1 \input texinfo @c -*-texinfo-*-
3 @setfilename standards.info
4 @settitle GNU Coding Standards
5 @c This date is automagically updated when you save this file:
6 @set lastupdate September 13, 2013
9 @dircategory GNU organization
11 * Standards: (standards). GNU coding standards.
14 @c @setchapternewpage odd
15 @setchapternewpage off
17 @c Put everything in one index (arbitrarily chosen to be the concept index).
23 @c This is used by a cross ref in make-stds.texi
27 The GNU coding standards, last updated @value{lastupdate}.
29 Copyright @copyright{} 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
30 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010,
31 2011, 2012, 2013 Free Software Foundation, Inc.
33 Permission is granted to copy, distribute and/or modify this document
34 under the terms of the GNU Free Documentation License, Version 1.3 or
35 any later version published by the Free Software Foundation; with no
36 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
37 Texts. A copy of the license is included in the section entitled
38 ``GNU Free Documentation License''.
42 @title GNU Coding Standards
43 @author Richard Stallman, et al.
44 @author last updated @value{lastupdate}
46 @vskip 0pt plus 1filll
54 @top GNU Coding Standards
60 * Preface:: About the GNU Coding Standards.
61 * Legal Issues:: Keeping free software free.
62 * Design Advice:: General program design.
63 * Program Behavior:: Program behavior for all programs
64 * Writing C:: Making the best use of C.
65 * Documentation:: Documenting programs.
66 * Managing Releases:: The release process.
67 * References:: Mentioning non-free software or documentation.
68 * GNU Free Documentation License:: Copying and sharing this manual.
74 @chapter About the GNU Coding Standards
76 The GNU Coding Standards were written by Richard Stallman and other GNU
77 Project volunteers. Their purpose is to make the GNU system clean,
78 consistent, and easy to install. This document can also be read as a
79 guide to writing portable, robust and reliable programs. It focuses on
80 programs written in C, but many of the rules and principles are useful
81 even if you write in another programming language. The rules often
82 state reasons for writing in a certain way.
84 @cindex where to obtain @code{standards.texi}
85 @cindex downloading this manual
86 If you did not obtain this file directly from the GNU project and
87 recently, please check for a newer version. You can get the GNU
88 Coding Standards from the GNU web server in many
89 different formats, including the Texinfo source, PDF, HTML, DVI, plain
90 text, and more, at: @uref{http://www.gnu.org/prep/standards/}.
92 If you are maintaining an official GNU package, in addition to this
93 document, please read and follow the GNU maintainer information
94 (@pxref{Top, , Contents, maintain, Information for Maintainers of GNU
97 @cindex @code{gnustandards-commit@@gnu.org} mailing list
98 If you want to receive diffs for every change to these GNU documents,
99 join the mailing list @code{gnustandards-commit@@gnu.org}, via the web
101 @url{http://lists.gnu.org/mailman/listinfo/gnustandards-commit}.
102 Archives are also available there.
104 @cindex @code{bug-standards@@gnu.org} email address
105 @cindex Savannah repository for gnustandards
106 @cindex gnustandards project repository
107 Please send corrections or suggestions for this document to
108 @email{bug-standards@@gnu.org}. If you make a suggestion, please
109 include a suggested new wording for it, to help us consider the
110 suggestion efficiently. We prefer a context diff to the Texinfo
111 source, but if that's difficult for you, you can make a context diff
112 for some other version of this document, or propose it in any way that
113 makes it clear. The source repository for this document can be found
114 at @url{http://savannah.gnu.org/projects/gnustandards}.
116 These standards cover the minimum of what is important when writing a
117 GNU package. Likely, the need for additional standards will come up.
118 Sometimes, you might suggest that such standards be added to this
119 document. If you think your standards would be generally useful, please
122 You should also set standards for your package on many questions not
123 addressed or not firmly specified here. The most important point is to
124 be self-consistent---try to stick to the conventions you pick, and try
125 to document them as much as possible. That way, your program will be
126 more maintainable by others.
128 The GNU Hello program serves as an example of how to follow the GNU
129 coding standards for a trivial program.
130 @uref{http://www.gnu.org/software/hello/hello.html}.
132 This release of the GNU Coding Standards was last updated
137 @chapter Keeping Free Software Free
138 @cindex legal aspects
140 This chapter discusses how you can make sure that GNU software
141 avoids legal difficulties, and other related issues.
144 * Reading Non-Free Code:: Referring to proprietary programs.
145 * Contributions:: Accepting contributions.
146 * Trademarks:: How we deal with trademark issues.
149 @node Reading Non-Free Code
150 @section Referring to Proprietary Programs
151 @cindex proprietary programs
152 @cindex avoiding proprietary code
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.)
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.
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 memory 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).
170 Or, on the contrary, emphasize simplicity instead of speed. For some
171 applications, the speed of today's computers makes simpler algorithms
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.
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.
186 @section Accepting Contributions
188 @cindex accepting contributions
190 If the program you are working on is copyrighted by the Free Software
191 Foundation, then when someone else sends you a piece of code to add to
192 the program, we need legal papers to use it---just as we asked you to
193 sign papers initially. @emph{Each} person who makes a nontrivial
194 contribution to a program must sign some sort of legal papers in order
195 for us to have clear title to the program; the main author alone is not
198 So, before adding in any contributions from other people, please tell
199 us, so we can arrange to get the papers. Then wait until we tell you
200 that we have received the signed papers, before you actually use the
203 This applies both before you release the program and afterward. If
204 you receive diffs to fix a bug, and they make significant changes, we
205 need legal papers for that change.
207 This also applies to comments and documentation files. For copyright
208 law, comments and code are just text. Copyright applies to all kinds of
209 text, so we need legal papers for all kinds.
211 We know it is frustrating to ask for legal papers; it's frustrating for
212 us as well. But if you don't wait, you are going out on a limb---for
213 example, what if the contributor's employer won't sign a disclaimer?
214 You might have to take that code out again!
216 You don't need papers for changes of a few lines here or there, since
217 they are not significant for copyright purposes. Also, you don't need
218 papers if all you get from the suggestion is some ideas, not actual code
219 which you use. For example, if someone sent you one implementation, but
220 you write a different implementation of the same idea, you don't need to
223 The very worst thing is if you forget to tell us about the other
224 contributor. We could be very embarrassed in court some day as a
227 We have more detailed advice for maintainers of GNU packages. If you
228 have reached the stage of maintaining a GNU program (whether released
229 or not), please take a look: @pxref{Legal Matters,,, maintain,
230 Information for GNU Maintainers}.
237 Please do not include any trademark acknowledgements in GNU software
238 packages or documentation.
240 Trademark acknowledgements are the statements that such-and-such is a
241 trademark of so-and-so. The GNU Project has no objection to the basic
242 idea of trademarks, but these acknowledgements feel like kowtowing,
243 and there is no legal requirement for them, so we don't use them.
245 What is legally required, as regards other people's trademarks, is to
246 avoid using them in ways which a reader might reasonably understand as
247 naming or labeling our own programs or activities. For example, since
248 ``Objective C'' is (or at least was) a trademark, we made sure to say
249 that we provide a ``compiler for the Objective C language'' rather
250 than an ``Objective C compiler''. The latter would have been meant as
251 a shorter way of saying the former, but it does not explicitly state
252 the relationship, so it could be misinterpreted as using ``Objective
253 C'' as a label for the compiler rather than for the language.
255 Please don't use ``win'' as an abbreviation for Microsoft Windows in
256 GNU software or documentation. In hacker terminology, calling
257 something a ``win'' is a form of praise. If you wish to praise
258 Microsoft Windows when speaking on your own, by all means do so, but
259 not in GNU software. Usually we write the name ``Windows'' in full,
260 but when brevity is very important (as in file names and sometimes
261 symbol names), we abbreviate it to ``w''. For instance, the files and
262 functions in Emacs that deal with Windows start with @samp{w32}.
