but we aim to list all the options that a new program might
+ − 886 want to be compatible with. If you use names not already in the table,
462
+ − 887 please send @email{bug-standards@@gnu.org} a list of them, with their
428
+ − 888 meanings, so we can update the table.
+ − 889
+ − 890 @c Please leave newlines between items in this table; it's much easier
+ − 891 @c to update when it isn't completely squashed together and unreadable.
+ − 892 @c When there is more than one short option for a long option name, put
+ − 893 @c a semicolon between the lists of the programs that use them, not a
+ − 894 @c period. --friedman
+ − 895
+ − 896 @table @samp
+ − 897 @item after-date
+ − 898 @samp{-N} in @code{tar}.
+ − 899
+ − 900 @item all
+ − 901 @samp{-a} in @code{du}, @code{ls}, @code{nm}, @code{stty}, @code{uname},
+ − 902 and @code{unexpand}.
+ − 903
+ − 904 @item all-text
+ − 905 @samp{-a} in @code{diff}.
+ − 906
+ − 907 @item almost-all
+ − 908 @samp{-A} in @code{ls}.
+ − 909
+ − 910 @item append
+ − 911 @samp{-a} in @code{etags}, @code{tee}, @code{time};
+ − 912 @samp{-r} in @code{tar}.
+ − 913
+ − 914 @item archive
+ − 915 @samp{-a} in @code{cp}.
+ − 916
+ − 917 @item archive-name
+ − 918 @samp{-n} in @code{shar}.
+ − 919
+ − 920 @item arglength
+ − 921 @samp{-l} in @code{m4}.
+ − 922
+ − 923 @item ascii
+ − 924 @samp{-a} in @code{diff}.
+ − 925
+ − 926 @item assign
+ − 927 @samp{-v} in @code{gawk}.
+ − 928
+ − 929 @item assume-new
+ − 930 @samp{-W} in Make.
+ − 931
+ − 932 @item assume-old
+ − 933 @samp{-o} in Make.
+ − 934
+ − 935 @item auto-check
+ − 936 @samp{-a} in @code{recode}.
+ − 937
+ − 938 @item auto-pager
+ − 939 @samp{-a} in @code{wdiff}.
+ − 940
+ − 941 @item auto-reference
+ − 942 @samp{-A} in @code{ptx}.
+ − 943
+ − 944 @item avoid-wraps
+ − 945 @samp{-n} in @code{wdiff}.
+ − 946
+ − 947 @item background
+ − 948 For server programs, run in the background.
+ − 949
+ − 950 @item backward-search
+ − 951 @samp{-B} in @code{ctags}.
+ − 952
+ − 953 @item basename
+ − 954 @samp{-f} in @code{shar}.
+ − 955
+ − 956 @item batch
+ − 957 Used in GDB.
+ − 958
+ − 959 @item baud
+ − 960 Used in GDB.
+ − 961
+ − 962 @item before
+ − 963 @samp{-b} in @code{tac}.
+ − 964
+ − 965 @item binary
+ − 966 @samp{-b} in @code{cpio} and @code{diff}.
+ − 967
+ − 968 @item bits-per-code
+ − 969 @samp{-b} in @code{shar}.
+ − 970
+ − 971 @item block-size
+ − 972 Used in @code{cpio} and @code{tar}.
+ − 973
+ − 974 @item blocks
+ − 975 @samp{-b} in @code{head} and @code{tail}.
+ − 976
+ − 977 @item break-file
+ − 978 @samp{-b} in @code{ptx}.
+ − 979
+ − 980 @item brief
+ − 981 Used in various programs to make output shorter.
+ − 982
+ − 983 @item bytes
+ − 984 @samp{-c} in @code{head}, @code{split}, and @code{tail}.
+ − 985
+ − 986 @item c@t{++}
+ − 987 @samp{-C} in @code{etags}.
+ − 988
+ − 989 @item catenate
+ − 990 @samp{-A} in @code{tar}.
+ − 991
+ − 992 @item cd
+ − 993 Used in various programs to specify the directory to use.
+ − 994
+ − 995 @item changes
+ − 996 @samp{-c} in @code{chgrp} and @code{chown}.
+ − 997
+ − 998 @item classify
+ − 999 @samp{-F} in @code{ls}.
+ − 1000
+ − 1001 @item colons
+ − 1002 @samp{-c} in @code{recode}.
+ − 1003
+ − 1004 @item command
+ − 1005 @samp{-c} in @code{su};
+ − 1006 @samp{-x} in GDB.
+ − 1007
+ − 1008 @item compare
+ − 1009 @samp{-d} in @code{tar}.
+ − 1010
+ − 1011 @item compat
+ − 1012 Used in @code{gawk}.
+ − 1013
+ − 1014 @item compress
+ − 1015 @samp{-Z} in @code{tar} and @code{shar}.
+ − 1016
+ − 1017 @item concatenate
+ − 1018 @samp{-A} in @code{tar}.
+ − 1019
+ − 1020 @item confirmation
+ − 1021 @samp{-w} in @code{tar}.
+ − 1022
+ − 1023 @item context
+ − 1024 Used in @code{diff}.
+ − 1025
+ − 1026 @item copyleft
+ − 1027 @samp{-W copyleft} in @code{gawk}.
+ − 1028
+ − 1029 @item copyright
+ − 1030 @samp{-C} in @code{ptx}, @code{recode}, and @code{wdiff};
+ − 1031 @samp{-W copyright} in @code{gawk}.
+ − 1032
+ − 1033 @item core
+ − 1034 Used in GDB.
+ − 1035
+ − 1036 @item count
+ − 1037 @samp{-q} in @code{who}.
+ − 1038
+ − 1039 @item count-links
+ − 1040 @samp{-l} in @code{du}.
+ − 1041
+ − 1042 @item create
+ − 1043 Used in @code{tar} and @code{cpio}.
+ − 1044
+ − 1045 @item cut-mark
+ − 1046 @samp{-c} in @code{shar}.
+ − 1047
+ − 1048 @item cxref
+ − 1049 @samp{-x} in @code{ctags}.
+ − 1050
+ − 1051 @item date
+ − 1052 @samp{-d} in @code{touch}.
+ − 1053
+ − 1054 @item debug
+ − 1055 @samp{-d} in Make and @code{m4};
+ − 1056 @samp{-t} in Bison.
+ − 1057
+ − 1058 @item define
+ − 1059 @samp{-D} in @code{m4}.
+ − 1060
+ − 1061 @item defines
+ − 1062 @samp{-d} in Bison and @code{ctags}.
+ − 1063
+ − 1064 @item delete
+ − 1065 @samp{-D} in @code{tar}.
+ − 1066
+ − 1067 @item dereference
+ − 1068 @samp{-L} in @code{chgrp}, @code{chown}, @code{cpio}, @code{du},
+ − 1069 @code{ls}, and @code{tar}.
+ − 1070
+ − 1071 @item dereference-args
+ − 1072 @samp{-D} in @code{du}.
+ − 1073
+ − 1074 @item device
+ − 1075 Specify an I/O device (special file name).
+ − 1076
+ − 1077 @item diacritics
+ − 1078 @samp{-d} in @code{recode}.
+ − 1079
+ − 1080 @item dictionary-order
+ − 1081 @samp{-d} in @code{look}.
+ − 1082
+ − 1083 @item diff
+ − 1084 @samp{-d} in @code{tar}.
+ − 1085
+ − 1086 @item digits
+ − 1087 @samp{-n} in @code{csplit}.
+ − 1088
+ − 1089 @item directory
+ − 1090 Specify the directory to use, in various programs. In @code{ls}, it
+ − 1091 means to show directories themselves rather than their contents. In
+ − 1092 @code{rm} and @code{ln}, it means to not treat links to directories
+ − 1093 specially.
+ − 1094
+ − 1095 @item discard-all
+ − 1096 @samp{-x} in @code{strip}.
+ − 1097
+ − 1098 @item discard-locals
+ − 1099 @samp{-X} in @code{strip}.
+ − 1100
+ − 1101 @item dry-run
+ − 1102 @samp{-n} in Make.
+ − 1103
+ − 1104 @item ed
+ − 1105 @samp{-e} in @code{diff}.
+ − 1106
+ − 1107 @item elide-empty-files
+ − 1108 @samp{-z} in @code{csplit}.
+ − 1109
+ − 1110 @item end-delete
+ − 1111 @samp{-x} in @code{wdiff}.
+ − 1112
+ − 1113 @item end-insert
+ − 1114 @samp{-z} in @code{wdiff}.
+ − 1115
+ − 1116 @item entire-new-file
+ − 1117 @samp{-N} in @code{diff}.
+ − 1118
+ − 1119 @item environment-overrides
+ − 1120 @samp{-e} in Make.
+ − 1121
+ − 1122 @item eof
+ − 1123 @samp{-e} in @code{xargs}.
+ − 1124
+ − 1125 @item epoch
+ − 1126 Used in GDB.
+ − 1127
+ − 1128 @item error-limit
+ − 1129 Used in @code{makeinfo}.
+ − 1130
+ − 1131 @item error-output
+ − 1132 @samp{-o} in @code{m4}.
+ − 1133
+ − 1134 @item escape
+ − 1135 @samp{-b} in @code{ls}.
+ − 1136
+ − 1137 @item exclude-from
+ − 1138 @samp{-X} in @code{tar}.
+ − 1139
+ − 1140 @item exec
+ − 1141 Used in GDB.
+ − 1142
+ − 1143 @item exit
+ − 1144 @samp{-x} in @code{xargs}.
+ − 1145
+ − 1146 @item exit-0
+ − 1147 @samp{-e} in @code{unshar}.
+ − 1148
+ − 1149 @item expand-tabs
+ − 1150 @samp{-t} in @code{diff}.
+ − 1151
+ − 1152 @item expression
+ − 1153 @samp{-e} in @code{sed}.
+ − 1154
+ − 1155 @item extern-only
+ − 1156 @samp{-g} in @code{nm}.
+ − 1157
+ − 1158 @item extract
+ − 1159 @samp{-i} in @code{cpio};
+ − 1160 @samp{-x} in @code{tar}.
+ − 1161
+ − 1162 @item faces
+ − 1163 @samp{-f} in @code{finger}.
+ − 1164
+ − 1165 @item fast
+ − 1166 @samp{-f} in @code{su}.
+ − 1167
+ − 1168 @item fatal-warnings
+ − 1169 @samp{-E} in @code{m4}.
+ − 1170
+ − 1171 @item file
+ − 1172 @samp{-f} in @code{info}, @code{gawk}, Make, @code{mt}, and @code{tar};
+ − 1173 @samp{-n} in @code{sed};
+ − 1174 @samp{-r} in @code{touch}.
+ − 1175
+ − 1176 @item field-separator
+ − 1177 @samp{-F} in @code{gawk}.
+ − 1178
+ − 1179 @item file-prefix
+ − 1180 @samp{-b} in Bison.
+ − 1181
+ − 1182 @item file-type
+ − 1183 @samp{-F} in @code{ls}.
+ − 1184
+ − 1185 @item files-from
+ − 1186 @samp{-T} in @code{tar}.
+ − 1187
+ − 1188 @item fill-column
+ − 1189 Used in @code{makeinfo}.
+ − 1190
+ − 1191 @item flag-truncation
+ − 1192 @samp{-F} in @code{ptx}.
+ − 1193
+ − 1194 @item fixed-output-files
+ − 1195 @samp{-y} in Bison.
+ − 1196
+ − 1197 @item follow
+ − 1198 @samp{-f} in @code{tail}.
+ − 1199
+ − 1200 @item footnote-style
+ − 1201 Used in @code{makeinfo}.
+ − 1202
+ − 1203 @item force
+ − 1204 @samp{-f} in @code{cp}, @code{ln}, @code{mv}, and @code{rm}.
+ − 1205
+ − 1206 @item force-prefix
+ − 1207 @samp{-F} in @code{shar}.
+ − 1208
+ − 1209 @item foreground
+ − 1210 For server programs, run in the foreground;
+ − 1211 in other words, don't do anything special to run the server
+ − 1212 in the background.
+ − 1213
+ − 1214 @item format
+ − 1215 Used in @code{ls}, @code{time}, and @code{ptx}.
+ − 1216
+ − 1217 @item freeze-state
+ − 1218 @samp{-F} in @code{m4}.
+ − 1219
+ − 1220 @item fullname
+ − 1221 Used in GDB.
+ − 1222
+ − 1223 @item gap-size
+ − 1224 @samp{-g} in @code{ptx}.
+ − 1225
+ − 1226 @item get
+ − 1227 @samp{-x} in @code{tar}.
+ − 1228
+ − 1229 @item graphic
+ − 1230 @samp{-i} in @code{ul}.
+ − 1231
+ − 1232 @item graphics
+ − 1233 @samp{-g} in @code{recode}.
+ − 1234
+ − 1235 @item group
+ − 1236 @samp{-g} in @code{install}.
+ − 1237
+ − 1238 @item gzip
+ − 1239 @samp{-z} in @code{tar} and @code{shar}.
+ − 1240
+ − 1241 @item hashsize
+ − 1242 @samp{-H} in @code{m4}.
+ − 1243
+ − 1244 @item header
+ − 1245 @samp{-h} in @code{objdump} and @code{recode}
+ − 1246
+ − 1247 @item heading
+ − 1248 @samp{-H} in @code{who}.
+ − 1249
+ − 1250 @item help
+ − 1251 Used to ask for brief usage information.
+ − 1252
+ − 1253 @item here-delimiter
+ − 1254 @samp{-d} in @code{shar}.
+ − 1255
+ − 1256 @item hide-control-chars
+ − 1257 @samp{-q} in @code{ls}.
+ − 1258
462
+ − 1259 @item html
+ − 1260 In @code{makeinfo}, output HTML.
+ − 1261
428
+ − 1262 @item idle
+ − 1263 @samp{-u} in @code{who}.
+ − 1264
+ − 1265 @item ifdef
+ − 1266 @samp{-D} in @code{diff}.
+ − 1267
+ − 1268 @item ignore
+ − 1269 @samp{-I} in @code{ls};
+ − 1270 @samp{-x} in @code{recode}.
+ − 1271
+ − 1272 @item ignore-all-space
+ − 1273 @samp{-w} in @code{diff}.
+ − 1274
+ − 1275 @item ignore-backups
+ − 1276 @samp{-B} in @code{ls}.
+ − 1277
+ − 1278 @item ignore-blank-lines
+ − 1279 @samp{-B} in @code{diff}.
+ − 1280
+ − 1281 @item ignore-case
+ − 1282 @samp{-f} in @code{look} and @code{ptx};
+ − 1283 @samp{-i} in @code{diff} and @code{wdiff}.
+ − 1284
+ − 1285 @item ignore-errors
+ − 1286 @samp{-i} in Make.
+ − 1287
+ − 1288 @item ignore-file
+ − 1289 @samp{-i} in @code{ptx}.
+ − 1290
+ − 1291 @item ignore-indentation
+ − 1292 @samp{-I} in @code{etags}.
+ − 1293
+ − 1294 @item ignore-init-file
+ − 1295 @samp{-f} in Oleo.
+ − 1296
+ − 1297 @item ignore-interrupts
+ − 1298 @samp{-i} in @code{tee}.