265 @chapter General Program Design
266 @cindex program design
268 This chapter discusses some of the issues you should take into
269 account when designing your program.
271 @c Standard or ANSI C
273 @c In 1989 the American National Standards Institute (ANSI) standardized
274 @c C as standard X3.159-1989. In December of that year the
275 @c International Standards Organization ISO adopted the ANSI C standard
276 @c making minor changes. In 1990 ANSI then re-adopted ISO standard
277 @c C. This version of C is known as either ANSI C or Standard C.
279 @c A major revision of the C Standard appeared in 1999.
282 * Source Language:: Which languages to use.
283 * Compatibility:: Compatibility with other implementations.
284 * Using Extensions:: Using non-standard features.
285 * Standard C:: Using standard C features.
286 * Conditional Compilation:: Compiling code only if a conditional is true.
289 @node Source Language
290 @section Which Languages to Use
291 @cindex programming languages
293 When you want to use a language that gets compiled and runs at high
294 speed, the best language to use is C. C++ is ok too, but please don't
295 make heavy use of templates. So is Java, if you compile it.
297 When highest efficiency is not required, other languages commonly used
298 in the free software community, such as Scheme, Python, Ruby, and
299 Java, are OK too. Scheme, as implemented by GNU@tie{}Guile, plays a
300 particular role in the GNU System: it is the preferred language to
301 extend programs written in C/C++, and also a fine language for a wide
302 range of applications. The more GNU components use Guile and Scheme,
303 the more users are able to extend and combine them (@pxref{The Emacs
304 Thesis,,, guile, GNU Guile Reference Manual}).
306 Many programs are designed to be extensible: they include an interpreter
307 for a language that is higher level than C. Often much of the program
308 is written in that language, too. The Emacs editor pioneered this
312 @cindex GNOME and Guile
313 The standard extensibility interpreter for GNU software is Guile
314 (@uref{http://www.gnu.org/@/software/@/guile/}), which implements the
315 language Scheme (an especially clean and simple dialect of Lisp).
316 Guile also includes bindings for GTK+/GNOME, making it practical to
317 write modern GUI functionality within Guile. We don't reject programs
318 written in other ``scripting languages'' such as Perl and Python, but
319 using Guile is the path that will lead to overall consistency of the
324 @section Compatibility with Other Implementations
325 @cindex compatibility with C and POSIX standards
326 @cindex C compatibility
327 @cindex POSIX compatibility
329 With occasional exceptions, utility programs and libraries for GNU
330 should be upward compatible with those in Berkeley Unix, and upward
331 compatible with Standard C if Standard C specifies their
332 behavior, and upward compatible with POSIX if POSIX specifies
335 When these standards conflict, it is useful to offer compatibility
336 modes for each of them.
338 @cindex options for compatibility
339 Standard C and POSIX prohibit many kinds of extensions. Feel
340 free to make the extensions anyway, and include a @samp{--ansi},
341 @samp{--posix}, or @samp{--compatible} option to turn them off.
342 However, if the extension has a significant chance of breaking any real
343 programs or scripts, then it is not really upward compatible. So you
344 should try to redesign its interface to make it upward compatible.
346 @cindex @code{POSIXLY_CORRECT}, environment variable
347 Many GNU programs suppress extensions that conflict with POSIX if the
348 environment variable @code{POSIXLY_CORRECT} is defined (even if it is
349 defined with a null value). Please make your program recognize this
350 variable if appropriate.
352 When a feature is used only by users (not by programs or command
353 files), and it is done poorly in Unix, feel free to replace it
354 completely with something totally different and better. (For example,
355 @code{vi} is replaced with Emacs.) But it is nice to offer a compatible
356 feature as well. (There is a free @code{vi} clone, so we offer it.)
358 Additional useful features are welcome regardless of whether
359 there is any precedent for them.
361 @node Using Extensions
362 @section Using Non-standard Features
363 @cindex non-standard extensions
365 Many GNU facilities that already exist support a number of convenient
366 extensions over the comparable Unix facilities. Whether to use these
367 extensions in implementing your program is a difficult question.
369 On the one hand, using the extensions can make a cleaner program.
370 On the other hand, people will not be able to build the program
371 unless the other GNU tools are available. This might cause the
372 program to work on fewer kinds of machines.
374 With some extensions, it might be easy to provide both alternatives.
375 For example, you can define functions with a ``keyword'' @code{INLINE}
376 and define that as a macro to expand into either @code{inline} or
377 nothing, depending on the compiler.
379 In general, perhaps it is best not to use the extensions if you can
380 straightforwardly do without them, but to use the extensions if they
381 are a big improvement.
383 An exception to this rule are the large, established programs (such as
384 Emacs) which run on a great variety of systems. Using GNU extensions in
385 such programs would make many users unhappy, so we don't do that.
387 Another exception is for programs that are used as part of compilation:
388 anything that must be compiled with other compilers in order to
389 bootstrap the GNU compilation facilities. If these require the GNU
390 compiler, then no one can compile them without having them installed
391 already. That would be extremely troublesome in certain cases.
394 @section Standard C and Pre-Standard C
395 @cindex ANSI C standard
397 1989 Standard C is widespread enough now that it is ok to use its
398 features in programs. There is one exception: do not ever use the
399 ``trigraph'' feature of Standard C.
401 The 1999 and 2011 editions of Standard C are not fully supported
402 on all platforms. If you aim to support compilation by
403 compilers other than GCC, you should not require these C
404 features in your programs. It is ok to use these features
405 conditionally when the compiler supports them.
407 If your program is only meant to compile with GCC, then you can
408 use these features if GCC supports them, when they give substantial
411 However, it is easy to support pre-standard compilers in most programs,
412 so if you know how to do that, feel free.
414 @cindex function prototypes
415 To support pre-standard C, instead of writing function definitions in
416 standard prototype form,
425 write the definition in pre-standard style like this,
435 and use a separate declaration to specify the argument prototype:
441 You need such a declaration anyway, in a header file, to get the benefit
442 of prototypes in all the files where the function is called. And once
443 you have the declaration, you normally lose nothing by writing the
444 function definition in the pre-standard style.
446 This technique does not work for integer types narrower than @code{int}.
447 If you think of an argument as being of a type narrower than @code{int},
448 declare it as @code{int} instead.
450 There are a few special cases where this technique is hard to use. For
451 example, if a function argument needs to hold the system type
452 @code{dev_t}, you run into trouble, because @code{dev_t} is shorter than
453 @code{int} on some machines; but you cannot use @code{int} instead,
454 because @code{dev_t} is wider than @code{int} on some machines. There
455 is no type you can safely use on all machines in a non-standard
456 definition. The only way to support non-standard C and pass such an
457 argument is to check the width of @code{dev_t} using Autoconf and choose
458 the argument type accordingly. This may not be worth the trouble.
460 In order to support pre-standard compilers that do not recognize
461 prototypes, you may want to use a preprocessor macro like this:
464 /* Declare the prototype for a general external function. */
465 #if defined (__STDC__) || defined (WINDOWSNT)
466 #define P_(proto) proto
472 @node Conditional Compilation
473 @section Conditional Compilation
475 When supporting configuration options already known when building your
476 program we prefer using @code{if (... )} over conditional compilation,
477 as in the former case the compiler is able to perform more extensive
478 checking of all possible code paths.
480 For example, please write
500 A modern compiler such as GCC will generate exactly the same code in
501 both cases, and we have been using similar techniques with good success
502 in several projects. Of course, the former method assumes that
503 @code{HAS_FOO} is defined as either 0 or 1.
505 While this is not a silver bullet solving all portability problems,
506 and is not always appropriate, following this policy would have saved
507 GCC developers many hours, or even days, per year.
509 In the case of function-like macros like @code{REVERSIBLE_CC_MODE} in
510 GCC which cannot be simply used in @code{if (...)} statements, there is
511 an easy workaround. Simply introduce another macro
512 @code{HAS_REVERSIBLE_CC_MODE} as in the following example:
515 #ifdef REVERSIBLE_CC_MODE
516 #define HAS_REVERSIBLE_CC_MODE 1
518 #define HAS_REVERSIBLE_CC_MODE 0
522 @node Program Behavior
523 @chapter Program Behavior for All Programs
525 This chapter describes conventions for writing robust
526 software. It also describes general standards for error messages, the
527 command line interface, and how libraries should behave.
530 * Non-GNU Standards:: We consider standards such as POSIX;
531 we don't "obey" them.
532 * Semantics:: Writing robust programs.
533 * Libraries:: Library behavior.
534 * Errors:: Formatting error messages.
535 * User Interfaces:: Standards about interfaces generally.
536 * Graphical Interfaces:: Standards for graphical interfaces.
537 * Command-Line Interfaces:: Standards for command line interfaces.
538 * Dynamic Plug-In Interfaces:: Standards for dynamic plug-in interfaces.
539 * Option Table:: Table of long options.
540 * OID Allocations:: Table of OID slots for GNU.
541 * Memory Usage:: When and how to care about memory needs.
542 * File Usage:: Which files to use, and where.
545 @node Non-GNU Standards
546 @section Non-GNU Standards
548 The GNU Project regards standards published by other organizations as
549 suggestions, not orders. We consider those standards, but we do not
550 ``obey'' them. In developing a GNU program, you should implement
551 an outside standard's specifications when that makes the GNU system
552 better overall in an objective sense. When it doesn't, you shouldn't.
554 In most cases, following published standards is convenient for
555 users---it means that their programs or scripts will work more
556 portably. For instance, GCC implements nearly all the features of
557 Standard C as specified by that standard. C program developers would
558 be unhappy if it did not. And GNU utilities mostly follow
559 specifications of POSIX.2; shell script writers and users would be
560 unhappy if our programs were incompatible.
562 But we do not follow either of these specifications rigidly, and there
563 are specific points on which we decided not to follow them, so as to
564 make the GNU system better for users.
566 For instance, Standard C says that nearly all extensions to C are
567 prohibited. How silly! GCC implements many extensions, some of which
568 were later adopted as part of the standard. If you want these
569 constructs to give an error message as ``required'' by the standard,
570 you must specify @samp{--pedantic}, which was implemented only so that
571 we can say ``GCC is a 100% implementation of the standard'', not
572 because there is any reason to actually use it.
574 POSIX.2 specifies that @samp{df} and @samp{du} must output sizes by
575 default in units of 512 bytes. What users want is units of 1k, so
576 that is what we do by default. If you want the ridiculous behavior
577 ``required'' by POSIX, you must set the environment variable
578 @samp{POSIXLY_CORRECT} (which was originally going to be named
579 @samp{POSIX_ME_HARDER}).
581 GNU utilities also depart from the letter of the POSIX.2 specification
582 when they support long-named command-line options, and intermixing
583 options with ordinary arguments. This minor incompatibility with
584 POSIX is never a problem in practice, and it is very useful.
586 In particular, don't reject a new feature, or remove an old one,
587 merely because a standard says it is ``forbidden'' or ``deprecated''.
591 @section Writing Robust Programs
593 @cindex arbitrary limits on data
594 Avoid arbitrary limits on the length or number of @emph{any} data
595 structure, including file names, lines, files, and symbols, by allocating
596 all data structures dynamically. In most Unix utilities, ``long lines
597 are silently truncated''. This is not acceptable in a GNU utility.
599 @cindex @code{NUL} characters
601 Utilities reading files should not drop NUL characters, or any other
602 nonprinting characters @emph{including those with codes above 0177}.
603 The only sensible exceptions would be utilities specifically intended
604 for interface to certain types of terminals or printers that can't
605 handle those characters. Whenever possible, try to make programs work
606 properly with sequences of bytes that represent multibyte characters;
607 UTF-8 is the most important.
609 @cindex error messages
610 Check every system call for an error return, unless you know you wish
611 to ignore errors. Include the system error text (from @code{perror},
612 @code{strerror}, or equivalent) in @emph{every} error message
613 resulting from a failing system call, as well as the name of the file
614 if any and the name of the utility. Just ``cannot open foo.c'' or
615 ``stat failed'' is not sufficient.
617 @cindex @code{malloc} return value
618 @cindex memory allocation failure
619 Check every call to @code{malloc} or @code{realloc} to see if it
620 returned zero. Check @code{realloc} even if you are making the block
621 smaller; in a system that rounds block sizes to a power of 2,
622 @code{realloc} may get a different block if you ask for less space.
624 In Unix, @code{realloc} can destroy the storage block if it returns
625 zero. GNU @code{realloc} does not have this bug: if it fails, the
626 original block is unchanged. Feel free to assume the bug is fixed. If
627 you wish to run your program on Unix, and wish to avoid lossage in this
628 case, you can use the GNU @code{malloc}.
630 You must expect @code{free} to alter the contents of the block that was
631 freed. Anything you want to fetch from the block, you must fetch before
634 If @code{malloc} fails in a noninteractive program, make that a fatal
635 error. In an interactive program (one that reads commands from the
636 user), it is better to abort the command and return to the command
637 reader loop. This allows the user to kill other processes to free up
638 virtual memory, and then try the command again.
640 @cindex command-line arguments, decoding
641 Use @code{getopt_long} to decode arguments, unless the argument syntax
642 makes this unreasonable.
644 When static storage is to be written in during program execution, use
645 explicit C code to initialize it. Reserve C initialized declarations
646 for data that will not be changed.
649 Try to avoid low-level interfaces to obscure Unix data structures (such
650 as file directories, utmp, or the layout of kernel memory), since these
651 are less likely to work compatibly. If you need to find all the files
652 in a directory, use @code{readdir} or some other high-level interface.
653 These are supported compatibly by GNU.
655 @cindex signal handling
656 The preferred signal handling facilities are the BSD variant of
657 @code{signal}, and the POSIX @code{sigaction} function; the
658 alternative USG @code{signal} interface is an inferior design.
660 Nowadays, using the POSIX signal functions may be the easiest way
661 to make a program portable. If you use @code{signal}, then on GNU/Linux
662 systems running GNU libc version 1, you should include
663 @file{bsd/signal.h} instead of @file{signal.h}, so as to get BSD
664 behavior. It is up to you whether to support systems where
665 @code{signal} has only the USG behavior, or give up on them.