+ − 1299
+ − 1300 @item ignore-matching-lines
+ − 1301 @samp{-I} in @code{diff}.
+ − 1302
+ − 1303 @item ignore-space-change
+ − 1304 @samp{-b} in @code{diff}.
+ − 1305
+ − 1306 @item ignore-zeros
+ − 1307 @samp{-i} in @code{tar}.
+ − 1308
+ − 1309 @item include
+ − 1310 @samp{-i} in @code{etags};
+ − 1311 @samp{-I} in @code{m4}.
+ − 1312
+ − 1313 @item include-dir
+ − 1314 @samp{-I} in Make.
+ − 1315
+ − 1316 @item incremental
+ − 1317 @samp{-G} in @code{tar}.
+ − 1318
+ − 1319 @item info
+ − 1320 @samp{-i}, @samp{-l}, and @samp{-m} in Finger.
+ − 1321
462
+ − 1322 @item init-file
+ − 1323 In some programs, specify the name of the file to read as the user's
+ − 1324 init file.
+ − 1325
428
+ − 1326 @item initial
+ − 1327 @samp{-i} in @code{expand}.
+ − 1328
+ − 1329 @item initial-tab
+ − 1330 @samp{-T} in @code{diff}.
+ − 1331
+ − 1332 @item inode
+ − 1333 @samp{-i} in @code{ls}.
+ − 1334
+ − 1335 @item interactive
+ − 1336 @samp{-i} in @code{cp}, @code{ln}, @code{mv}, @code{rm};
+ − 1337 @samp{-e} in @code{m4};
+ − 1338 @samp{-p} in @code{xargs};
+ − 1339 @samp{-w} in @code{tar}.
+ − 1340
+ − 1341 @item intermix-type
+ − 1342 @samp{-p} in @code{shar}.
+ − 1343
462
+ − 1344 @item iso-8601
+ − 1345 Used in @code{date}
+ − 1346
428
+ − 1347 @item jobs
+ − 1348 @samp{-j} in Make.
+ − 1349
+ − 1350 @item just-print
+ − 1351 @samp{-n} in Make.
+ − 1352
+ − 1353 @item keep-going
+ − 1354 @samp{-k} in Make.
+ − 1355
+ − 1356 @item keep-files
+ − 1357 @samp{-k} in @code{csplit}.
+ − 1358
+ − 1359 @item kilobytes
+ − 1360 @samp{-k} in @code{du} and @code{ls}.
+ − 1361
+ − 1362 @item language
+ − 1363 @samp{-l} in @code{etags}.
+ − 1364
+ − 1365 @item less-mode
+ − 1366 @samp{-l} in @code{wdiff}.
+ − 1367
+ − 1368 @item level-for-gzip
+ − 1369 @samp{-g} in @code{shar}.
+ − 1370
+ − 1371 @item line-bytes
+ − 1372 @samp{-C} in @code{split}.
+ − 1373
+ − 1374 @item lines
+ − 1375 Used in @code{split}, @code{head}, and @code{tail}.
+ − 1376
+ − 1377 @item link
+ − 1378 @samp{-l} in @code{cpio}.
+ − 1379
+ − 1380 @item lint
+ − 1381 @itemx lint-old
+ − 1382 Used in @code{gawk}.
+ − 1383
+ − 1384 @item list
+ − 1385 @samp{-t} in @code{cpio};
+ − 1386 @samp{-l} in @code{recode}.
+ − 1387
+ − 1388 @item list
+ − 1389 @samp{-t} in @code{tar}.
+ − 1390
+ − 1391 @item literal
+ − 1392 @samp{-N} in @code{ls}.
+ − 1393
+ − 1394 @item load-average
+ − 1395 @samp{-l} in Make.
+ − 1396
+ − 1397 @item login
+ − 1398 Used in @code{su}.
+ − 1399
+ − 1400 @item machine
+ − 1401 No listing of which programs already use this;
+ − 1402 someone should check to
+ − 1403 see if any actually do, and tell @email{gnu@@gnu.org}.
+ − 1404
+ − 1405 @item macro-name
+ − 1406 @samp{-M} in @code{ptx}.
+ − 1407
+ − 1408 @item mail
+ − 1409 @samp{-m} in @code{hello} and @code{uname}.
+ − 1410
+ − 1411 @item make-directories
+ − 1412 @samp{-d} in @code{cpio}.
+ − 1413
+ − 1414 @item makefile
+ − 1415 @samp{-f} in Make.
+ − 1416
+ − 1417 @item mapped
+ − 1418 Used in GDB.
+ − 1419
+ − 1420 @item max-args
+ − 1421 @samp{-n} in @code{xargs}.
+ − 1422
+ − 1423 @item max-chars
+ − 1424 @samp{-n} in @code{xargs}.
+ − 1425
+ − 1426 @item max-lines
+ − 1427 @samp{-l} in @code{xargs}.
+ − 1428
+ − 1429 @item max-load
+ − 1430 @samp{-l} in Make.
+ − 1431
+ − 1432 @item max-procs
+ − 1433 @samp{-P} in @code{xargs}.
+ − 1434
+ − 1435 @item mesg
+ − 1436 @samp{-T} in @code{who}.
+ − 1437
+ − 1438 @item message
+ − 1439 @samp{-T} in @code{who}.
+ − 1440
+ − 1441 @item minimal
+ − 1442 @samp{-d} in @code{diff}.
+ − 1443
+ − 1444 @item mixed-uuencode
+ − 1445 @samp{-M} in @code{shar}.
+ − 1446
+ − 1447 @item mode
+ − 1448 @samp{-m} in @code{install}, @code{mkdir}, and @code{mkfifo}.
+ − 1449
+ − 1450 @item modification-time
+ − 1451 @samp{-m} in @code{tar}.
+ − 1452
+ − 1453 @item multi-volume
+ − 1454 @samp{-M} in @code{tar}.
+ − 1455
+ − 1456 @item name-prefix
+ − 1457 @samp{-a} in Bison.
+ − 1458
+ − 1459 @item nesting-limit
+ − 1460 @samp{-L} in @code{m4}.
+ − 1461
+ − 1462 @item net-headers
+ − 1463 @samp{-a} in @code{shar}.
+ − 1464
+ − 1465 @item new-file
+ − 1466 @samp{-W} in Make.
+ − 1467
+ − 1468 @item no-builtin-rules
+ − 1469 @samp{-r} in Make.
+ − 1470
+ − 1471 @item no-character-count
+ − 1472 @samp{-w} in @code{shar}.
+ − 1473
+ − 1474 @item no-check-existing
+ − 1475 @samp{-x} in @code{shar}.
+ − 1476
+ − 1477 @item no-common
+ − 1478 @samp{-3} in @code{wdiff}.
+ − 1479
+ − 1480 @item no-create
+ − 1481 @samp{-c} in @code{touch}.
+ − 1482
+ − 1483 @item no-defines
+ − 1484 @samp{-D} in @code{etags}.
+ − 1485
+ − 1486 @item no-deleted
+ − 1487 @samp{-1} in @code{wdiff}.
+ − 1488
+ − 1489 @item no-dereference
+ − 1490 @samp{-d} in @code{cp}.
+ − 1491
+ − 1492 @item no-inserted
+ − 1493 @samp{-2} in @code{wdiff}.
+ − 1494
+ − 1495 @item no-keep-going
+ − 1496 @samp{-S} in Make.
+ − 1497
+ − 1498 @item no-lines
+ − 1499 @samp{-l} in Bison.
+ − 1500
+ − 1501 @item no-piping
+ − 1502 @samp{-P} in @code{shar}.
+ − 1503
+ − 1504 @item no-prof
+ − 1505 @samp{-e} in @code{gprof}.
+ − 1506
+ − 1507 @item no-regex
+ − 1508 @samp{-R} in @code{etags}.
+ − 1509
+ − 1510 @item no-sort
+ − 1511 @samp{-p} in @code{nm}.
+ − 1512
+ − 1513 @item no-split
+ − 1514 Used in @code{makeinfo}.
+ − 1515
+ − 1516 @item no-static
+ − 1517 @samp{-a} in @code{gprof}.
+ − 1518
+ − 1519 @item no-time
+ − 1520 @samp{-E} in @code{gprof}.
+ − 1521
+ − 1522 @item no-timestamp
+ − 1523 @samp{-m} in @code{shar}.
+ − 1524
+ − 1525 @item no-validate
+ − 1526 Used in @code{makeinfo}.
+ − 1527
+ − 1528 @item no-wait
+ − 1529 Used in @code{emacsclient}.
+ − 1530
+ − 1531 @item no-warn
+ − 1532 Used in various programs to inhibit warnings.
+ − 1533
+ − 1534 @item node
+ − 1535 @samp{-n} in @code{info}.
+ − 1536
+ − 1537 @item nodename
+ − 1538 @samp{-n} in @code{uname}.
+ − 1539
+ − 1540 @item nonmatching
+ − 1541 @samp{-f} in @code{cpio}.
+ − 1542
+ − 1543 @item nstuff
+ − 1544 @samp{-n} in @code{objdump}.
+ − 1545
+ − 1546 @item null
+ − 1547 @samp{-0} in @code{xargs}.
+ − 1548
+ − 1549 @item number
+ − 1550 @samp{-n} in @code{cat}.
+ − 1551
+ − 1552 @item number-nonblank
+ − 1553 @samp{-b} in @code{cat}.
+ − 1554
+ − 1555 @item numeric-sort
+ − 1556 @samp{-n} in @code{nm}.
+ − 1557
+ − 1558 @item numeric-uid-gid
+ − 1559 @samp{-n} in @code{cpio} and @code{ls}.
+ − 1560
+ − 1561 @item nx
+ − 1562 Used in GDB.
+ − 1563
+ − 1564 @item old-archive
+ − 1565 @samp{-o} in @code{tar}.
+ − 1566
+ − 1567 @item old-file
+ − 1568 @samp{-o} in Make.
+ − 1569
+ − 1570 @item one-file-system
+ − 1571 @samp{-l} in @code{tar}, @code{cp}, and @code{du}.
+ − 1572
+ − 1573 @item only-file
+ − 1574 @samp{-o} in @code{ptx}.
+ − 1575
+ − 1576 @item only-prof
+ − 1577 @samp{-f} in @code{gprof}.
+ − 1578
+ − 1579 @item only-time
+ − 1580 @samp{-F} in @code{gprof}.
+ − 1581
+ − 1582 @item options
+ − 1583 @samp{-o} in @code{getopt}, @code{fdlist}, @code{fdmount},
+ − 1584 @code{fdmountd}, and @code{fdumount}.
+ − 1585
+ − 1586 @item output
+ − 1587 In various programs, specify the output file name.
+ − 1588
+ − 1589 @item output-prefix
+ − 1590 @samp{-o} in @code{shar}.
+ − 1591
+ − 1592 @item override
+ − 1593 @samp{-o} in @code{rm}.
+ − 1594
+ − 1595 @item overwrite
+ − 1596 @samp{-c} in @code{unshar}.
+ − 1597
+ − 1598 @item owner
+ − 1599 @samp{-o} in @code{install}.
+ − 1600
+ − 1601 @item paginate
+ − 1602 @samp{-l} in @code{diff}.
+ − 1603
+ − 1604 @item paragraph-indent
+ − 1605 Used in @code{makeinfo}.
+ − 1606
+ − 1607 @item parents
+ − 1608 @samp{-p} in @code{mkdir} and @code{rmdir}.
+ − 1609
+ − 1610 @item pass-all
+ − 1611 @samp{-p} in @code{ul}.
+ − 1612
+ − 1613 @item pass-through
+ − 1614 @samp{-p} in @code{cpio}.
+ − 1615
+ − 1616 @item port
+ − 1617 @samp{-P} in @code{finger}.
+ − 1618
+ − 1619 @item portability
+ − 1620 @samp{-c} in @code{cpio} and @code{tar}.
+ − 1621
+ − 1622 @item posix
+ − 1623 Used in @code{gawk}.
+ − 1624
+ − 1625 @item prefix-builtins
+ − 1626 @samp{-P} in @code{m4}.
+ − 1627
+ − 1628 @item prefix
+ − 1629 @samp{-f} in @code{csplit}.
+ − 1630
+ − 1631 @item preserve
+ − 1632 Used in @code{tar} and @code{cp}.
+ − 1633
+ − 1634 @item preserve-environment
+ − 1635 @samp{-p} in @code{su}.
+ − 1636
+ − 1637 @item preserve-modification-time
+ − 1638 @samp{-m} in @code{cpio}.
+ − 1639
+ − 1640 @item preserve-order
+ − 1641 @samp{-s} in @code{tar}.
+ − 1642
+ − 1643 @item preserve-permissions
+ − 1644 @samp{-p} in @code{tar}.
+ − 1645
+ − 1646 @item print
+ − 1647 @samp{-l} in @code{diff}.
+ − 1648
+ − 1649 @item print-chars
+ − 1650 @samp{-L} in @code{cmp}.
+ − 1651
+ − 1652 @item print-data-base
+ − 1653 @samp{-p} in Make.
+ − 1654
+ − 1655 @item print-directory
+ − 1656 @samp{-w} in Make.
+ − 1657
+ − 1658 @item print-file-name
+ − 1659 @samp{-o} in @code{nm}.
+ − 1660
+ − 1661 @item print-symdefs
+ − 1662 @samp{-s} in @code{nm}.
+ − 1663
+ − 1664 @item printer
+ − 1665 @samp{-p} in @code{wdiff}.
+ − 1666
+ − 1667 @item prompt
+ − 1668 @samp{-p} in @code{ed}.
+ − 1669
+ − 1670 @item proxy
+ − 1671 Specify an HTTP proxy.
+ − 1672
+ − 1673 @item query-user
+ − 1674 @samp{-X} in @code{shar}.
+ − 1675
+ − 1676 @item question
+ − 1677 @samp{-q} in Make.
+ − 1678
+ − 1679 @item quiet
2665
+ − 1680 Used in many programs to inhibit the usual output. @strong{Every
+ − 1681 program accepting @samp{--quiet}} should accept @samp{--silent} as a
428
+ − 1682 synonym.
+ − 1683
+ − 1684 @item quiet-unshar
+ − 1685 @samp{-Q} in @code{shar}
+ − 1686
+ − 1687 @item quote-name
+ − 1688 @samp{-Q} in @code{ls}.
+ − 1689
+ − 1690 @item rcs
+ − 1691 @samp{-n} in @code{diff}.
+ − 1692
+ − 1693 @item re-interval
+ − 1694 Used in @code{gawk}.
+ − 1695
+ − 1696 @item read-full-blocks
+ − 1697 @samp{-B} in @code{tar}.
+ − 1698
+ − 1699 @item readnow
+ − 1700 Used in GDB.
+ − 1701
+ − 1702 @item recon
+ − 1703 @samp{-n} in Make.
+ − 1704
+ − 1705 @item record-number
+ − 1706 @samp{-R} in @code{tar}.
+ − 1707
+ − 1708 @item recursive
+ − 1709 Used in @code{chgrp}, @code{chown}, @code{cp}, @code{ls}, @code{diff},
+ − 1710 and @code{rm}.
+ − 1711
+ − 1712 @item reference-limit
+ − 1713 Used in @code{makeinfo}.
+ − 1714
+ − 1715 @item references
+ − 1716 @samp{-r} in @code{ptx}.
+ − 1717
+ − 1718 @item regex
+ − 1719 @samp{-r} in @code{tac} and @code{etags}.
+ − 1720
+ − 1721 @item release
+ − 1722 @samp{-r} in @code{uname}.
+ − 1723
+ − 1724 @item reload-state
+ − 1725 @samp{-R} in @code{m4}.
+ − 1726
+ − 1727 @item relocation
+ − 1728 @samp{-r} in @code{objdump}.
+ − 1729
+ − 1730 @item rename
+ − 1731 @samp{-r} in @code{cpio}.
+ − 1732
+ − 1733 @item replace
+ − 1734 @samp{-i} in @code{xargs}.
+ − 1735
+ − 1736 @item report-identical-files
+ − 1737 @samp{-s} in @code{diff}.
+ − 1738
+ − 1739 @item reset-access-time
+ − 1740 @samp{-a} in @code{cpio}.
+ − 1741
+ − 1742 @item reverse
+ − 1743 @samp{-r} in @code{ls} and @code{nm}.
+ − 1744
+ − 1745 @item reversed-ed
+ − 1746 @samp{-f} in @code{diff}.
+ − 1747
+ − 1748 @item right-side-defs
+ − 1749 @samp{-R} in @code{ptx}.
+ − 1750
+ − 1751 @item same-order
+ − 1752 @samp{-s} in @code{tar}.
+ − 1753
+ − 1754 @item same-permissions
+ − 1755 @samp{-p} in @code{tar}.
+ − 1756
+ − 1757 @item save
+ − 1758 @samp{-g} in @code{stty}.
+ − 1759
+ − 1760 @item se
+ − 1761 Used in GDB.
+ − 1762
+ − 1763 @item sentence-regexp
+ − 1764 @samp{-S} in @code{ptx}.
+ − 1765
+ − 1766 @item separate-dirs
+ − 1767 @samp{-S} in @code{du}.
+ − 1768
+ − 1769 @item separator
+ − 1770 @samp{-s} in @code{tac}.
+ − 1771
+ − 1772 @item sequence
+ − 1773 Used by @code{recode} to chose files or pipes for sequencing passes.
+ − 1774
+ − 1775 @item shell
+ − 1776 @samp{-s} in @code{su}.
+ − 1777
+ − 1778 @item show-all
+ − 1779 @samp{-A} in @code{cat}.
+ − 1780
+ − 1781 @item show-c-function
+ − 1782 @samp{-p} in @code{diff}.
+ − 1783
+ − 1784 @item show-ends
+ − 1785 @samp{-E} in @code{cat}.
+ − 1786
+ − 1787 @item show-function-line
+ − 1788 @samp{-F} in @code{diff}.
+ − 1789
+ − 1790 @item show-tabs
+ − 1791 @samp{-T} in @code{cat}.
+ − 1792
+ − 1793 @item silent
+ − 1794 Used in many programs to inhibit the usual output.
2665
+ − 1795 @strong{Every program accepting
+ − 1796 @samp{--silent}} should accept @samp{--quiet} as a synonym.
428
+ − 1797
+ − 1798 @item size
+ − 1799 @samp{-s} in @code{ls}.
+ − 1800
+ − 1801 @item socket
+ − 1802 Specify a file descriptor for a network server to use for its socket,
+ − 1803 instead of opening and binding a new socket. This provides a way to
+ − 1804 run, in a nonpriveledged process, a server that normally needs a
+ − 1805 reserved port number.
+ − 1806
+ − 1807 @item sort
+ − 1808 Used in @code{ls}.
+ − 1809
+ − 1810 @item source
+ − 1811 @samp{-W source} in @code{gawk}.
+ − 1812
+ − 1813 @item sparse
+ − 1814 @samp{-S} in @code{tar}.
+ − 1815
+ − 1816 @item speed-large-files
+ − 1817 @samp{-H} in @code{diff}.
+ − 1818
+ − 1819 @item split-at
+ − 1820 @samp{-E} in @code{unshar}.
+ − 1821
+ − 1822 @item split-size-limit
+ − 1823 @samp{-L} in @code{shar}.
+ − 1824
+ − 1825 @item squeeze-blank
+ − 1826 @samp{-s} in @code{cat}.
+ − 1827
+ − 1828 @item start-delete
+ − 1829 @samp{-w} in @code{wdiff}.
+ − 1830
+ − 1831 @item start-insert
+ − 1832 @samp{-y} in @code{wdiff}.
+ − 1833
+ − 1834 @item starting-file
+ − 1835 Used in @code{tar} and @code{diff} to specify which file within
+ − 1836 a directory to start processing with.
+ − 1837
+ − 1838 @item statistics
+ − 1839 @samp{-s} in @code{wdiff}.
+ − 1840
+ − 1841 @item stdin-file-list
+ − 1842 @samp{-S} in @code{shar}.
+ − 1843
+ − 1844 @item stop
+ − 1845 @samp{-S} in Make.
+ − 1846
+ − 1847 @item strict
+ − 1848 @samp{-s} in @code{recode}.
+ − 1849
+ − 1850 @item strip
+ − 1851 @samp{-s} in @code{install}.
+ − 1852
+ − 1853 @item strip-all
+ − 1854 @samp{-s} in @code{strip}.
+ − 1855
+ − 1856 @item strip-debug
+ − 1857 @samp{-S} in @code{strip}.
+ − 1858
+ − 1859 @item submitter
+ − 1860 @samp{-s} in @code{shar}.
+ − 1861
+ − 1862 @item suffix
+ − 1863 @samp{-S} in @code{cp}, @code{ln}, @code{mv}.
+ − 1864
+ − 1865 @item suffix-format
+ − 1866 @samp{-b} in @code{csplit}.
+ − 1867
+ − 1868 @item sum
+ − 1869 @samp{-s} in @code{gprof}.
+ − 1870
+ − 1871 @item summarize
+ − 1872 @samp{-s} in @code{du}.
+ − 1873
+ − 1874 @item symbolic
+ − 1875 @samp{-s} in @code{ln}.
+ − 1876
+ − 1877 @item symbols
+ − 1878 Used in GDB and @code{objdump}.
+ − 1879
+ − 1880 @item synclines
+ − 1881 @samp{-s} in @code{m4}.
+ − 1882
+ − 1883 @item sysname
+ − 1884 @samp{-s} in @code{uname}.
+ − 1885
+ − 1886 @item tabs
+ − 1887 @samp{-t} in @code{expand} and @code{unexpand}.
+ − 1888
+ − 1889 @item tabsize
+ − 1890 @samp{-T} in @code{ls}.
+ − 1891
+ − 1892 @item terminal
+ − 1893 @samp{-T} in @code{tput} and @code{ul}.
+ − 1894 @samp{-t} in @code{wdiff}.
+ − 1895
+ − 1896 @item text
+ − 1897 @samp{-a} in @code{diff}.
+ − 1898
+ − 1899 @item text-files
+ − 1900 @samp{-T} in @code{shar}.
+ − 1901
+ − 1902 @item time
+ − 1903 Used in @code{ls} and @code{touch}.
+ − 1904
+ − 1905 @item timeout
+ − 1906 Specify how long to wait before giving up on some operation.
+ − 1907
+ − 1908 @item to-stdout
+ − 1909 @samp{-O} in @code{tar}.
+ − 1910
+ − 1911 @item total
+ − 1912 @samp{-c} in @code{du}.
+ − 1913
+ − 1914 @item touch
+ − 1915 @samp{-t} in Make, @code{ranlib}, and @code{recode}.
+ − 1916
+ − 1917 @item trace
+ − 1918 @samp{-t} in @code{m4}.
+ − 1919
+ − 1920 @item traditional
+ − 1921 @samp{-t} in @code{hello};
+ − 1922 @samp{-W traditional} in @code{gawk};
+ − 1923 @samp{-G} in @code{ed}, @code{m4}, and @code{ptx}.
+ − 1924
+ − 1925 @item tty
+ − 1926 Used in GDB.
+ − 1927
+ − 1928 @item typedefs
+ − 1929 @samp{-t} in @code{ctags}.
+ − 1930
+ − 1931 @item typedefs-and-c++
+ − 1932 @samp{-T} in @code{ctags}.
+ − 1933
+ − 1934 @item typeset-mode
+ − 1935 @samp{-t} in @code{ptx}.
+ − 1936
+ − 1937 @item uncompress
+ − 1938 @samp{-z} in @code{tar}.
+ − 1939
+ − 1940 @item unconditional
+ − 1941 @samp{-u} in @code{cpio}.
+ − 1942
+ − 1943 @item undefine
+ − 1944 @samp{-U} in @code{m4}.
+ − 1945
+ − 1946 @item undefined-only
+ − 1947 @samp{-u} in @code{nm}.
+ − 1948
+ − 1949 @item update
+ − 1950 @samp{-u} in @code{cp}, @code{ctags}, @code{mv}, @code{tar}.
+ − 1951
+ − 1952 @item usage
+ − 1953 Used in @code{gawk}; same as @samp{--help}.
+ − 1954
+ − 1955 @item uuencode
+ − 1956 @samp{-B} in @code{shar}.
+ − 1957
+ − 1958 @item vanilla-operation
+ − 1959 @samp{-V} in @code{shar}.
+ − 1960
+ − 1961 @item verbose
+ − 1962 Print more information about progress. Many programs support this.
+ − 1963
+ − 1964 @item verify
+ − 1965 @samp{-W} in @code{tar}.
+ − 1966
+ − 1967 @item version
+ − 1968 Print the version number.
+ − 1969
+ − 1970 @item version-control
+ − 1971 @samp{-V} in @code{cp}, @code{ln}, @code{mv}.
+ − 1972
+ − 1973 @item vgrind
+ − 1974 @samp{-v} in @code{ctags}.
+ − 1975
+ − 1976 @item volume
+ − 1977 @samp{-V} in @code{tar}.
+ − 1978
+ − 1979 @item what-if
+ − 1980 @samp{-W} in Make.
+ − 1981
+ − 1982 @item whole-size-limit
+ − 1983 @samp{-l} in @code{shar}.
+ − 1984
+ − 1985 @item width
+ − 1986 @samp{-w} in @code{ls} and @code{ptx}.
+ − 1987
+ − 1988 @item word-regexp
+ − 1989 @samp{-W} in @code{ptx}.
+ − 1990
+ − 1991 @item writable
+ − 1992 @samp{-T} in @code{who}.
+ − 1993
+ − 1994 @item zeros
+ − 1995 @samp{-z} in @code{gprof}.
+ − 1996 @end table
+ − 1997
+ − 1998 @node Memory Usage
+ − 1999 @section Memory Usage
462
+ − 2000 @cindex memory usage
+ − 2001
+ − 2002 If a program typically uses just a few meg of memory, don't bother making any
428
+ − 2003 effort to reduce memory usage. For example, if it is impractical for
+ − 2004 other reasons to operate on files more than a few meg long, it is
+ − 2005 reasonable to read entire input files into core to operate on them.
+ − 2006
+ − 2007 However, for programs such as @code{cat} or @code{tail}, that can
+ − 2008 usefully operate on very large files, it is important to avoid using a
+ − 2009 technique that would artificially limit the size of files it can handle.
+ − 2010 If a program works by lines and could be applied to arbitrary
+ − 2011 user-supplied input files, it should keep only a line in memory, because
+ − 2012 this is not very hard and users will want to be able to operate on input
+ − 2013 files that are bigger than will fit in core all at once.
+ − 2014
+ − 2015 If your program creates complicated data structures, just make them in
+ − 2016 core and give a fatal error if @code{malloc} returns zero.
+ − 2017
462
+ − 2018 @node File Usage
+ − 2019 @section File Usage
+ − 2020 @cindex file usage
+ − 2021
+ − 2022 Programs should be prepared to operate when @file{/usr} and @file{/etc}
+ − 2023 are read-only file systems. Thus, if the program manages log files,
+ − 2024 lock files, backup files, score files, or any other files which are
+ − 2025 modified for internal purposes, these files should not be stored in
+ − 2026 @file{/usr} or @file{/etc}.
+ − 2027
+ − 2028 There are two exceptions. @file{/etc} is used to store system
+ − 2029 configuration information; it is reasonable for a program to modify
+ − 2030 files in @file{/etc} when its job is to update the system configuration.
+ − 2031 Also, if the user explicitly asks to modify one file in a directory, it
+ − 2032 is reasonable for the program to store other files in the same
+ − 2033 directory.
+ − 2034
428
+ − 2035 @node Writing C
+ − 2036 @chapter Making The Best Use of C
+ − 2037
+ − 2038 This @value{CHAPTER} provides advice on how best to use the C language
+ − 2039 when writing GNU software.
+ − 2040
+ − 2041 @menu
+ − 2042 * Formatting:: Formatting Your Source Code
+ − 2043 * Comments:: Commenting Your Work
+ − 2044 * Syntactic Conventions:: Clean Use of C Constructs
+ − 2045 * Names:: Naming Variables and Functions
+ − 2046 * System Portability:: Portability between different operating systems
+ − 2047 * CPU Portability:: Supporting the range of CPU types
+ − 2048 * System Functions:: Portability and ``standard'' library functions
+ − 2049 * Internationalization:: Techniques for internationalization
+ − 2050 * Mmap:: How you can safely use @code{mmap}.
+ − 2051 @end menu
+ − 2052
+ − 2053 @node Formatting
+ − 2054 @section Formatting Your Source Code
462
+ − 2055 @cindex formatting source code
+ − 2056
+ − 2057 @cindex open brace
+ − 2058 @cindex braces, in C source
428
+ − 2059 It is important to put the open-brace that starts the body of a C
+ − 2060 function in column zero, and avoid putting any other open-brace or
+ − 2061 open-parenthesis or open-bracket in column zero. Several tools look
+ − 2062 for open-braces in column zero to find the beginnings of C functions.
+ − 2063 These tools will not work on code not formatted that way.
+ − 2064
+ − 2065 It is also important for function definitions to start the name of the
+ − 2066 function in column zero. This helps people to search for function
+ − 2067 definitions, and may also help certain tools recognize them. Thus,
+ − 2068 the proper format is this:
+ − 2069
+ − 2070 @example
+ − 2071 static char *
+ − 2072 concat (s1, s2) /* Name starts in column zero here */
+ − 2073 char *s1, *s2;
+ − 2074 @{ /* Open brace in column zero here */
+ − 2075 @dots{}
+ − 2076 @}
+ − 2077 @end example
+ − 2078
+ − 2079 @noindent
462
+ − 2080 or, if you want to use Standard C syntax, format the definition like
+ − 2081 this:
428
+ − 2082
+ − 2083 @example
+ − 2084 static char *
+ − 2085 concat (char *s1, char *s2)
+ − 2086 @{
+ − 2087 @dots{}
+ − 2088 @}
+ − 2089 @end example
+ − 2090
462
+ − 2091 In Standard C, if the arguments don't fit nicely on one line,
428
+ − 2092 split it like this:
+ − 2093
+ − 2094 @example
+ − 2095 int
+ − 2096 lots_of_args (int an_integer, long a_long, short a_short,
+ − 2097 double a_double, float a_float)
+ − 2098 @dots{}
+ − 2099 @end example
+ − 2100
462
+ − 2101 The rest of this section gives our recommendations for other aspects of
+ − 2102 C formatting style, which is also the default style of the @code{indent}
+ − 2103 program in version 1.2 and newer. It corresponds to the options
+ − 2104
+ − 2105 @smallexample
+ − 2106 -nbad -bap -nbc -bbo -bl -bli2 -bls -ncdb -nce -cp1 -cs -di2
+ − 2107 -ndj -nfc1 -nfca -hnl -i2 -ip5 -lp -pcs -psl -nsc -nsob
+ − 2108 @end smallexample
+ − 2109
+ − 2110 We don't think of these recommendations as requirements, because it
+ − 2111 causes no problems for users if two different programs have different
+ − 2112 formatting styles.