667 @cindex impossible conditions
668 In error checks that detect ``impossible'' conditions, just abort.
669 There is usually no point in printing any message. These checks
670 indicate the existence of bugs. Whoever wants to fix the bugs will have
671 to read the source code and run a debugger. So explain the problem with
672 comments in the source. The relevant data will be in variables, which
673 are easy to examine with the debugger, so there is no point moving them
676 Do not use a count of errors as the exit status for a program.
677 @emph{That does not work}, because exit status values are limited to 8
678 bits (0 through 255). A single run of the program might have 256
679 errors; if you try to return 256 as the exit status, the parent process
680 will see 0 as the status, and it will appear that the program succeeded.
682 @cindex temporary files
683 @cindex @code{TMPDIR} environment variable
684 If you make temporary files, check the @code{TMPDIR} environment
685 variable; if that variable is defined, use the specified directory
686 instead of @file{/tmp}.
688 In addition, be aware that there is a possible security problem when
689 creating temporary files in world-writable directories. In C, you can
690 avoid this problem by creating temporary files in this manner:
693 fd = open (filename, O_WRONLY | O_CREAT | O_EXCL, 0600);
697 or by using the @code{mkstemps} function from Gnulib
698 (@pxref{mkstemps,,, gnulib, Gnulib}).
700 In bash, use @code{set -C} (long name @code{noclobber}) to avoid this
701 problem. In addition, the @code{mktemp} utility is a more general
702 solution for creating temporary files from shell scripts
703 (@pxref{mktemp invocation,,, coreutils, GNU Coreutils}).
707 @section Library Behavior
710 Try to make library functions reentrant. If they need to do dynamic
711 storage allocation, at least try to avoid any nonreentrancy aside from
712 that of @code{malloc} itself.
714 Here are certain name conventions for libraries, to avoid name
717 Choose a name prefix for the library, more than two characters long.
718 All external function and variable names should start with this
719 prefix. In addition, there should only be one of these in any given
720 library member. This usually means putting each one in a separate
723 An exception can be made when two external symbols are always used
724 together, so that no reasonable program could use one without the
725 other; then they can both go in the same file.
727 External symbols that are not documented entry points for the user
728 should have names beginning with @samp{_}. The @samp{_} should be
729 followed by the chosen name prefix for the library, to prevent
730 collisions with other libraries. These can go in the same files with
731 user entry points if you like.
733 Static functions and variables can be used as you like and need not
734 fit any naming convention.
737 @section Formatting Error Messages
738 @cindex formatting error messages
739 @cindex error messages, formatting
741 Error messages from compilers should look like this:
744 @var{sourcefile}:@var{lineno}: @var{message}
748 If you want to mention the column number, use one of these formats:
751 @var{sourcefile}:@var{lineno}:@var{column}: @var{message}
752 @var{sourcefile}:@var{lineno}.@var{column}: @var{message}
757 Line numbers should start from 1 at the beginning of the file, and
758 column numbers should start from 1 at the beginning of the line.
759 (Both of these conventions are chosen for compatibility.) Calculate
760 column numbers assuming that space and all ASCII printing characters
761 have equal width, and assuming tab stops every 8 columns. For
762 non-ASCII characters, Unicode character widths should be used when in
763 a UTF-8 locale; GNU libc and GNU gnulib provide suitable
764 @code{wcwidth} functions.
766 The error message can also give both the starting and ending positions
767 of the erroneous text. There are several formats so that you can
768 avoid redundant information such as a duplicate line number.
769 Here are the possible formats:
772 @var{sourcefile}:@var{line1}.@var{column1}-@var{line2}.@var{column2}: @var{message}
773 @var{sourcefile}:@var{line1}.@var{column1}-@var{column2}: @var{message}
774 @var{sourcefile}:@var{line1}-@var{line2}: @var{message}
778 When an error is spread over several files, you can use this format:
781 @var{file1}:@var{line1}.@var{column1}-@var{file2}:@var{line2}.@var{column2}: @var{message}
784 Error messages from other noninteractive programs should look like this:
787 @var{program}:@var{sourcefile}:@var{lineno}: @var{message}
791 when there is an appropriate source file, or like this:
794 @var{program}: @var{message}
798 when there is no relevant source file.
800 If you want to mention the column number, use this format:
803 @var{program}:@var{sourcefile}:@var{lineno}:@var{column}: @var{message}
806 In an interactive program (one that is reading commands from a
807 terminal), it is better not to include the program name in an error
808 message. The place to indicate which program is running is in the
809 prompt or with the screen layout. (When the same program runs with
810 input from a source other than a terminal, it is not interactive and
811 would do best to print error messages using the noninteractive style.)
813 The string @var{message} should not begin with a capital letter when
814 it follows a program name and/or file name, because that isn't the
815 beginning of a sentence. (The sentence conceptually starts at the
816 beginning of the line.) Also, it should not end with a period.
818 Error messages from interactive programs, and other messages such as
819 usage messages, should start with a capital letter. But they should not
822 @node User Interfaces
823 @section Standards for Interfaces Generally
825 @cindex program name and its behavior
826 @cindex behavior, dependent on program's name
827 Please don't make the behavior of a utility depend on the name used
828 to invoke it. It is useful sometimes to make a link to a utility
829 with a different name, and that should not change what it does.
831 Instead, use a run time option or a compilation switch or both
832 to select among the alternate behaviors.
834 @cindex output device and program's behavior
835 Likewise, please don't make the behavior of the program depend on the
836 type of output device it is used with. Device independence is an
837 important principle of the system's design; do not compromise it merely
838 to save someone from typing an option now and then. (Variation in error
839 message syntax when using a terminal is ok, because that is a side issue
840 that people do not depend on.)
842 If you think one behavior is most useful when the output is to a
843 terminal, and another is most useful when the output is a file or a
844 pipe, then it is usually best to make the default behavior the one that
845 is useful with output to a terminal, and have an option for the other
848 Compatibility requires certain programs to depend on the type of output
849 device. It would be disastrous if @code{ls} or @code{sh} did not do so
850 in the way all users expect. In some of these cases, we supplement the
851 program with a preferred alternate version that does not depend on the
852 output device type. For example, we provide a @code{dir} program much
853 like @code{ls} except that its default output format is always
857 @node Graphical Interfaces
858 @section Standards for Graphical Interfaces
859 @cindex graphical user interface
860 @cindex interface styles
861 @cindex user interface styles
864 When you write a program that provides a graphical user interface,
865 please make it work with the X Window System and the GTK+ toolkit
866 unless the functionality specifically requires some alternative (for
867 example, ``displaying jpeg images while in console mode'').
869 In addition, please provide a command-line interface to control the
870 functionality. (In many cases, the graphical user interface can be a
871 separate program which invokes the command-line program.) This is
872 so that the same jobs can be done from scripts.
877 @cindex keyboard interface
878 @cindex library interface
879 Please also consider providing a D-bus interface for use from other
880 running programs, such as within GNOME. (GNOME used to use CORBA
881 for this, but that is being phased out.) In addition, consider
882 providing a library interface (for use from C), and perhaps a
883 keyboard-driven console interface (for use by users from console
884 mode). Once you are doing the work to provide the functionality and
885 the graphical interface, these won't be much extra work.
887 @node Command-Line Interfaces
888 @section Standards for Command Line Interfaces
889 @cindex command-line interface
892 It is a good idea to follow the POSIX guidelines for the
893 command-line options of a program. The easiest way to do this is to use
894 @code{getopt} to parse them. Note that the GNU version of @code{getopt}
895 will normally permit options anywhere among the arguments unless the
896 special argument @samp{--} is used. This is not what POSIX
897 specifies; it is a GNU extension.
899 @cindex long-named options
900 Please define long-named options that are equivalent to the
901 single-letter Unix-style options. We hope to make GNU more user
902 friendly this way. This is easy to do with the GNU function
905 One of the advantages of long-named options is that they can be
906 consistent from program to program. For example, users should be able
907 to expect the ``verbose'' option of any GNU program which has one, to be
908 spelled precisely @samp{--verbose}. To achieve this uniformity, look at
909 the table of common long-option names when you choose the option names
910 for your program (@pxref{Option Table}).
912 It is usually a good idea for file names given as ordinary arguments to
913 be input files only; any output files would be specified using options
914 (preferably @samp{-o} or @samp{--output}). Even if you allow an output
915 file name as an ordinary argument for compatibility, try to provide an
916 option as another way to specify it. This will lead to more consistency
917 among GNU utilities, and fewer idiosyncrasies for users to remember.
919 @cindex standard command-line options
920 @cindex options, standard command-line
921 @cindex CGI programs, standard options for
922 @cindex PATH_INFO, specifying standard options as
923 All programs should support two standard options: @samp{--version}
924 and @samp{--help}. CGI programs should accept these as command-line
925 options, and also if given as the @env{PATH_INFO}; for instance,
926 visiting @url{http://example.org/p.cgi/--help} in a browser should
927 output the same information as invoking @samp{p.cgi --help} from the
931 * --version:: The standard output for --version.
932 * --help:: The standard output for --help.
936 @subsection @option{--version}
938 @cindex @samp{--version} output
940 The standard @code{--version} option should direct the program to
941 print information about its name, version, origin and legal status,
942 all on standard output, and then exit successfully. Other options and
943 arguments should be ignored once this is seen, and the program should
944 not perform its normal function.
946 @cindex canonical name of a program
947 @cindex program's canonical name
948 The first line is meant to be easy for a program to parse; the version
949 number proper starts after the last space. In addition, it contains
950 the canonical name for this program, in this format:
957 The program's name should be a constant string; @emph{don't} compute it
958 from @code{argv[0]}. The idea is to state the standard or canonical
959 name for the program, not its file name. There are other ways to find
960 out the precise file name where a command is found in @code{PATH}.
962 If the program is a subsidiary part of a larger package, mention the
963 package name in parentheses, like this:
966 emacsserver (GNU Emacs) 19.30
970 If the package has a version number which is different from this
971 program's version number, you can mention the package version number
972 just before the close-parenthesis.
974 If you @emph{need} to mention the version numbers of libraries which
975 are distributed separately from the package which contains this program,
976 you can do so by printing an additional line of version info for each
977 library you want to mention. Use the same format for these lines as for
980 Please do not mention all of the libraries that the program uses ``just
981 for completeness''---that would produce a lot of unhelpful clutter.
982 Please mention library version numbers only if you find in practice that
983 they are very important to you in debugging.
985 The following line, after the version number line or lines, should be a
986 copyright notice. If more than one copyright notice is called for, put
987 each on a separate line.
989 Next should follow a line stating the license, preferably using one of
990 abbreviations below, and a brief statement that the program is free
991 software, and that users are free to copy and change it. Also mention
992 that there is no warranty, to the extent permitted by law. See
993 recommended wording below.
995 It is ok to finish the output with a list of the major authors of the
996 program, as a way of giving credit.
998 Here's an example of output that follows these rules:
1002 Copyright (C) 2007 Free Software Foundation, Inc.
1003 License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
1004 This is free software: you are free to change and redistribute it.
1005 There is NO WARRANTY, to the extent permitted by law.
1008 You should adapt this to your program, of course, filling in the proper
1009 year, copyright holder, name of program, and the references to
1010 distribution terms, and changing the rest of the wording as necessary.
1012 This copyright notice only needs to mention the most recent year in
1013 which changes were made---there's no need to list the years for previous
1014 versions' changes. You don't have to mention the name of the program in
1015 these notices, if that is inconvenient, since it appeared in the first
1016 line. (The rules are different for copyright notices in source files;
1017 @pxref{Copyright Notices,,,maintain,Information for GNU Maintainers}.)
1019 Translations of the above lines must preserve the validity of the
1020 copyright notices (@pxref{Internationalization}). If the translation's
1021 character set supports it, the @samp{(C)} should be replaced with the
1022 copyright symbol, as follows:
1025 (the official copyright symbol, which is the letter C in a circle);
1031 Write the word ``Copyright'' exactly like that, in English. Do not
1032 translate it into another language. International treaties recognize
1033 the English word ``Copyright''; translations into other languages do not
1034 have legal significance.
1036 Finally, here is the table of our suggested license abbreviations.
1037 Any abbreviation can be followed by @samp{v@var{version}[+]}, meaning
1038 that particular version, or later versions with the @samp{+}, as shown
1039 above. In the case of a GNU license, @emph{always} indicate the permitted
1040 versions in this way.
1042 In the case of exceptions for extra permissions with the GPL, we use
1043 @samp{/} for a separator; the version number can follow the license
1044 abbreviation as usual, as in the examples below.
1048 GNU General Public License, @url{http://www.gnu.org/@/licenses/@/gpl.html}.
1051 GNU Lesser General Public License, @url{http://www.gnu.org/@/licenses/@/lgpl.html}.
1054 GNU GPL with the exception for Ada.
1057 The Apache Software Foundation license,
1058 @url{http://www.apache.org/@/licenses}.
1061 The Artistic license used for Perl, @url{http://dev.perl.org/licenses/artistic.html}.
1064 The Expat license, @url{http://www.jclark.com/@/xml/@/copying.txt}.
1067 The Mozilla Public License, @url{http://www.mozilla.org/@/MPL/}.
1070 The original (4-clause) BSD license, incompatible with the GNU GPL
1071 @url{http://www.xfree86.org/@/3.3.6/@/COPYRIGHT2.html#6}.
1074 The license used for PHP, @url{http://www.php.net/@/license/}.
1077 The non-license that is being in the public domain,
1078 @url{http://www.gnu.org/@/licenses/@/license-list.html#PublicDomain}.
1081 The license for Python, @url{http://www.python.org/@/2.0.1/@/license.html}.
1084 The revised (3-clause) BSD, compatible with the GNU GPL,@*
1085 @url{http://www.xfree86.org/@/3.3.6/@/COPYRIGHT2.html#5}.
1088 The simple non-copyleft license used for most versions of the X Window
1089 System, @url{http://www.xfree86.org/@/3.3.6/@/COPYRIGHT2.html#3}.
1092 The license for Zlib, @url{http://www.gzip.org/@/zlib/@/zlib_license.html}.
1096 More information about these licenses and many more are on the GNU
1097 licensing web pages,
1098 @url{http://www.gnu.org/@/licenses/@/license-list.html}.
1102 @subsection @option{--help}
1104 @cindex @samp{--help} output
1106 The standard @code{--help} option should output brief documentation
1107 for how to invoke the program, on standard output, then exit
1108 successfully. Other options and arguments should be ignored once this
1109 is seen, and the program should not perform its normal function.