+ − 2113
+ − 2114 But whatever style you use, please use it consistently, since a mixture
+ − 2115 of styles within one program tends to look ugly. If you are
+ − 2116 contributing changes to an existing program, please follow the style of
+ − 2117 that program.
+ − 2118
+ − 2119 For the body of the function, our recommended style looks like this:
428
+ − 2120
+ − 2121 @example
+ − 2122 if (x < foo (y, z))
+ − 2123 haha = bar[4] + 5;
+ − 2124 else
+ − 2125 @{
+ − 2126 while (z)
+ − 2127 @{
+ − 2128 haha += foo (z, z);
+ − 2129 z--;
+ − 2130 @}
+ − 2131 return ++x + bar ();
+ − 2132 @}
+ − 2133 @end example
+ − 2134
462
+ − 2135 @cindex spaces before open-paren
428
+ − 2136 We find it easier to read a program when it has spaces before the
+ − 2137 open-parentheses and after the commas. Especially after the commas.
+ − 2138
+ − 2139 When you split an expression into multiple lines, split it
+ − 2140 before an operator, not after one. Here is the right way:
+ − 2141
462
+ − 2142 @cindex expressions, splitting
428
+ − 2143 @example
+ − 2144 if (foo_this_is_long && bar > win (x, y, z)
+ − 2145 && remaining_condition)
+ − 2146 @end example
+ − 2147
+ − 2148 Try to avoid having two operators of different precedence at the same
+ − 2149 level of indentation. For example, don't write this:
+ − 2150
+ − 2151 @example
+ − 2152 mode = (inmode[j] == VOIDmode
+ − 2153 || GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])
+ − 2154 ? outmode[j] : inmode[j]);
+ − 2155 @end example
+ − 2156
+ − 2157 Instead, use extra parentheses so that the indentation shows the nesting:
+ − 2158
+ − 2159 @example
+ − 2160 mode = ((inmode[j] == VOIDmode
+ − 2161 || (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])))
+ − 2162 ? outmode[j] : inmode[j]);
+ − 2163 @end example
+ − 2164
+ − 2165 Insert extra parentheses so that Emacs will indent the code properly.
+ − 2166 For example, the following indentation looks nice if you do it by hand,
+ − 2167
+ − 2168 @example
+ − 2169 v = rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000
+ − 2170 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000;
+ − 2171 @end example
+ − 2172
462
+ − 2173 @noindent
+ − 2174 but Emacs would alter it. Adding a set of parentheses produces
+ − 2175 something that looks equally nice, and which Emacs will preserve:
428
+ − 2176
+ − 2177 @example
+ − 2178 v = (rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000
+ − 2179 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000);
+ − 2180 @end example
+ − 2181
+ − 2182 Format do-while statements like this:
+ − 2183
+ − 2184 @example
+ − 2185 do
+ − 2186 @{
+ − 2187 a = foo (a);
+ − 2188 @}
+ − 2189 while (a > 0);
+ − 2190 @end example
+ − 2191
462
+ − 2192 @cindex formfeed
+ − 2193 @cindex control-L
428
+ − 2194 Please use formfeed characters (control-L) to divide the program into
+ − 2195 pages at logical places (but not within a function). It does not matter
+ − 2196 just how long the pages are, since they do not have to fit on a printed
+ − 2197 page. The formfeeds should appear alone on lines by themselves.
+ − 2198
+ − 2199 @node Comments
+ − 2200 @section Commenting Your Work
462
+ − 2201 @cindex commenting
428
+ − 2202
+ − 2203 Every program should start with a comment saying briefly what it is for.
+ − 2204 Example: @samp{fmt - filter for simple filling of text}.
+ − 2205
+ − 2206 Please write the comments in a GNU program in English, because English
+ − 2207 is the one language that nearly all programmers in all countries can
+ − 2208 read. If you do not write English well, please write comments in
+ − 2209 English as well as you can, then ask other people to help rewrite them.
+ − 2210 If you can't write comments in English, please find someone to work with
+ − 2211 you and translate your comments into English.
+ − 2212
+ − 2213 Please put a comment on each function saying what the function does,
+ − 2214 what sorts of arguments it gets, and what the possible values of
+ − 2215 arguments mean and are used for. It is not necessary to duplicate in
+ − 2216 words the meaning of the C argument declarations, if a C type is being
+ − 2217 used in its customary fashion. If there is anything nonstandard about
+ − 2218 its use (such as an argument of type @code{char *} which is really the
+ − 2219 address of the second character of a string, not the first), or any
+ − 2220 possible values that would not work the way one would expect (such as,
+ − 2221 that strings containing newlines are not guaranteed to work), be sure
+ − 2222 to say so.
+ − 2223
+ − 2224 Also explain the significance of the return value, if there is one.
+ − 2225
+ − 2226 Please put two spaces after the end of a sentence in your comments, so
+ − 2227 that the Emacs sentence commands will work. Also, please write
+ − 2228 complete sentences and capitalize the first word. If a lower-case
+ − 2229 identifier comes at the beginning of a sentence, don't capitalize it!
+ − 2230 Changing the spelling makes it a different identifier. If you don't
+ − 2231 like starting a sentence with a lower case letter, write the sentence
+ − 2232 differently (e.g., ``The identifier lower-case is @dots{}'').
+ − 2233
+ − 2234 The comment on a function is much clearer if you use the argument
+ − 2235 names to speak about the argument values. The variable name itself
+ − 2236 should be lower case, but write it in upper case when you are speaking
+ − 2237 about the value rather than the variable itself. Thus, ``the inode
+ − 2238 number NODE_NUM'' rather than ``an inode''.
+ − 2239
+ − 2240 There is usually no purpose in restating the name of the function in
+ − 2241 the comment before it, because the reader can see that for himself.
+ − 2242 There might be an exception when the comment is so long that the function
+ − 2243 itself would be off the bottom of the screen.
+ − 2244
+ − 2245 There should be a comment on each static variable as well, like this:
+ − 2246
+ − 2247 @example
+ − 2248 /* Nonzero means truncate lines in the display;
+ − 2249 zero means continue them. */
+ − 2250 int truncate_lines;
+ − 2251 @end example
+ − 2252
462
+ − 2253 @cindex conditionals, comments for
+ − 2254 @cindex @code{#endif}, commenting
428
+ − 2255 Every @samp{#endif} should have a comment, except in the case of short
+ − 2256 conditionals (just a few lines) that are not nested. The comment should
+ − 2257 state the condition of the conditional that is ending, @emph{including
+ − 2258 its sense}. @samp{#else} should have a comment describing the condition
+ − 2259 @emph{and sense} of the code that follows. For example:
+ − 2260
+ − 2261 @example
+ − 2262 @group
+ − 2263 #ifdef foo
+ − 2264 @dots{}
+ − 2265 #else /* not foo */
+ − 2266 @dots{}
+ − 2267 #endif /* not foo */
+ − 2268 @end group
+ − 2269 @group
+ − 2270 #ifdef foo
+ − 2271 @dots{}
+ − 2272 #endif /* foo */
+ − 2273 @end group
+ − 2274 @end example
+ − 2275
+ − 2276 @noindent
+ − 2277 but, by contrast, write the comments this way for a @samp{#ifndef}:
+ − 2278
+ − 2279 @example
+ − 2280 @group
+ − 2281 #ifndef foo
+ − 2282 @dots{}
+ − 2283 #else /* foo */
+ − 2284 @dots{}
+ − 2285 #endif /* foo */
+ − 2286 @end group
+ − 2287 @group
+ − 2288 #ifndef foo
+ − 2289 @dots{}
+ − 2290 #endif /* not foo */
+ − 2291 @end group
+ − 2292 @end example
+ − 2293
+ − 2294 @node Syntactic Conventions
+ − 2295 @section Clean Use of C Constructs
462
+ − 2296 @cindex syntactic conventions
+ − 2297
+ − 2298 @cindex implicit @code{int}
+ − 2299 @cindex function argument, declaring
+ − 2300 Please explicitly declare the types of all objects. For example, you
+ − 2301 should explicitly declare all arguments to functions, and you should
+ − 2302 declare functions to return @code{int} rather than omitting the
+ − 2303 @code{int}.
+ − 2304
+ − 2305 @cindex compiler warnings
+ − 2306 @cindex @samp{-Wall} compiler option
+ − 2307 Some programmers like to use the GCC @samp{-Wall} option, and change the
+ − 2308 code whenever it issues a warning. If you want to do this, then do.
+ − 2309 Other programmers prefer not to use @samp{-Wall}, because it gives
+ − 2310 warnings for valid and legitimate code which they do not want to change.
+ − 2311 If you want to do this, then do. The compiler should be your servant,
+ − 2312 not your master.
428
+ − 2313
+ − 2314 Declarations of external functions and functions to appear later in the
+ − 2315 source file should all go in one place near the beginning of the file
+ − 2316 (somewhere before the first function definition in the file), or else
+ − 2317 should go in a header file. Don't put @code{extern} declarations inside
+ − 2318 functions.
+ − 2319
462
+ − 2320 @cindex temporary variables
428
+ − 2321 It used to be common practice to use the same local variables (with
+ − 2322 names like @code{tem}) over and over for different values within one
+ − 2323 function. Instead of doing this, it is better declare a separate local
+ − 2324 variable for each distinct purpose, and give it a name which is
+ − 2325 meaningful. This not only makes programs easier to understand, it also
+ − 2326 facilitates optimization by good compilers. You can also move the
+ − 2327 declaration of each local variable into the smallest scope that includes
+ − 2328 all its uses. This makes the program even cleaner.
+ − 2329
+ − 2330 Don't use local variables or parameters that shadow global identifiers.
+ − 2331
462
+ − 2332 @cindex multiple variables in a line
428
+ − 2333 Don't declare multiple variables in one declaration that spans lines.
+ − 2334 Start a new declaration on each line, instead. For example, instead
+ − 2335 of this:
+ − 2336
+ − 2337 @example
+ − 2338 @group
+ − 2339 int foo,
+ − 2340 bar;
+ − 2341 @end group
+ − 2342 @end example
+ − 2343
+ − 2344 @noindent
+ − 2345 write either this:
+ − 2346
+ − 2347 @example
+ − 2348 int foo, bar;
+ − 2349 @end example
+ − 2350
+ − 2351 @noindent
+ − 2352 or this:
+ − 2353
+ − 2354 @example
+ − 2355 int foo;
+ − 2356 int bar;
+ − 2357 @end example
+ − 2358
+ − 2359 @noindent
+ − 2360 (If they are global variables, each should have a comment preceding it
+ − 2361 anyway.)
+ − 2362
+ − 2363 When you have an @code{if}-@code{else} statement nested in another
+ − 2364 @code{if} statement, always put braces around the @code{if}-@code{else}.
+ − 2365 Thus, never write like this:
+ − 2366
+ − 2367 @example
+ − 2368 if (foo)
+ − 2369 if (bar)
+ − 2370 win ();
+ − 2371 else
+ − 2372 lose ();
+ − 2373 @end example
+ − 2374
+ − 2375 @noindent
+ − 2376 always like this:
+ − 2377
+ − 2378 @example
+ − 2379 if (foo)
+ − 2380 @{
+ − 2381 if (bar)
+ − 2382 win ();
+ − 2383 else
+ − 2384 lose ();
+ − 2385 @}
+ − 2386 @end example
+ − 2387
+ − 2388 If you have an @code{if} statement nested inside of an @code{else}
+ − 2389 statement, either write @code{else if} on one line, like this,
+ − 2390
+ − 2391 @example
+ − 2392 if (foo)
+ − 2393 @dots{}
+ − 2394 else if (bar)
+ − 2395 @dots{}
+ − 2396 @end example
+ − 2397
+ − 2398 @noindent
+ − 2399 with its @code{then}-part indented like the preceding @code{then}-part,
+ − 2400 or write the nested @code{if} within braces like this:
+ − 2401
+ − 2402 @example
+ − 2403 if (foo)
+ − 2404 @dots{}
+ − 2405 else
+ − 2406 @{
+ − 2407 if (bar)
+ − 2408 @dots{}
+ − 2409 @}
+ − 2410 @end example
+ − 2411
+ − 2412 Don't declare both a structure tag and variables or typedefs in the
+ − 2413 same declaration. Instead, declare the structure tag separately
+ − 2414 and then use it to declare the variables or typedefs.
+ − 2415
+ − 2416 Try to avoid assignments inside @code{if}-conditions. For example,
+ − 2417 don't write this:
+ − 2418
+ − 2419 @example
+ − 2420 if ((foo = (char *) malloc (sizeof *foo)) == 0)
+ − 2421 fatal ("virtual memory exhausted");
+ − 2422 @end example
+ − 2423
+ − 2424 @noindent
+ − 2425 instead, write this:
+ − 2426
+ − 2427 @example
+ − 2428 foo = (char *) malloc (sizeof *foo);
+ − 2429 if (foo == 0)
+ − 2430 fatal ("virtual memory exhausted");
+ − 2431 @end example
+ − 2432
462
+ − 2433 @pindex lint
428
+ − 2434 Don't make the program ugly to placate @code{lint}. Please don't insert any
+ − 2435 casts to @code{void}. Zero without a cast is perfectly fine as a null
+ − 2436 pointer constant, except when calling a varargs function.
+ − 2437
+ − 2438 @node Names
+ − 2439 @section Naming Variables and Functions
+ − 2440
462
+ − 2441 @cindex names of variables and functions
428
+ − 2442 The names of global variables and functions in a program serve as
+ − 2443 comments of a sort. So don't choose terse names---instead, look for
+ − 2444 names that give useful information about the meaning of the variable or
+ − 2445 function. In a GNU program, names should be English, like other
+ − 2446 comments.
+ − 2447
+ − 2448 Local variable names can be shorter, because they are used only within
+ − 2449 one context, where (presumably) comments explain their purpose.
+ − 2450
+ − 2451 Try to limit your use of abbreviations in symbol names. It is ok to
+ − 2452 make a few abbreviations, explain what they mean, and then use them
+ − 2453 frequently, but don't use lots of obscure abbreviations.
+ − 2454
+ − 2455 Please use underscores to separate words in a name, so that the Emacs
+ − 2456 word commands can be useful within them. Stick to lower case; reserve
+ − 2457 upper case for macros and @code{enum} constants, and for name-prefixes
+ − 2458 that follow a uniform convention.
+ − 2459
+ − 2460 For example, you should use names like @code{ignore_space_change_flag};
+ − 2461 don't use names like @code{iCantReadThis}.
+ − 2462
+ − 2463 Variables that indicate whether command-line options have been
+ − 2464 specified should be named after the meaning of the option, not after
+ − 2465 the option-letter. A comment should state both the exact meaning of
+ − 2466 the option and its letter. For example,
+ − 2467
+ − 2468 @example
+ − 2469 @group
+ − 2470 /* Ignore changes in horizontal whitespace (-b). */
+ − 2471 int ignore_space_change_flag;
+ − 2472 @end group
+ − 2473 @end example
+ − 2474
+ − 2475 When you want to define names with constant integer values, use
+ − 2476 @code{enum} rather than @samp{#define}. GDB knows about enumeration
+ − 2477 constants.
+ − 2478
462
+ − 2479 @cindex file-name limitations
+ − 2480 @pindex doschk
+ − 2481 You might want to make sure that none of the file names would conflict
+ − 2482 the files were loaded onto an MS-DOS file system which shortens the
+ − 2483 names. You can use the program @code{doschk} to test for this.
+ − 2484
+ − 2485 Some GNU programs were designed to limit themselves to file names of 14
+ − 2486 characters or less, to avoid file name conflicts if they are read into
+ − 2487 older System V systems. Please preserve this feature in the existing
+ − 2488 GNU programs that have it, but there is no need to do this in new GNU
+ − 2489 programs. @code{doschk} also reports file names longer than 14
+ − 2490 characters.
428
+ − 2491
+ − 2492 @node System Portability
+ − 2493 @section Portability between System Types
462
+ − 2494 @cindex portability, between system types
428
+ − 2495
+ − 2496 In the Unix world, ``portability'' refers to porting to different Unix
+ − 2497 versions. For a GNU program, this kind of portability is desirable, but
+ − 2498 not paramount.
+ − 2499
+ − 2500 The primary purpose of GNU software is to run on top of the GNU kernel,
462
+ − 2501 compiled with the GNU C compiler, on various types of @sc{cpu}. So the
+ − 2502 kinds of portability that are absolutely necessary are quite limited.
+ − 2503 But it is important to support Linux-based GNU systems, since they
+ − 2504 are the form of GNU that is popular.
+ − 2505
+ − 2506 Beyond that, it is good to support the other free operating systems
+ − 2507 (*BSD), and it is nice to support other Unix-like systems if you want
+ − 2508 to. Supporting a variety of Unix-like systems is desirable, although
+ − 2509 not paramount. It is usually not too hard, so you may as well do it.
+ − 2510 But you don't have to consider it an obligation, if it does turn out to
+ − 2511 be hard.
+ − 2512
+ − 2513 @pindex autoconf
428
+ − 2514 The easiest way to achieve portability to most Unix-like systems is to
+ − 2515 use Autoconf. It's unlikely that your program needs to know more
+ − 2516 information about the host platform than Autoconf can provide, simply
+ − 2517 because most of the programs that need such knowledge have already been
+ − 2518 written.
+ − 2519
+ − 2520 Avoid using the format of semi-internal data bases (e.g., directories)
+ − 2521 when there is a higher-level alternative (@code{readdir}).
+ − 2522
462
+ − 2523 @cindex non-@sc{posix} systems, and portability
428
+ − 2524 As for systems that are not like Unix, such as MSDOS, Windows, the
+ − 2525 Macintosh, VMS, and MVS, supporting them is often a lot of work. When
+ − 2526 that is the case, it is better to spend your time adding features that
+ − 2527 will be useful on GNU and GNU/Linux, rather than on supporting other
+ − 2528 incompatible systems.
+ − 2529
462
+ − 2530 It is a good idea to define the ``feature test macro''
+ − 2531 @code{_GNU_SOURCE} when compiling your C files. When you compile on GNU
+ − 2532 or GNU/Linux, this will enable the declarations of GNU library extension
+ − 2533 functions, and that will usually give you a compiler error message if
+ − 2534 you define the same function names in some other way in your program.
+ − 2535 (You don't have to actually @emph{use} these functions, if you prefer
+ − 2536 to make the program more portable to other systems.)
+ − 2537
+ − 2538 But whether or not you use these GNU extensions, you should avoid
+ − 2539 using their names for any other meanings. Doing so would make it hard
+ − 2540 to move your code into other GNU programs.
+ − 2541
428
+ − 2542 @node CPU Portability
+ − 2543 @section Portability between @sc{cpu}s
+ − 2544
462
+ − 2545 @cindex data types, and portability
+ − 2546 @cindex portability, and data types
428
+ − 2547 Even GNU systems will differ because of differences among @sc{cpu}
+ − 2548 types---for example, difference in byte ordering and alignment
+ − 2549 requirements. It is absolutely essential to handle these differences.
+ − 2550 However, don't make any effort to cater to the possibility that an
+ − 2551 @code{int} will be less than 32 bits. We don't support 16-bit machines
+ − 2552 in GNU.
+ − 2553
462
+ − 2554 Similarly, don't make any effort to cater to the possibility that
+ − 2555 @code{long} will be smaller than predefined types like @code{size_t}.
+ − 2556 For example, the following code is ok:
+ − 2557
+ − 2558 @example
+ − 2559 printf ("size = %lu\n", (unsigned long) sizeof array);
+ − 2560 printf ("diff = %ld\n", (long) (pointer2 - pointer1));
+ − 2561 @end example
+ − 2562
+ − 2563 1989 Standard C requires this to work, and we know of only one
+ − 2564 counterexample: 64-bit programs on Microsoft Windows IA-64. We will
+ − 2565 leave it to those who want to port GNU programs to that environment
+ − 2566 to figure out how to do it.
+ − 2567
+ − 2568 Predefined file-size types like @code{off_t} are an exception: they are
+ − 2569 longer than @code{long} on many platforms, so code like the above won't
+ − 2570 work with them. One way to print an @code{off_t} value portably is to
+ − 2571 print its digits yourself, one by one.
+ − 2572
428
+ − 2573 Don't assume that the address of an @code{int} object is also the
+ − 2574 address of its least-significant byte. This is false on big-endian
+ − 2575 machines. Thus, don't make the following mistake:
+ − 2576
+ − 2577 @example
+ − 2578 int c;
+ − 2579 @dots{}
+ − 2580 while ((c = getchar()) != EOF)
+ − 2581 write(file_descriptor, &c, 1);
+ − 2582 @end example
+ − 2583
+ − 2584 When calling functions, you need not worry about the difference between
+ − 2585 pointers of various types, or between pointers and integers. On most
+ − 2586 machines, there's no difference anyway. As for the few machines where
462
+ − 2587 there is a difference, all of them support Standard C prototypes, so you can
+ − 2588 use prototypes (perhaps conditionalized to be active only in Standard C)
+ − 2589 to make the code work on those systems.
428
+ − 2590
+ − 2591 In certain cases, it is ok to pass integer and pointer arguments
+ − 2592 indiscriminately to the same function, and use no prototype on any
+ − 2593 system. For example, many GNU programs have error-reporting functions
+ − 2594 that pass their arguments along to @code{printf} and friends:
+ − 2595
+ − 2596 @example
+ − 2597 error (s, a1, a2, a3)
+ − 2598 char *s;
+ − 2599 char *a1, *a2, *a3;
+ − 2600 @{
+ − 2601 fprintf (stderr, "error: ");
+ − 2602 fprintf (stderr, s, a1, a2, a3);
+ − 2603 @}
+ − 2604 @end example
+ − 2605
+ − 2606 @noindent
+ − 2607 In practice, this works on all machines, since a pointer is generally
462
+ − 2608 the widest possible kind of argument; it is much simpler than any
428
+ − 2609 ``correct'' alternative. Be sure @emph{not} to use a prototype for such
+ − 2610 functions.
+ − 2611
462
+ − 2612 If you have decided to use Standard C, then you can instead define
+ − 2613 @code{error} using @file{stdarg.h}, and pass the arguments along to
+ − 2614 @code{vfprintf}.
+ − 2615
+ − 2616 @cindex casting pointers to integers
+ − 2617 Avoid casting pointers to integers if you can. Such casts greatly
+ − 2618 reduce portability, and in most programs they are easy to avoid. In the
+ − 2619 cases where casting pointers to integers is essential---such as, a Lisp
+ − 2620 interpreter which stores type information as well as an address in one
+ − 2621 word---you'll have to make explicit provisions to handle different word
+ − 2622 sizes. You will also need to make provision for systems in which the
+ − 2623 normal range of addresses you can get from @code{malloc} starts far away
+ − 2624 from zero.
428
+ − 2625
+ − 2626 @node System Functions
+ − 2627 @section Calling System Functions
462
+ − 2628 @cindex library functions, and portability
+ − 2629 @cindex portability, and library functions
+ − 2630
+ − 2631 C implementations differ substantially. Standard C reduces but does
+ − 2632 not eliminate the incompatibilities; meanwhile, many GNU packages still
+ − 2633 support pre-standard compilers because this is not hard to do. This
+ − 2634 chapter gives recommendations for how to use the more-or-less standard C
+ − 2635 library functions to avoid unnecessary loss of portability.
428
+ − 2636
+ − 2637 @itemize @bullet
+ − 2638 @item
462
+ − 2639 Don't use the return value of @code{sprintf}. It returns the number of
428
+ − 2640 characters written on some systems, but not on all systems.
+ − 2641
+ − 2642 @item
462
+ − 2643 Be aware that @code{vfprintf} is not always available.
+ − 2644
+ − 2645 @item
428
+ − 2646 @code{main} should be declared to return type @code{int}. It should
+ − 2647 terminate either by calling @code{exit} or by returning the integer
+ − 2648 status code; make sure it cannot ever return an undefined value.
+ − 2649
462
+ − 2650 @cindex declaration for system functions
428
+ − 2651 @item
+ − 2652 Don't declare system functions explicitly.
+ − 2653
+ − 2654 Almost any declaration for a system function is wrong on some system.
+ − 2655 To minimize conflicts, leave it to the system header files to declare
+ − 2656 system functions. If the headers don't declare a function, let it
+ − 2657 remain undeclared.
+ − 2658
+ − 2659 While it may seem unclean to use a function without declaring it, in
+ − 2660 practice this works fine for most system library functions on the
+ − 2661 systems where this really happens; thus, the disadvantage is only
+ − 2662 theoretical. By contrast, actual declarations have frequently caused
+ − 2663 actual conflicts.
+ − 2664
+ − 2665 @item
+ − 2666 If you must declare a system function, don't specify the argument types.
462
+ − 2667 Use an old-style declaration, not a Standard C prototype. The more you
428
+ − 2668 specify about the function, the more likely a conflict.
+ − 2669
+ − 2670 @item
+ − 2671 In particular, don't unconditionally declare @code{malloc} or
+ − 2672 @code{realloc}.
+ − 2673
+ − 2674 Most GNU programs use those functions just once, in functions
+ − 2675 conventionally named @code{xmalloc} and @code{xrealloc}. These
+ − 2676 functions call @code{malloc} and @code{realloc}, respectively, and
+ − 2677 check the results.
+ − 2678
+ − 2679 Because @code{xmalloc} and @code{xrealloc} are defined in your program,
+ − 2680 you can declare them in other files without any risk of type conflict.
+ − 2681
+ − 2682 On most systems, @code{int} is the same length as a pointer; thus, the
+ − 2683 calls to @code{malloc} and @code{realloc} work fine. For the few
+ − 2684 exceptional systems (mostly 64-bit machines), you can use
+ − 2685 @strong{conditionalized} declarations of @code{malloc} and
+ − 2686 @code{realloc}---or put these declarations in configuration files
+ − 2687 specific to those systems.
+ − 2688
462
+ − 2689 @cindex string library functions
428
+ − 2690 @item
+ − 2691 The string functions require special treatment. Some Unix systems have
+ − 2692 a header file @file{string.h}; others have @file{strings.h}. Neither
+ − 2693 file name is portable. There are two things you can do: use Autoconf to
+ − 2694 figure out which file to include, or don't include either file.
+ − 2695
+ − 2696 @item
+ − 2697 If you don't include either strings file, you can't get declarations for
+ − 2698 the string functions from the header file in the usual way.
+ − 2699
462
+ − 2700 That causes less of a problem than you might think. The newer standard
428
+ − 2701 string functions should be avoided anyway because many systems still
+ − 2702 don't support them. The string functions you can use are these:
+ − 2703
+ − 2704 @example
+ − 2705 strcpy strncpy strcat strncat
+ − 2706 strlen strcmp strncmp
+ − 2707 strchr strrchr
+ − 2708 @end example
+ − 2709
+ − 2710 The copy and concatenate functions work fine without a declaration as
+ − 2711 long as you don't use their values. Using their values without a
+ − 2712 declaration fails on systems where the width of a pointer differs from
+ − 2713 the width of @code{int}, and perhaps in other cases. It is trivial to
+ − 2714 avoid using their values, so do that.
+ − 2715
+ − 2716 The compare functions and @code{strlen} work fine without a declaration
+ − 2717 on most systems, possibly all the ones that GNU software runs on.
+ − 2718 You may find it necessary to declare them @strong{conditionally} on a
+ − 2719 few systems.
+ − 2720
+ − 2721 The search functions must be declared to return @code{char *}. Luckily,
+ − 2722 there is no variation in the data type they return. But there is
+ − 2723 variation in their names. Some systems give these functions the names
+ − 2724 @code{index} and @code{rindex}; other systems use the names
+ − 2725 @code{strchr} and @code{strrchr}. Some systems support both pairs of
+ − 2726 names, but neither pair works on all systems.
+ − 2727
+ − 2728 You should pick a single pair of names and use it throughout your
+ − 2729 program. (Nowadays, it is better to choose @code{strchr} and
462
+ − 2730 @code{strrchr} for new programs, since those are the standard
428
+ − 2731 names.) Declare both of those names as functions returning @code{char
+ − 2732 *}. On systems which don't support those names, define them as macros
+ − 2733 in terms of the other pair. For example, here is what to put at the
+ − 2734 beginning of your file (or in a header) if you want to use the names
+ − 2735 @code{strchr} and @code{strrchr} throughout:
+ − 2736
+ − 2737 @example
+ − 2738 #ifndef HAVE_STRCHR
+ − 2739 #define strchr index
+ − 2740 #endif
+ − 2741 #ifndef HAVE_STRRCHR
+ − 2742 #define strrchr rindex
+ − 2743 #endif
+ − 2744
+ − 2745 char *strchr ();
+ − 2746 char *strrchr ();
+ − 2747 @end example
+ − 2748 @end itemize
+ − 2749
+ − 2750 Here we assume that @code{HAVE_STRCHR} and @code{HAVE_STRRCHR} are
+ − 2751 macros defined in systems where the corresponding functions exist.
+ − 2752 One way to get them properly defined is to use Autoconf.
+ − 2753
+ − 2754 @node Internationalization
+ − 2755 @section Internationalization
462
+ − 2756 @cindex internationalization
+ − 2757
+ − 2758 @pindex gettext
428
+ − 2759 GNU has a library called GNU gettext that makes it easy to translate the
+ − 2760 messages in a program into various languages. You should use this
+ − 2761 library in every program. Use English for the messages as they appear
+ − 2762 in the program, and let gettext provide the way to translate them into
+ − 2763 other languages.