1111 @cindex address for bug reports
1113 Near the end of the @samp{--help} option's output, please place lines
1114 giving the email address for bug reports, the package's home page
1115 (normally @indicateurl{http://www.gnu.org/software/@var{pkg}}, and the
1116 general page for help using GNU programs. The format should be like this:
1119 Report bugs to: @var{mailing-address}
1120 @var{pkg} home page: <http://www.gnu.org/software/@var{pkg}/>
1121 General help using GNU software: <http://www.gnu.org/gethelp/>
1124 It is ok to mention other appropriate mailing lists and web pages.
1127 @node Dynamic Plug-In Interfaces
1128 @section Standards for Dynamic Plug-in Interfaces
1130 @cindex dynamic plug-ins
1132 Another aspect of keeping free programs free is encouraging
1133 development of free plug-ins, and discouraging development of
1134 proprietary plug-ins. Many GNU programs will not have anything like
1135 plug-ins at all, but those that do should follow these
1138 First, the general plug-in architecture design should closely tie the
1139 plug-in to the original code, such that the plug-in and the base
1140 program are parts of one extended program. For GCC, for example,
1141 plug-ins receive and modify GCC's internal data structures, and so
1142 clearly form an extended program with the base GCC.
1144 @vindex plugin_is_GPL_compatible
1145 Second, you should require plug-in developers to affirm that their
1146 plug-ins are released under an appropriate license. This should be
1147 enforced with a simple programmatic check. For GCC, again for
1148 example, a plug-in must define the global symbol
1149 @code{plugin_is_GPL_compatible}, thus asserting that the plug-in is
1150 released under a GPL-compatible license (@pxref{Plugins,, Plugins,
1151 gccint, GCC Internals}).
1153 By adding this check to your program you are not creating a new legal
1154 requirement. The GPL itself requires plug-ins to be free software,
1155 licensed compatibly. As long as you have followed the first rule above
1156 to keep plug-ins closely tied to your original program, the GPL and AGPL
1157 already require those plug-ins to be released under a compatible
1158 license. The symbol definition in the plug-in---or whatever equivalent
1159 works best in your program---makes it harder for anyone who might
1160 distribute proprietary plug-ins to legally defend themselves. If a case
1161 about this got to court, we can point to that symbol as evidence that
1162 the plug-in developer understood that the license had this requirement.
1166 @section Table of Long Options
1167 @cindex long option names
1168 @cindex table of long options
1170 Here is a table of long options used by GNU programs. It is surely
1171 incomplete, but we aim to list all the options that a new program might
1172 want to be compatible with. If you use names not already in the table,
1173 please send @email{bug-standards@@gnu.org} a list of them, with their
1174 meanings, so we can update the table.
1176 @c Please leave newlines between items in this table; it's much easier
1177 @c to update when it isn't completely squashed together and unreadable.
1178 @c When there is more than one short option for a long option name, put
1179 @c a semicolon between the lists of the programs that use them, not a
1180 @c period. --friedman
1184 @samp{-N} in @code{tar}.
1187 @samp{-a} in @code{du}, @code{ls}, @code{nm}, @code{stty}, @code{uname},
1188 and @code{unexpand}.
1191 @samp{-a} in @code{diff}.
1194 @samp{-A} in @code{ls}.
1197 @samp{-a} in @code{etags}, @code{tee}, @code{time};
1198 @samp{-r} in @code{tar}.
1201 @samp{-a} in @code{cp}.
1204 @samp{-n} in @code{shar}.
1207 @samp{-l} in @code{m4}.
1210 @samp{-a} in @code{diff}.
1213 @samp{-v} in @code{gawk}.
1216 @samp{-W} in @code{make}.
1219 @samp{-o} in @code{make}.
1222 @samp{-a} in @code{recode}.
1225 @samp{-a} in @code{wdiff}.
1227 @item auto-reference
1228 @samp{-A} in @code{ptx}.
1231 @samp{-n} in @code{wdiff}.
1234 For server programs, run in the background.
1236 @item backward-search
1237 @samp{-B} in @code{ctags}.
1240 @samp{-f} in @code{shar}.
1249 @samp{-b} in @code{tac}.
1252 @samp{-b} in @code{cpio} and @code{diff}.
1255 @samp{-b} in @code{shar}.
1258 Used in @code{cpio} and @code{tar}.
1261 @samp{-b} in @code{head} and @code{tail}.
1264 @samp{-b} in @code{ptx}.
1267 Used in various programs to make output shorter.
1270 @samp{-c} in @code{head}, @code{split}, and @code{tail}.
1273 @samp{-C} in @code{etags}.
1276 @samp{-A} in @code{tar}.
1279 Used in various programs to specify the directory to use.
1282 @samp{-c} in @code{chgrp} and @code{chown}.
1285 @samp{-F} in @code{ls}.
1288 @samp{-c} in @code{recode}.
1291 @samp{-c} in @code{su};
1295 @samp{-d} in @code{tar}.
1298 Used in @code{gawk}.
1301 @samp{-Z} in @code{tar} and @code{shar}.
1304 @samp{-A} in @code{tar}.
1307 @samp{-w} in @code{tar}.
1310 Used in @code{diff}.
1313 @samp{-W copyleft} in @code{gawk}.
1316 @samp{-C} in @code{ptx}, @code{recode}, and @code{wdiff};
1317 @samp{-W copyright} in @code{gawk}.
1323 @samp{-q} in @code{who}.
1326 @samp{-l} in @code{du}.
1329 Used in @code{tar} and @code{cpio}.
1332 @samp{-c} in @code{shar}.
1335 @samp{-x} in @code{ctags}.
1338 @samp{-d} in @code{touch}.
1341 @samp{-d} in @code{make} and @code{m4};
1345 @samp{-D} in @code{m4}.
1348 @samp{-d} in Bison and @code{ctags}.
1351 @samp{-D} in @code{tar}.
1354 @samp{-L} in @code{chgrp}, @code{chown}, @code{cpio}, @code{du},
1355 @code{ls}, and @code{tar}.
1357 @item dereference-args
1358 @samp{-D} in @code{du}.
1361 Specify an I/O device (special file name).
1364 @samp{-d} in @code{recode}.
1366 @item dictionary-order
1367 @samp{-d} in @code{look}.
1370 @samp{-d} in @code{tar}.
1373 @samp{-n} in @code{csplit}.
1376 Specify the directory to use, in various programs. In @code{ls}, it
1377 means to show directories themselves rather than their contents. In
1378 @code{rm} and @code{ln}, it means to not treat links to directories
1382 @samp{-x} in @code{strip}.
1384 @item discard-locals
1385 @samp{-X} in @code{strip}.
1388 @samp{-n} in @code{make}.
1391 @samp{-e} in @code{diff}.
1393 @item elide-empty-files
1394 @samp{-z} in @code{csplit}.
1397 @samp{-x} in @code{wdiff}.
1400 @samp{-z} in @code{wdiff}.
1402 @item entire-new-file
1403 @samp{-N} in @code{diff}.
1405 @item environment-overrides
1406 @samp{-e} in @code{make}.
1409 @samp{-e} in @code{xargs}.
1415 Used in @code{makeinfo}.
1418 @samp{-o} in @code{m4}.
1421 @samp{-b} in @code{ls}.
1424 @samp{-X} in @code{tar}.
1430 @samp{-x} in @code{xargs}.
1433 @samp{-e} in @code{unshar}.
1436 @samp{-t} in @code{diff}.