+ − 2764
+ − 2765 Using GNU gettext involves putting a call to the @code{gettext} macro
+ − 2766 around each string that might need translation---like this:
+ − 2767
+ − 2768 @example
+ − 2769 printf (gettext ("Processing file `%s'..."));
+ − 2770 @end example
+ − 2771
+ − 2772 @noindent
+ − 2773 This permits GNU gettext to replace the string @code{"Processing file
+ − 2774 `%s'..."} with a translated version.
+ − 2775
+ − 2776 Once a program uses gettext, please make a point of writing calls to
+ − 2777 @code{gettext} when you add new strings that call for translation.
+ − 2778
+ − 2779 Using GNU gettext in a package involves specifying a @dfn{text domain
+ − 2780 name} for the package. The text domain name is used to separate the
+ − 2781 translations for this package from the translations for other packages.
+ − 2782 Normally, the text domain name should be the same as the name of the
+ − 2783 package---for example, @samp{fileutils} for the GNU file utilities.
+ − 2784
462
+ − 2785 @cindex message text, and internationalization
428
+ − 2786 To enable gettext to work well, avoid writing code that makes
+ − 2787 assumptions about the structure of words or sentences. When you want
+ − 2788 the precise text of a sentence to vary depending on the data, use two or
+ − 2789 more alternative string constants each containing a complete sentences,
+ − 2790 rather than inserting conditionalized words or phrases into a single
+ − 2791 sentence framework.
+ − 2792
+ − 2793 Here is an example of what not to do:
+ − 2794
+ − 2795 @example
+ − 2796 printf ("%d file%s processed", nfiles,
+ − 2797 nfiles != 1 ? "s" : "");
+ − 2798 @end example
+ − 2799
+ − 2800 @noindent
+ − 2801 The problem with that example is that it assumes that plurals are made
+ − 2802 by adding `s'. If you apply gettext to the format string, like this,
+ − 2803
+ − 2804 @example
+ − 2805 printf (gettext ("%d file%s processed"), nfiles,
+ − 2806 nfiles != 1 ? "s" : "");
+ − 2807 @end example
+ − 2808
+ − 2809 @noindent
+ − 2810 the message can use different words, but it will still be forced to use
+ − 2811 `s' for the plural. Here is a better way:
+ − 2812
+ − 2813 @example
+ − 2814 printf ((nfiles != 1 ? "%d files processed"
+ − 2815 : "%d file processed"),
+ − 2816 nfiles);
+ − 2817 @end example
+ − 2818
+ − 2819 @noindent
+ − 2820 This way, you can apply gettext to each of the two strings
+ − 2821 independently:
+ − 2822
+ − 2823 @example
+ − 2824 printf ((nfiles != 1 ? gettext ("%d files processed")
+ − 2825 : gettext ("%d file processed")),
+ − 2826 nfiles);
+ − 2827 @end example
+ − 2828
+ − 2829 @noindent
+ − 2830 This can be any method of forming the plural of the word for ``file'', and
+ − 2831 also handles languages that require agreement in the word for
+ − 2832 ``processed''.
+ − 2833
+ − 2834 A similar problem appears at the level of sentence structure with this
+ − 2835 code:
+ − 2836
+ − 2837 @example
+ − 2838 printf ("# Implicit rule search has%s been done.\n",
+ − 2839 f->tried_implicit ? "" : " not");
+ − 2840 @end example
+ − 2841
+ − 2842 @noindent
+ − 2843 Adding @code{gettext} calls to this code cannot give correct results for
+ − 2844 all languages, because negation in some languages requires adding words
+ − 2845 at more than one place in the sentence. By contrast, adding
+ − 2846 @code{gettext} calls does the job straightfowardly if the code starts
+ − 2847 out like this:
+ − 2848
+ − 2849 @example
+ − 2850 printf (f->tried_implicit
+ − 2851 ? "# Implicit rule search has been done.\n",
+ − 2852 : "# Implicit rule search has not been done.\n");
+ − 2853 @end example
+ − 2854
+ − 2855 @node Mmap
+ − 2856 @section Mmap
462
+ − 2857 @findex mmap
428
+ − 2858
+ − 2859 Don't assume that @code{mmap} either works on all files or fails
+ − 2860 for all files. It may work on some files and fail on others.
+ − 2861
+ − 2862 The proper way to use @code{mmap} is to try it on the specific file for
+ − 2863 which you want to use it---and if @code{mmap} doesn't work, fall back on
+ − 2864 doing the job in another way using @code{read} and @code{write}.
+ − 2865
+ − 2866 The reason this precaution is needed is that the GNU kernel (the HURD)
+ − 2867 provides a user-extensible file system, in which there can be many
+ − 2868 different kinds of ``ordinary files.'' Many of them support
+ − 2869 @code{mmap}, but some do not. It is important to make programs handle
+ − 2870 all these kinds of files.
+ − 2871
+ − 2872 @node Documentation
+ − 2873 @chapter Documenting Programs
462
+ − 2874 @cindex documentation
+ − 2875
+ − 2876 A GNU program should ideally come with full free documentation, adequate
+ − 2877 for both reference and tutorial purposes. If the package can be
+ − 2878 programmed or extended, the documentation should cover programming or
+ − 2879 extending it, as well as just using it.
428
+ − 2880
+ − 2881 @menu
+ − 2882 * GNU Manuals:: Writing proper manuals.
462
+ − 2883 * Doc Strings and Manuals:: Compiling doc strings doesn't make a manual.
428
+ − 2884 * Manual Structure Details:: Specific structure conventions.
+ − 2885 * License for Manuals:: Writing the distribution terms for a manual.
462
+ − 2886 * Manual Credits:: Giving credit to documentation contributors.
+ − 2887 * Printed Manuals:: Mentioning the printed manual.
428
+ − 2888 * NEWS File:: NEWS files supplement manuals.
+ − 2889 * Change Logs:: Recording Changes
+ − 2890 * Man Pages:: Man pages are secondary.
+ − 2891 * Reading other Manuals:: How far you can go in learning
+ − 2892 from other manuals.
+ − 2893 @end menu
+ − 2894
+ − 2895 @node GNU Manuals
+ − 2896 @section GNU Manuals
+ − 2897
462
+ − 2898 The preferred document format for the GNU system is the Texinfo
+ − 2899 formatting language. Every GNU package should (ideally) have
+ − 2900 documentation in Texinfo both for reference and for learners. Texinfo
+ − 2901 makes it possible to produce a good quality formatted book, using
+ − 2902 @TeX{}, and to generate an Info file. It is also possible to generate
+ − 2903 HTML output from Texinfo source. See the Texinfo manual, either the
+ − 2904 hardcopy, or the on-line version available through @code{info} or the
+ − 2905 Emacs Info subsystem (@kbd{C-h i}).
+ − 2906
+ − 2907 Nowadays some other formats such as Docbook and Sgmltexi can be
+ − 2908 converted automatically into Texinfo. It is ok to produce the Texinfo
+ − 2909 documentation by conversion this way, as long as it gives good results.
428
+ − 2910
+ − 2911 Programmers often find it most natural to structure the documentation
+ − 2912 following the structure of the implementation, which they know. But
+ − 2913 this structure is not necessarily good for explaining how to use the
+ − 2914 program; it may be irrelevant and confusing for a user.
+ − 2915
+ − 2916 At every level, from the sentences in a paragraph to the grouping of
+ − 2917 topics into separate manuals, the right way to structure documentation
+ − 2918 is according to the concepts and questions that a user will have in mind
+ − 2919 when reading it. Sometimes this structure of ideas matches the
+ − 2920 structure of the implementation of the software being documented---but
+ − 2921 often they are different. Often the most important part of learning to
+ − 2922 write good documentation is learning to notice when you are structuring
+ − 2923 the documentation like the implementation, and think about better
+ − 2924 alternatives.
+ − 2925
+ − 2926 For example, each program in the GNU system probably ought to be
+ − 2927 documented in one manual; but this does not mean each program should
+ − 2928 have its own manual. That would be following the structure of the
+ − 2929 implementation, rather than the structure that helps the user
+ − 2930 understand.
+ − 2931
+ − 2932 Instead, each manual should cover a coherent @emph{topic}. For example,
+ − 2933 instead of a manual for @code{diff} and a manual for @code{diff3}, we
+ − 2934 have one manual for ``comparison of files'' which covers both of those
+ − 2935 programs, as well as @code{cmp}. By documenting these programs
+ − 2936 together, we can make the whole subject clearer.
+ − 2937
462
+ − 2938 The manual which discusses a program should certainly document all of
+ − 2939 the program's command-line options and all of its commands. It should
+ − 2940 give examples of their use. But don't organize the manual as a list of
428
+ − 2941 features. Instead, organize it logically, by subtopics. Address the
+ − 2942 questions that a user will ask when thinking about the job that the
+ − 2943 program does.
+ − 2944
+ − 2945 In general, a GNU manual should serve both as tutorial and reference.
+ − 2946 It should be set up for convenient access to each topic through Info,
+ − 2947 and for reading straight through (appendixes aside). A GNU manual
+ − 2948 should give a good introduction to a beginner reading through from the
+ − 2949 start, and should also provide all the details that hackers want.
+ − 2950 The Bison manual is a good example of this---please take a look at it
+ − 2951 to see what we mean.
+ − 2952
+ − 2953 That is not as hard as it first sounds. Arrange each chapter as a
+ − 2954 logical breakdown of its topic, but order the sections, and write their
+ − 2955 text, so that reading the chapter straight through makes sense. Do
+ − 2956 likewise when structuring the book into chapters, and when structuring a
+ − 2957 section into paragraphs. The watchword is, @emph{at each point, address
+ − 2958 the most fundamental and important issue raised by the preceding text.}
+ − 2959
+ − 2960 If necessary, add extra chapters at the beginning of the manual which
+ − 2961 are purely tutorial and cover the basics of the subject. These provide
+ − 2962 the framework for a beginner to understand the rest of the manual. The
+ − 2963 Bison manual provides a good example of how to do this.
+ − 2964
462
+ − 2965 To serve as a reference, a manual should have an Index that list all the
+ − 2966 functions, variables, options, and important concepts that are part of
+ − 2967 the program. One combined Index should do for a short manual, but
+ − 2968 sometimes for a complex package it is better to use multiple indices.
+ − 2969 The Texinfo manual includes advice on preparing good index entries, see
+ − 2970 @ref{Index Entries, , Making Index Entries, texinfo, The GNU Texinfo
+ − 2971 Manual}, and see @ref{Indexing Commands, , Defining the Entries of an
+ − 2972 Index, texinfo, The GNU Texinfo manual}.
+ − 2973
428
+ − 2974 Don't use Unix man pages as a model for how to write GNU documentation;
+ − 2975 most of them are terse, badly structured, and give inadequate
462
+ − 2976 explanation of the underlying concepts. (There are, of course, some
+ − 2977 exceptions.) Also, Unix man pages use a particular format which is
428
+ − 2978 different from what we use in GNU manuals.
+ − 2979
+ − 2980 Please include an email address in the manual for where to report
+ − 2981 bugs @emph{in the manual}.
+ − 2982
+ − 2983 Please do not use the term ``pathname'' that is used in Unix
+ − 2984 documentation; use ``file name'' (two words) instead. We use the term
+ − 2985 ``path'' only for search paths, which are lists of directory names.
+ − 2986
+ − 2987 Please do not use the term ``illegal'' to refer to erroneous input to a
+ − 2988 computer program. Please use ``invalid'' for this, and reserve the term
462
+ − 2989 ``illegal'' for activities punishable by law.
+ − 2990
+ − 2991 @node Doc Strings and Manuals
+ − 2992 @section Doc Strings and Manuals
+ − 2993
+ − 2994 Some programming systems, such as Emacs, provide a documentation string
+ − 2995 for each function, command or variable. You may be tempted to write a
+ − 2996 reference manual by compiling the documentation strings and writing a
+ − 2997 little additional text to go around them---but you must not do it. That
+ − 2998 approach is a fundamental mistake. The text of well-written
+ − 2999 documentation strings will be entirely wrong for a manual.
+ − 3000
+ − 3001 A documentation string needs to stand alone---when it appears on the
+ − 3002 screen, there will be no other text to introduce or explain it.
+ − 3003 Meanwhile, it can be rather informal in style.
+ − 3004
+ − 3005 The text describing a function or variable in a manual must not stand
+ − 3006 alone; it appears in the context of a section or subsection. Other text
+ − 3007 at the beginning of the section should explain some of the concepts, and
+ − 3008 should often make some general points that apply to several functions or
+ − 3009 variables. The previous descriptions of functions and variables in the
+ − 3010 section will also have given information about the topic. A description
+ − 3011 written to stand alone would repeat some of that information; this
+ − 3012 redundance looks bad. Meanwhile, the informality that is acceptable in
+ − 3013 a documentation string is totally unacceptable in a manual.
+ − 3014
+ − 3015 The only good way to use documentation strings in writing a good manual
+ − 3016 is to use them as a source of information for writing good text.
428
+ − 3017
+ − 3018 @node Manual Structure Details
+ − 3019 @section Manual Structure Details
462
+ − 3020 @cindex manual structure
428
+ − 3021
+ − 3022 The title page of the manual should state the version of the programs or
+ − 3023 packages documented in the manual. The Top node of the manual should
+ − 3024 also contain this information. If the manual is changing more
+ − 3025 frequently than or independent of the program, also state a version
+ − 3026 number for the manual in both of these places.
+ − 3027
+ − 3028 Each program documented in the manual should have a node named
+ − 3029 @samp{@var{program} Invocation} or @samp{Invoking @var{program}}. This
+ − 3030 node (together with its subnodes, if any) should describe the program's
+ − 3031 command line arguments and how to run it (the sort of information people
+ − 3032 would look in a man page for). Start with an @samp{@@example}
+ − 3033 containing a template for all the options and arguments that the program
+ − 3034 uses.
+ − 3035
+ − 3036 Alternatively, put a menu item in some menu whose item name fits one of
+ − 3037 the above patterns. This identifies the node which that item points to
+ − 3038 as the node for this purpose, regardless of the node's actual name.
+ − 3039
462
+ − 3040 The @samp{--usage} feature of the Info reader looks for such a node
+ − 3041 or menu item in order to find the relevant text, so it is essential
+ − 3042 for every Texinfo file to have one.
428
+ − 3043
+ − 3044 If one manual describes several programs, it should have such a node for
462
+ − 3045 each program described in the manual.
428
+ − 3046
+ − 3047 @node License for Manuals
+ − 3048 @section License for Manuals
462
+ − 3049 @cindex license for manuals
+ − 3050
+ − 3051 Please use the GNU Free Documentation License for all GNU manuals that
+ − 3052 are more than a few pages long. Likewise for a collection of short
+ − 3053 documents---you only need one copy of the GNU FDL for the whole
+ − 3054 collection. For a single short document, you can use a very permissive
+ − 3055 non-copyleft license, to avoid taking up space with a long license.
+ − 3056
+ − 3057 See @uref{http://www.gnu.org/copyleft/fdl-howto.html} for more explanation
+ − 3058 of how to employ the GFDL.
+ − 3059
+ − 3060 Note that it is not obligatory to include a copy of the GNU GPL or GNU
+ − 3061 LGPL in a manual whose license is neither the GPL nor the LGPL. It can
+ − 3062 be a good idea to include the program's license in a large manual; in a
+ − 3063 short manual, whose size would be increased considerably by including
+ − 3064 the program's license, it is probably better not to include it.
+ − 3065
+ − 3066 @node Manual Credits
+ − 3067 @section Manual Credits
+ − 3068 @cindex credits for manuals
+ − 3069
+ − 3070 Please credit the principal human writers of the manual as the authors,
+ − 3071 on the title page of the manual. If a company sponsored the work, thank
+ − 3072 the company in a suitable place in the manual, but do not cite the
+ − 3073 company as an author.
+ − 3074
+ − 3075 @node Printed Manuals
+ − 3076 @section Printed Manuals
+ − 3077
+ − 3078 The FSF publishes some GNU manuals in printed form. To encourage sales
+ − 3079 of these manuals, the on-line versions of the manual should mention at
+ − 3080 the very start that the printed manual is available and should point at
+ − 3081 information for getting it---for instance, with a link to the page
+ − 3082 @url{http://www.gnu.org/order/order.html}. This should not be included
+ − 3083 in the printed manual, though, because there it is redundant.
+ − 3084
+ − 3085 It is also useful to explain in the on-line forms of the manual how the
+ − 3086 user can print out the manual from the sources.
428
+ − 3087
+ − 3088 @node NEWS File
+ − 3089 @section The NEWS File
462
+ − 3090 @cindex @file{NEWS} file
428
+ − 3091
+ − 3092 In addition to its manual, the package should have a file named
+ − 3093 @file{NEWS} which contains a list of user-visible changes worth
+ − 3094 mentioning. In each new release, add items to the front of the file and
+ − 3095 identify the version they pertain to. Don't discard old items; leave
+ − 3096 them in the file after the newer items. This way, a user upgrading from
+ − 3097 any previous version can see what is new.
+ − 3098
+ − 3099 If the @file{NEWS} file gets very long, move some of the older items
+ − 3100 into a file named @file{ONEWS} and put a note at the end referring the
+ − 3101 user to that file.
+ − 3102
+ − 3103 @node Change Logs
+ − 3104 @section Change Logs
462
+ − 3105 @cindex change logs
428
+ − 3106
+ − 3107 Keep a change log to describe all the changes made to program source
+ − 3108 files. The purpose of this is so that people investigating bugs in the
+ − 3109 future will know about the changes that might have introduced the bug.
+ − 3110 Often a new bug can be found by looking at what was recently changed.
+ − 3111 More importantly, change logs can help you eliminate conceptual
+ − 3112 inconsistencies between different parts of a program, by giving you a
+ − 3113 history of how the conflicting concepts arose and who they came from.
+ − 3114
+ − 3115 @menu
+ − 3116 * Change Log Concepts::
+ − 3117 * Style of Change Logs::
+ − 3118 * Simple Changes::
+ − 3119 * Conditional Changes::
462
+ − 3120 * Indicating the Part Changed::
428
+ − 3121 @end menu
+ − 3122
+ − 3123 @node Change Log Concepts
+ − 3124 @subsection Change Log Concepts
+ − 3125
+ − 3126 You can think of the change log as a conceptual ``undo list'' which
+ − 3127 explains how earlier versions were different from the current version.
+ − 3128 People can see the current version; they don't need the change log
+ − 3129 to tell them what is in it. What they want from a change log is a
+ − 3130 clear explanation of how the earlier version differed.
+ − 3131
+ − 3132 The change log file is normally called @file{ChangeLog} and covers an
+ − 3133 entire directory. Each directory can have its own change log, or a
+ − 3134 directory can use the change log of its parent directory--it's up to
+ − 3135 you.
+ − 3136
+ − 3137 Another alternative is to record change log information with a version
+ − 3138 control system such as RCS or CVS. This can be converted automatically
+ − 3139 to a @file{ChangeLog} file using @code{rcs2log}; in Emacs, the command
+ − 3140 @kbd{C-x v a} (@code{vc-update-change-log}) does the job.
+ − 3141
+ − 3142 There's no need to describe the full purpose of the changes or how they
+ − 3143 work together. If you think that a change calls for explanation, you're
+ − 3144 probably right. Please do explain it---but please put the explanation
+ − 3145 in comments in the code, where people will see it whenever they see the
+ − 3146 code. For example, ``New function'' is enough for the change log when
+ − 3147 you add a function, because there should be a comment before the
+ − 3148 function definition to explain what it does.
+ − 3149
+ − 3150 However, sometimes it is useful to write one line to describe the
+ − 3151 overall purpose of a batch of changes.
+ − 3152
+ − 3153 The easiest way to add an entry to @file{ChangeLog} is with the Emacs
+ − 3154 command @kbd{M-x add-change-log-entry}. An entry should have an
+ − 3155 asterisk, the name of the changed file, and then in parentheses the name
+ − 3156 of the changed functions, variables or whatever, followed by a colon.
+ − 3157 Then describe the changes you made to that function or variable.
+ − 3158
+ − 3159 @node Style of Change Logs
+ − 3160 @subsection Style of Change Logs
462
+ − 3161 @cindex change logs, style
+ − 3162
+ − 3163 Here are some simple examples of change log entries, starting with the
+ − 3164 header line that says who made the change and when, followed by
+ − 3165 descriptions of specific changes. (These examples are drawn from Emacs
+ − 3166 and GCC.)
428
+ − 3167
+ − 3168 @example
462
+ − 3169 1998-08-17 Richard Stallman <rms@@gnu.org>
+ − 3170
428
+ − 3171 * register.el (insert-register): Return nil.
+ − 3172 (jump-to-register): Likewise.
+ − 3173
+ − 3174 * sort.el (sort-subr): Return nil.
+ − 3175
+ − 3176 * tex-mode.el (tex-bibtex-file, tex-file, tex-region):
+ − 3177 Restart the tex shell if process is gone or stopped.
+ − 3178 (tex-shell-running): New function.
+ − 3179
+ − 3180 * expr.c (store_one_arg): Round size up for move_block_to_reg.
+ − 3181 (expand_call): Round up when emitting USE insns.
+ − 3182 * stmt.c (assign_parms): Round size up for move_block_from_reg.
+ − 3183 @end example
+ − 3184
+ − 3185 It's important to name the changed function or variable in full. Don't
+ − 3186 abbreviate function or variable names, and don't combine them.
+ − 3187 Subsequent maintainers will often search for a function name to find all
+ − 3188 the change log entries that pertain to it; if you abbreviate the name,
+ − 3189 they won't find it when they search.
+ − 3190
+ − 3191 For example, some people are tempted to abbreviate groups of function
+ − 3192 names by writing @samp{* register.el (@{insert,jump-to@}-register)};
+ − 3193 this is not a good idea, since searching for @code{jump-to-register} or
+ − 3194 @code{insert-register} would not find that entry.
+ − 3195
+ − 3196 Separate unrelated change log entries with blank lines. When two
+ − 3197 entries represent parts of the same change, so that they work together,
+ − 3198 then don't put blank lines between them. Then you can omit the file
+ − 3199 name and the asterisk when successive entries are in the same file.
+ − 3200
462
+ − 3201 Break long lists of function names by closing continued lines with
+ − 3202 @samp{)}, rather than @samp{,}, and opening the continuation with
+ − 3203 @samp{(} as in this example:
+ − 3204
+ − 3205 @example
+ − 3206 * keyboard.c (menu_bar_items, tool_bar_items)
+ − 3207 (Fexecute_extended_command): Deal with `keymap' property.
+ − 3208 @end example
+ − 3209
428
+ − 3210 @node Simple Changes
+ − 3211 @subsection Simple Changes
+ − 3212
+ − 3213 Certain simple kinds of changes don't need much detail in the change
+ − 3214 log.
+ − 3215
+ − 3216 When you change the calling sequence of a function in a simple fashion,
462
+ − 3217 and you change all the callers of the function to use the new calling
+ − 3218 sequence, there is no need to make individual entries for all the
+ − 3219 callers that you changed. Just write in the entry for the function
+ − 3220 being called, ``All callers changed''---like this:
428
+ − 3221
+ − 3222 @example
+ − 3223 * keyboard.c (Fcommand_execute): New arg SPECIAL.
+ − 3224 All callers changed.
+ − 3225 @end example
+ − 3226
+ − 3227 When you change just comments or doc strings, it is enough to write an
+ − 3228 entry for the file, without mentioning the functions. Just ``Doc
+ − 3229 fixes'' is enough for the change log.
+ − 3230
+ − 3231 There's no need to make change log entries for documentation files.
+ − 3232 This is because documentation is not susceptible to bugs that are hard
+ − 3233 to fix. Documentation does not consist of parts that must interact in a
+ − 3234 precisely engineered fashion. To correct an error, you need not know
+ − 3235 the history of the erroneous passage; it is enough to compare what the
+ − 3236 documentation says with the way the program actually works.
+ − 3237
+ − 3238 @node Conditional Changes
+ − 3239 @subsection Conditional Changes
462
+ − 3240 @cindex conditional changes, and change logs
+ − 3241 @cindex change logs, conditional changes
428
+ − 3242
+ − 3243 C programs often contain compile-time @code{#if} conditionals. Many
+ − 3244 changes are conditional; sometimes you add a new definition which is
+ − 3245 entirely contained in a conditional. It is very useful to indicate in
+ − 3246 the change log the conditions for which the change applies.
+ − 3247
+ − 3248 Our convention for indicating conditional changes is to use square
+ − 3249 brackets around the name of the condition.
+ − 3250
+ − 3251 Here is a simple example, describing a change which is conditional but
+ − 3252 does not have a function or entity name associated with it:
+ − 3253
+ − 3254 @example
+ − 3255 * xterm.c [SOLARIS2]: Include string.h.
+ − 3256 @end example
+ − 3257
+ − 3258 Here is an entry describing a new definition which is entirely
+ − 3259 conditional. This new definition for the macro @code{FRAME_WINDOW_P} is
+ − 3260 used only when @code{HAVE_X_WINDOWS} is defined:
+ − 3261
+ − 3262 @example
+ − 3263 * frame.h [HAVE_X_WINDOWS] (FRAME_WINDOW_P): Macro defined.
+ − 3264 @end example
+ − 3265
+ − 3266 Here is an entry for a change within the function @code{init_display},
+ − 3267 whose definition as a whole is unconditional, but the changes themselves
+ − 3268 are contained in a @samp{#ifdef HAVE_LIBNCURSES} conditional:
+ − 3269
+ − 3270 @example
+ − 3271 * dispnew.c (init_display) [HAVE_LIBNCURSES]: If X, call tgetent.
+ − 3272 @end example
+ − 3273
+ − 3274 Here is an entry for a change that takes affect only when
+ − 3275 a certain macro is @emph{not} defined:
+ − 3276
+ − 3277 @example
+ − 3278 (gethostname) [!HAVE_SOCKETS]: Replace with winsock version.
+ − 3279 @end example
+ − 3280
462
+ − 3281 @node Indicating the Part Changed
+ − 3282 @subsection Indicating the Part Changed
+ − 3283
+ − 3284 Indicate the part of a function which changed by using angle brackets
+ − 3285 enclosing an indication of what the changed part does. Here is an entry
+ − 3286 for a change in the part of the function @code{sh-while-getopts} that
+ − 3287 deals with @code{sh} commands:
+ − 3288
+ − 3289 @example
+ − 3290 * progmodes/sh-script.el (sh-while-getopts) <sh>: Handle case that
+ − 3291 user-specified option string is empty.
+ − 3292 @end example
+ − 3293
+ − 3294
428
+ − 3295 @node Man Pages
+ − 3296 @section Man Pages
462
+ − 3297 @cindex man pages
428
+ − 3298
+ − 3299 In the GNU project, man pages are secondary. It is not necessary or
+ − 3300 expected for every GNU program to have a man page, but some of them do.
+ − 3301 It's your choice whether to include a man page in your program.
+ − 3302
+ − 3303 When you make this decision, consider that supporting a man page
+ − 3304 requires continual effort each time the program is changed. The time
+ − 3305 you spend on the man page is time taken away from more useful work.
+ − 3306
+ − 3307 For a simple program which changes little, updating the man page may be
+ − 3308 a small job. Then there is little reason not to include a man page, if
+ − 3309 you have one.
+ − 3310
+ − 3311 For a large program that changes a great deal, updating a man page may
+ − 3312 be a substantial burden. If a user offers to donate a man page, you may
+ − 3313 find this gift costly to accept. It may be better to refuse the man
+ − 3314 page unless the same person agrees to take full responsibility for
+ − 3315 maintaining it---so that you can wash your hands of it entirely. If
+ − 3316 this volunteer later ceases to do the job, then don't feel obliged to
+ − 3317 pick it up yourself; it may be better to withdraw the man page from the
+ − 3318 distribution until someone else agrees to update it.
+ − 3319
+ − 3320 When a program changes only a little, you may feel that the
+ − 3321 discrepancies are small enough that the man page remains useful without
+ − 3322 updating. If so, put a prominent note near the beginning of the man
+ − 3323 page explaining that you don't maintain it and that the Texinfo manual
+ − 3324 is more authoritative. The note should say how to access the Texinfo
+ − 3325 documentation.
+ − 3326
+ − 3327 @node Reading other Manuals
+ − 3328 @section Reading other Manuals
+ − 3329
+ − 3330 There may be non-free books or documentation files that describe the
+ − 3331 program you are documenting.
+ − 3332
+ − 3333 It is ok to use these documents for reference, just as the author of a
+ − 3334 new algebra textbook can read other books on algebra. A large portion
+ − 3335 of any non-fiction book consists of facts, in this case facts about how
+ − 3336 a certain program works, and these facts are necessarily the same for
+ − 3337 everyone who writes about the subject. But be careful not to copy your
+ − 3338 outline structure, wording, tables or examples from preexisting non-free
+ − 3339 documentation. Copying from free documentation may be ok; please check
+ − 3340 with the FSF about the individual case.
+ − 3341
+ − 3342 @node Managing Releases
+ − 3343 @chapter The Release Process
462
+ − 3344 @cindex releasing
428
+ − 3345
+ − 3346 Making a release is more than just bundling up your source files in a
+ − 3347 tar file and putting it up for FTP. You should set up your software so
+ − 3348 that it can be configured to run on a variety of systems. Your Makefile
+ − 3349 should conform to the GNU standards described below, and your directory
+ − 3350 layout should also conform to the standards discussed below. Doing so
+ − 3351 makes it easy to include your package into the larger framework of
+ − 3352 all GNU software.
+ − 3353
+ − 3354 @menu
+ − 3355 * Configuration:: How Configuration Should Work
+ − 3356 * Makefile Conventions:: Makefile Conventions
+ − 3357 * Releases:: Making Releases
+ − 3358 @end menu
+ − 3359
+ − 3360 @node Configuration
+ − 3361 @section How Configuration Should Work
462
+ − 3362 @cindex program configuration
+ − 3363
+ − 3364 @pindex configure
428
+ − 3365 Each GNU distribution should come with a shell script named
+ − 3366 @code{configure}. This script is given arguments which describe the
+ − 3367 kind of machine and system you want to compile the program for.