1439 @samp{-e} in @code{sed}.
1442 @samp{-g} in @code{nm}.
1445 @samp{-i} in @code{cpio};
1446 @samp{-x} in @code{tar}.
1449 @samp{-f} in @code{finger}.
1452 @samp{-f} in @code{su}.
1454 @item fatal-warnings
1455 @samp{-E} in @code{m4}.
1458 @samp{-f} in @code{gawk}, @code{info}, @code{make}, @code{mt},
1459 @code{sed}, and @code{tar}.
1461 @item field-separator
1462 @samp{-F} in @code{gawk}.
1468 @samp{-F} in @code{ls}.
1471 @samp{-T} in @code{tar}.
1474 Used in @code{makeinfo}.
1476 @item flag-truncation
1477 @samp{-F} in @code{ptx}.
1479 @item fixed-output-files
1483 @samp{-f} in @code{tail}.
1485 @item footnote-style
1486 Used in @code{makeinfo}.
1489 @samp{-f} in @code{cp}, @code{ln}, @code{mv}, and @code{rm}.
1492 @samp{-F} in @code{shar}.
1495 For server programs, run in the foreground;
1496 in other words, don't do anything special to run the server
1500 Used in @code{ls}, @code{time}, and @code{ptx}.
1503 @samp{-F} in @code{m4}.
1509 @samp{-g} in @code{ptx}.
1512 @samp{-x} in @code{tar}.
1515 @samp{-i} in @code{ul}.
1518 @samp{-g} in @code{recode}.
1521 @samp{-g} in @code{install}.
1524 @samp{-z} in @code{tar} and @code{shar}.
1527 @samp{-H} in @code{m4}.
1530 @samp{-h} in @code{objdump} and @code{recode}
1533 @samp{-H} in @code{who}.
1536 Used to ask for brief usage information.
1538 @item here-delimiter
1539 @samp{-d} in @code{shar}.
1541 @item hide-control-chars
1542 @samp{-q} in @code{ls}.
1545 In @code{makeinfo}, output HTML.
1548 @samp{-u} in @code{who}.
1551 @samp{-D} in @code{diff}.
1554 @samp{-I} in @code{ls};
1555 @samp{-x} in @code{recode}.
1557 @item ignore-all-space
1558 @samp{-w} in @code{diff}.
1560 @item ignore-backups
1561 @samp{-B} in @code{ls}.
1563 @item ignore-blank-lines
1564 @samp{-B} in @code{diff}.
1567 @samp{-f} in @code{look} and @code{ptx};
1568 @samp{-i} in @code{diff} and @code{wdiff}.
1571 @samp{-i} in @code{make}.
1574 @samp{-i} in @code{ptx}.
1576 @item ignore-indentation
1577 @samp{-I} in @code{etags}.
1579 @item ignore-init-file
1582 @item ignore-interrupts
1583 @samp{-i} in @code{tee}.
1585 @item ignore-matching-lines
1586 @samp{-I} in @code{diff}.
1588 @item ignore-space-change
1589 @samp{-b} in @code{diff}.
1592 @samp{-i} in @code{tar}.
1595 @samp{-i} in @code{etags};
1596 @samp{-I} in @code{m4}.
1599 @samp{-I} in @code{make}.
1602 @samp{-G} in @code{tar}.
1605 @samp{-i}, @samp{-l}, and @samp{-m} in Finger.
1608 In some programs, specify the name of the file to read as the user's
1612 @samp{-i} in @code{expand}.
1615 @samp{-T} in @code{diff}.
1618 @samp{-i} in @code{ls}.
1621 @samp{-i} in @code{cp}, @code{ln}, @code{mv}, @code{rm};
1622 @samp{-e} in @code{m4};
1623 @samp{-p} in @code{xargs};
1624 @samp{-w} in @code{tar}.
1627 @samp{-p} in @code{shar}.
1633 @samp{-j} in @code{make}.
1636 @samp{-n} in @code{make}.
1639 @samp{-k} in @code{make}.
1642 @samp{-k} in @code{csplit}.
1645 @samp{-k} in @code{du} and @code{ls}.
1648 @samp{-l} in @code{etags}.
1651 @samp{-l} in @code{wdiff}.
1653 @item level-for-gzip
1654 @samp{-g} in @code{shar}.
1657 @samp{-C} in @code{split}.
1660 Used in @code{split}, @code{head}, and @code{tail}.
1663 @samp{-l} in @code{cpio}.
1667 Used in @code{gawk}.
1670 @samp{-t} in @code{cpio};
1671 @samp{-l} in @code{recode}.
1674 @samp{-t} in @code{tar}.
1677 @samp{-N} in @code{ls}.
1680 @samp{-l} in @code{make}.
1686 Used in @code{uname}.
1689 @samp{-M} in @code{ptx}.
1692 @samp{-m} in @code{hello} and @code{uname}.
1694 @item make-directories
1695 @samp{-d} in @code{cpio}.
1698 @samp{-f} in @code{make}.
1704 @samp{-n} in @code{xargs}.
1707 @samp{-n} in @code{xargs}.
1710 @samp{-l} in @code{xargs}.
1713 @samp{-l} in @code{make}.
1716 @samp{-P} in @code{xargs}.
1719 @samp{-T} in @code{who}.
1722 @samp{-T} in @code{who}.
1725 @samp{-d} in @code{diff}.
1727 @item mixed-uuencode
1728 @samp{-M} in @code{shar}.
1731 @samp{-m} in @code{install}, @code{mkdir}, and @code{mkfifo}.
1733 @item modification-time
1734 @samp{-m} in @code{tar}.
1737 @samp{-M} in @code{tar}.
1743 @samp{-L} in @code{m4}.
1746 @samp{-a} in @code{shar}.
1749 @samp{-W} in @code{make}.
1751 @item no-builtin-rules
1752 @samp{-r} in @code{make}.
1754 @item no-character-count
1755 @samp{-w} in @code{shar}.
1757 @item no-check-existing
1758 @samp{-x} in @code{shar}.
1761 @samp{-3} in @code{wdiff}.
1764 @samp{-c} in @code{touch}.
1767 @samp{-D} in @code{etags}.
1770 @samp{-1} in @code{wdiff}.
1772 @item no-dereference
1773 @samp{-d} in @code{cp}.
1776 @samp{-2} in @code{wdiff}.
1779 @samp{-S} in @code{make}.
1785 @samp{-P} in @code{shar}.
1788 @samp{-e} in @code{gprof}.
1791 @samp{-R} in @code{etags}.
1794 @samp{-p} in @code{nm}.
1797 Don't print a startup splash screen.
1800 Used in @code{makeinfo}.
1803 @samp{-a} in @code{gprof}.
1806 @samp{-E} in @code{gprof}.
1809 @samp{-m} in @code{shar}.
1812 Used in @code{makeinfo}.
1815 Used in @code{emacsclient}.
1818 Used in various programs to inhibit warnings.
1821 @samp{-n} in @code{info}.
1824 @samp{-n} in @code{uname}.
1827 @samp{-f} in @code{cpio}.
1830 @samp{-n} in @code{objdump}.
1833 @samp{-0} in @code{xargs}.
1836 @samp{-n} in @code{cat}.
1838 @item number-nonblank
1839 @samp{-b} in @code{cat}.
1842 @samp{-n} in @code{nm}.
1844 @item numeric-uid-gid
1845 @samp{-n} in @code{cpio} and @code{ls}.