+ − 3368
+ − 3369 The @code{configure} script must record the configuration options so
+ − 3370 that they affect compilation.
+ − 3371
+ − 3372 One way to do this is to make a link from a standard name such as
+ − 3373 @file{config.h} to the proper configuration file for the chosen system.
+ − 3374 If you use this technique, the distribution should @emph{not} contain a
+ − 3375 file named @file{config.h}. This is so that people won't be able to
+ − 3376 build the program without configuring it first.
+ − 3377
+ − 3378 Another thing that @code{configure} can do is to edit the Makefile. If
+ − 3379 you do this, the distribution should @emph{not} contain a file named
+ − 3380 @file{Makefile}. Instead, it should include a file @file{Makefile.in} which
+ − 3381 contains the input used for editing. Once again, this is so that people
+ − 3382 won't be able to build the program without configuring it first.
+ − 3383
+ − 3384 If @code{configure} does write the @file{Makefile}, then @file{Makefile}
+ − 3385 should have a target named @file{Makefile} which causes @code{configure}
+ − 3386 to be rerun, setting up the same configuration that was set up last
+ − 3387 time. The files that @code{configure} reads should be listed as
+ − 3388 dependencies of @file{Makefile}.
+ − 3389
+ − 3390 All the files which are output from the @code{configure} script should
+ − 3391 have comments at the beginning explaining that they were generated
+ − 3392 automatically using @code{configure}. This is so that users won't think
+ − 3393 of trying to edit them by hand.
+ − 3394
+ − 3395 The @code{configure} script should write a file named @file{config.status}
+ − 3396 which describes which configuration options were specified when the
+ − 3397 program was last configured. This file should be a shell script which,
+ − 3398 if run, will recreate the same configuration.
+ − 3399
+ − 3400 The @code{configure} script should accept an option of the form
+ − 3401 @samp{--srcdir=@var{dirname}} to specify the directory where sources are found
+ − 3402 (if it is not the current directory). This makes it possible to build
+ − 3403 the program in a separate directory, so that the actual source directory
+ − 3404 is not modified.
+ − 3405
+ − 3406 If the user does not specify @samp{--srcdir}, then @code{configure} should
+ − 3407 check both @file{.} and @file{..} to see if it can find the sources. If
+ − 3408 it finds the sources in one of these places, it should use them from
+ − 3409 there. Otherwise, it should report that it cannot find the sources, and
+ − 3410 should exit with nonzero status.
+ − 3411
+ − 3412 Usually the easy way to support @samp{--srcdir} is by editing a
+ − 3413 definition of @code{VPATH} into the Makefile. Some rules may need to
+ − 3414 refer explicitly to the specified source directory. To make this
+ − 3415 possible, @code{configure} can add to the Makefile a variable named
+ − 3416 @code{srcdir} whose value is precisely the specified directory.
+ − 3417
+ − 3418 The @code{configure} script should also take an argument which specifies the
+ − 3419 type of system to build the program for. This argument should look like
+ − 3420 this:
+ − 3421
+ − 3422 @example
+ − 3423 @var{cpu}-@var{company}-@var{system}
+ − 3424 @end example
+ − 3425
+ − 3426 For example, a Sun 3 might be @samp{m68k-sun-sunos4.1}.
+ − 3427
+ − 3428 The @code{configure} script needs to be able to decode all plausible
+ − 3429 alternatives for how to describe a machine. Thus, @samp{sun3-sunos4.1}
+ − 3430 would be a valid alias. For many programs, @samp{vax-dec-ultrix} would
+ − 3431 be an alias for @samp{vax-dec-bsd}, simply because the differences
462
+ − 3432 between Ultrix and @sc{bsd} are rarely noticeable, but a few programs
428
+ − 3433 might need to distinguish them.
+ − 3434 @c Real 4.4BSD now runs on some Suns.
+ − 3435
+ − 3436 There is a shell script called @file{config.sub} that you can use
+ − 3437 as a subroutine to validate system types and canonicalize aliases.
+ − 3438
462
+ − 3439 @cindex optional features, configure-time
428
+ − 3440 Other options are permitted to specify in more detail the software
+ − 3441 or hardware present on the machine, and include or exclude optional
+ − 3442 parts of the package:
+ − 3443
+ − 3444 @table @samp
+ − 3445 @item --enable-@var{feature}@r{[}=@var{parameter}@r{]}
+ − 3446 Configure the package to build and install an optional user-level
+ − 3447 facility called @var{feature}. This allows users to choose which
+ − 3448 optional features to include. Giving an optional @var{parameter} of
+ − 3449 @samp{no} should omit @var{feature}, if it is built by default.
+ − 3450
+ − 3451 No @samp{--enable} option should @strong{ever} cause one feature to
+ − 3452 replace another. No @samp{--enable} option should ever substitute one
+ − 3453 useful behavior for another useful behavior. The only proper use for
+ − 3454 @samp{--enable} is for questions of whether to build part of the program
+ − 3455 or exclude it.
+ − 3456
+ − 3457 @item --with-@var{package}
+ − 3458 @c @r{[}=@var{parameter}@r{]}
+ − 3459 The package @var{package} will be installed, so configure this package
+ − 3460 to work with @var{package}.
+ − 3461
+ − 3462 @c Giving an optional @var{parameter} of
+ − 3463 @c @samp{no} should omit @var{package}, if it is used by default.
+ − 3464
+ − 3465 Possible values of @var{package} include
+ − 3466 @samp{gnu-as} (or @samp{gas}), @samp{gnu-ld}, @samp{gnu-libc},
+ − 3467 @samp{gdb},
+ − 3468 @samp{x},
+ − 3469 and
+ − 3470 @samp{x-toolkit}.
+ − 3471
+ − 3472 Do not use a @samp{--with} option to specify the file name to use to
+ − 3473 find certain files. That is outside the scope of what @samp{--with}
+ − 3474 options are for.
+ − 3475 @end table
+ − 3476
+ − 3477 All @code{configure} scripts should accept all of these ``detail''
+ − 3478 options, whether or not they make any difference to the particular
+ − 3479 package at hand. In particular, they should accept any option that
+ − 3480 starts with @samp{--with-} or @samp{--enable-}. This is so users will
+ − 3481 be able to configure an entire GNU source tree at once with a single set
+ − 3482 of options.
+ − 3483
+ − 3484 You will note that the categories @samp{--with-} and @samp{--enable-}
+ − 3485 are narrow: they @strong{do not} provide a place for any sort of option
+ − 3486 you might think of. That is deliberate. We want to limit the possible
+ − 3487 configuration options in GNU software. We do not want GNU programs to
+ − 3488 have idiosyncratic configuration options.
+ − 3489
462
+ − 3490 Packages that perform part of the compilation process may support
+ − 3491 cross-compilation. In such a case, the host and target machines for the
+ − 3492 program may be different.
+ − 3493
+ − 3494 The @code{configure} script should normally treat the specified type of
+ − 3495 system as both the host and the target, thus producing a program which
+ − 3496 works for the same type of machine that it runs on.
+ − 3497
+ − 3498 To configure a cross-compiler, cross-assembler, or what have you, you
+ − 3499 should specify a target different from the host, using the configure
+ − 3500 option @samp{--target=@var{targettype}}. The syntax for
+ − 3501 @var{targettype} is the same as for the host type. So the command would
+ − 3502 look like this:
+ − 3503
+ − 3504 @example
+ − 3505 ./configure @var{hosttype} --target=@var{targettype}
+ − 3506 @end example
+ − 3507
+ − 3508 Programs for which cross-operation is not meaningful need not accept the
+ − 3509 @samp{--target} option, because configuring an entire operating system for
+ − 3510 cross-operation is not a meaningful operation.
428
+ − 3511
+ − 3512 Bootstrapping a cross-compiler requires compiling it on a machine other
+ − 3513 than the host it will run on. Compilation packages accept a
462
+ − 3514 configuration option @samp{--build=@var{buildtype}} for specifying the
+ − 3515 configuration on which you will compile them, but the configure script
+ − 3516 should normally guess the build machine type (using
+ − 3517 @file{config.guess}), so this option is probably not necessary. The
+ − 3518 host and target types normally default from the build type, so in
+ − 3519 bootstrapping a cross-compiler you must specify them both explicitly.
428
+ − 3520
+ − 3521 Some programs have ways of configuring themselves automatically. If
+ − 3522 your program is set up to do this, your @code{configure} script can simply
+ − 3523 ignore most of its arguments.
+ − 3524
+ − 3525 @comment The makefile standards are in a separate file that is also
+ − 3526 @comment included by make.texinfo. Done by roland@gnu.ai.mit.edu on 1/6/93.
+ − 3527 @comment For this document, turn chapters into sections, etc.
+ − 3528 @lowersections
+ − 3529 @include make-stds.texi
+ − 3530 @raisesections
+ − 3531
+ − 3532 @node Releases
+ − 3533 @section Making Releases
462
+ − 3534 @cindex packaging
428
+ − 3535
+ − 3536 Package the distribution of @code{Foo version 69.96} up in a gzipped tar
+ − 3537 file with the name @file{foo-69.96.tar.gz}. It should unpack into a
+ − 3538 subdirectory named @file{foo-69.96}.
+ − 3539
+ − 3540 Building and installing the program should never modify any of the files
+ − 3541 contained in the distribution. This means that all the files that form
+ − 3542 part of the program in any way must be classified into @dfn{source
+ − 3543 files} and @dfn{non-source files}. Source files are written by humans
+ − 3544 and never changed automatically; non-source files are produced from
+ − 3545 source files by programs under the control of the Makefile.
+ − 3546
462
+ − 3547 @cindex @file{README} file
428
+ − 3548 The distribution should contain a file named @file{README} which gives
+ − 3549 the name of the package, and a general description of what it does. It
+ − 3550 is also good to explain the purpose of each of the first-level
+ − 3551 subdirectories in the package, if there are any. The @file{README} file
+ − 3552 should either state the version number of the package, or refer to where
+ − 3553 in the package it can be found.
+ − 3554
+ − 3555 The @file{README} file should refer to the file @file{INSTALL}, which
+ − 3556 should contain an explanation of the installation procedure.
+ − 3557
+ − 3558 The @file{README} file should also refer to the file which contains the
+ − 3559 copying conditions. The GNU GPL, if used, should be in a file called
+ − 3560 @file{COPYING}. If the GNU LGPL is used, it should be in a file called
+ − 3561 @file{COPYING.LIB}.
+ − 3562
+ − 3563 Naturally, all the source files must be in the distribution. It is okay
+ − 3564 to include non-source files in the distribution, provided they are
+ − 3565 up-to-date and machine-independent, so that building the distribution
+ − 3566 normally will never modify them. We commonly include non-source files
+ − 3567 produced by Bison, @code{lex}, @TeX{}, and @code{makeinfo}; this helps avoid
+ − 3568 unnecessary dependencies between our distributions, so that users can
+ − 3569 install whichever packages they want to install.
+ − 3570
+ − 3571 Non-source files that might actually be modified by building and
+ − 3572 installing the program should @strong{never} be included in the
+ − 3573 distribution. So if you do distribute non-source files, always make
+ − 3574 sure they are up to date when you make a new distribution.
+ − 3575
+ − 3576 Make sure that the directory into which the distribution unpacks (as
+ − 3577 well as any subdirectories) are all world-writable (octal mode 777).
+ − 3578 This is so that old versions of @code{tar} which preserve the
+ − 3579 ownership and permissions of the files from the tar archive will be
+ − 3580 able to extract all the files even if the user is unprivileged.
+ − 3581
+ − 3582 Make sure that all the files in the distribution are world-readable.
+ − 3583
+ − 3584 Make sure that no file name in the distribution is more than 14
+ − 3585 characters long. Likewise, no file created by building the program
+ − 3586 should have a name longer than 14 characters. The reason for this is
+ − 3587 that some systems adhere to a foolish interpretation of the @sc{posix}
+ − 3588 standard, and refuse to open a longer name, rather than truncating as
+ − 3589 they did in the past.
+ − 3590
+ − 3591 Don't include any symbolic links in the distribution itself. If the tar
+ − 3592 file contains symbolic links, then people cannot even unpack it on
+ − 3593 systems that don't support symbolic links. Also, don't use multiple
+ − 3594 names for one file in different directories, because certain file
+ − 3595 systems cannot handle this and that prevents unpacking the
+ − 3596 distribution.
+ − 3597
+ − 3598 Try to make sure that all the file names will be unique on MS-DOS. A
+ − 3599 name on MS-DOS consists of up to 8 characters, optionally followed by a
+ − 3600 period and up to three characters. MS-DOS will truncate extra
+ − 3601 characters both before and after the period. Thus,
+ − 3602 @file{foobarhacker.c} and @file{foobarhacker.o} are not ambiguous; they
+ − 3603 are truncated to @file{foobarha.c} and @file{foobarha.o}, which are
+ − 3604 distinct.
+ − 3605
462
+ − 3606 @cindex @file{texinfo.tex}, in a distribution
428
+ − 3607 Include in your distribution a copy of the @file{texinfo.tex} you used
+ − 3608 to test print any @file{*.texinfo} or @file{*.texi} files.
+ − 3609
+ − 3610 Likewise, if your program uses small GNU software packages like regex,
+ − 3611 getopt, obstack, or termcap, include them in the distribution file.
+ − 3612 Leaving them out would make the distribution file a little smaller at
+ − 3613 the expense of possible inconvenience to a user who doesn't know what
+ − 3614 other files to get.
+ − 3615
+ − 3616 @node References
+ − 3617 @chapter References to Non-Free Software and Documentation
462
+ − 3618 @cindex references to non-free material
428
+ − 3619
+ − 3620 A GNU program should not recommend use of any non-free program. We
+ − 3621 can't stop some people from writing proprietary programs, or stop other
+ − 3622 people from using them. But we can and should avoid helping to
+ − 3623 advertise them to new customers.
+ − 3624
+ − 3625 Sometimes it is important to mention how to build your package on top of
+ − 3626 some non-free operating system or other non-free base package. In such
+ − 3627 cases, please mention the name of the non-free package or system in the
+ − 3628 briefest possible way. Don't include any references for where to find
+ − 3629 more information about the proprietary program. The goal should be that
+ − 3630 people already using the proprietary program will get the advice they
+ − 3631 need about how to use your free program, while people who don't already
+ − 3632 use the proprietary program will not see anything to encourage them to
+ − 3633 take an interest in it.
+ − 3634
+ − 3635 Likewise, a GNU package should not refer the user to any non-free
+ − 3636 documentation for free software. The need for free documentation to go
+ − 3637 with free software is now a major focus of the GNU project; to show that
+ − 3638 we are serious about the need for free documentation, we must not
+ − 3639 undermine our position by recommending use of documentation that isn't
+ − 3640 free.
+ − 3641
462
+ − 3642 @node Index
+ − 3643 @unnumbered Index
+ − 3644 @printindex cp
+ − 3645
428
+ − 3646 @contents
+ − 3647
+ − 3648 @bye