1851 @samp{-o} in @code{tar}.
1854 @samp{-o} in @code{make}.
1856 @item one-file-system
1857 @samp{-l} in @code{tar}, @code{cp}, and @code{du}.
1860 @samp{-o} in @code{ptx}.
1863 @samp{-f} in @code{gprof}.
1866 @samp{-F} in @code{gprof}.
1869 @samp{-o} in @code{getopt}, @code{fdlist}, @code{fdmount},
1870 @code{fdmountd}, and @code{fdumount}.
1873 In various programs, specify the output file name.
1876 @samp{-o} in @code{shar}.
1879 @samp{-o} in @code{rm}.
1882 @samp{-c} in @code{unshar}.
1885 @samp{-o} in @code{install}.
1888 @samp{-l} in @code{diff}.
1890 @item paragraph-indent
1891 Used in @code{makeinfo}.
1894 @samp{-p} in @code{mkdir} and @code{rmdir}.
1897 @samp{-p} in @code{ul}.
1900 @samp{-p} in @code{cpio}.
1903 @samp{-P} in @code{finger}.
1906 @samp{-c} in @code{cpio} and @code{tar}.
1909 Used in @code{gawk}.
1911 @item prefix-builtins
1912 @samp{-P} in @code{m4}.
1915 @samp{-f} in @code{csplit}.
1918 Used in @code{tar} and @code{cp}.
1920 @item preserve-environment
1921 @samp{-p} in @code{su}.
1923 @item preserve-modification-time
1924 @samp{-m} in @code{cpio}.
1926 @item preserve-order
1927 @samp{-s} in @code{tar}.
1929 @item preserve-permissions
1930 @samp{-p} in @code{tar}.
1933 @samp{-l} in @code{diff}.
1936 @samp{-L} in @code{cmp}.
1938 @item print-data-base
1939 @samp{-p} in @code{make}.
1941 @item print-directory
1942 @samp{-w} in @code{make}.
1944 @item print-file-name
1945 @samp{-o} in @code{nm}.
1948 @samp{-s} in @code{nm}.
1951 @samp{-p} in @code{wdiff}.
1954 @samp{-p} in @code{ed}.
1957 Specify an HTTP proxy.
1960 @samp{-X} in @code{shar}.
1963 @samp{-q} in @code{make}.
1966 Used in many programs to inhibit the usual output. Every
1967 program accepting @samp{--quiet} should accept @samp{--silent} as a
1971 @samp{-Q} in @code{shar}
1974 @samp{-Q} in @code{ls}.
1977 @samp{-n} in @code{diff}.
1980 Used in @code{gawk}.
1982 @item read-full-blocks
1983 @samp{-B} in @code{tar}.
1989 @samp{-n} in @code{make}.
1992 @samp{-R} in @code{tar}.
1995 Used in @code{chgrp}, @code{chown}, @code{cp}, @code{ls}, @code{diff},
1999 @samp{-r} in @code{touch}.
2002 @samp{-r} in @code{ptx}.
2005 @samp{-r} in @code{tac} and @code{etags}.
2008 @samp{-r} in @code{uname}.
2011 @samp{-R} in @code{m4}.
2014 @samp{-r} in @code{objdump}.
2017 @samp{-r} in @code{cpio}.
2020 @samp{-i} in @code{xargs}.
2022 @item report-identical-files
2023 @samp{-s} in @code{diff}.
2025 @item reset-access-time
2026 @samp{-a} in @code{cpio}.
2029 @samp{-r} in @code{ls} and @code{nm}.
2032 @samp{-f} in @code{diff}.
2034 @item right-side-defs
2035 @samp{-R} in @code{ptx}.
2038 @samp{-s} in @code{tar}.
2040 @item same-permissions
2041 @samp{-p} in @code{tar}.
2044 @samp{-g} in @code{stty}.
2049 @item sentence-regexp
2050 @samp{-S} in @code{ptx}.
2053 @samp{-S} in @code{du}.
2056 @samp{-s} in @code{tac}.
2059 Used by @code{recode} to chose files or pipes for sequencing passes.
2062 @samp{-s} in @code{su}.
2065 @samp{-A} in @code{cat}.
2067 @item show-c-function
2068 @samp{-p} in @code{diff}.
2071 @samp{-E} in @code{cat}.
2073 @item show-function-line
2074 @samp{-F} in @code{diff}.
2077 @samp{-T} in @code{cat}.
2080 Used in many programs to inhibit the usual output.
2081 Every program accepting
2082 @samp{--silent} should accept @samp{--quiet} as a synonym.
2085 @samp{-s} in @code{ls}.
2088 Specify a file descriptor for a network server to use for its socket,
2089 instead of opening and binding a new socket. This provides a way to
2090 run, in a non-privileged process, a server that normally needs a
2091 reserved port number.
2097 @samp{-W source} in @code{gawk}.
2100 @samp{-S} in @code{tar}.
2102 @item speed-large-files
2103 @samp{-H} in @code{diff}.
2106 @samp{-E} in @code{unshar}.
2108 @item split-size-limit
2109 @samp{-L} in @code{shar}.
2112 @samp{-s} in @code{cat}.
2115 @samp{-w} in @code{wdiff}.
2118 @samp{-y} in @code{wdiff}.
2121 Used in @code{tar} and @code{diff} to specify which file within
2122 a directory to start processing with.
2125 @samp{-s} in @code{wdiff}.
2127 @item stdin-file-list
2128 @samp{-S} in @code{shar}.
2131 @samp{-S} in @code{make}.
2134 @samp{-s} in @code{recode}.
2137 @samp{-s} in @code{install}.
2140 @samp{-s} in @code{strip}.
2143 @samp{-S} in @code{strip}.
2146 @samp{-s} in @code{shar}.
2149 @samp{-S} in @code{cp}, @code{ln}, @code{mv}.
2152 @samp{-b} in @code{csplit}.
2155 @samp{-s} in @code{gprof}.
2158 @samp{-s} in @code{du}.
2161 @samp{-s} in @code{ln}.
2164 Used in GDB and @code{objdump}.
2167 @samp{-s} in @code{m4}.
2170 @samp{-s} in @code{uname}.
2173 @samp{-t} in @code{expand} and @code{unexpand}.
2176 @samp{-T} in @code{ls}.
2179 @samp{-T} in @code{tput} and @code{ul}.
2180 @samp{-t} in @code{wdiff}.
2183 @samp{-a} in @code{diff}.
2186 @samp{-T} in @code{shar}.
2189 Used in @code{ls} and @code{touch}.
2192 Specify how long to wait before giving up on some operation.
2195 @samp{-O} in @code{tar}.
2198 @samp{-c} in @code{du}.
2201 @samp{-t} in @code{make}, @code{ranlib}, and @code{recode}.
2204 @samp{-t} in @code{m4}.
2207 @samp{-t} in @code{hello};
2208 @samp{-W traditional} in @code{gawk};
2209 @samp{-G} in @code{ed}, @code{m4}, and @code{ptx}.
2215 @samp{-t} in @code{ctags}.
2217 @item typedefs-and-c++
2218 @samp{-T} in @code{ctags}.
2221 @samp{-t} in @code{ptx}.
2224 @samp{-z} in @code{tar}.
2227 @samp{-u} in @code{cpio}.
2230 @samp{-U} in @code{m4}.
2232 @item undefined-only
2233 @samp{-u} in @code{nm}.
2236 @samp{-u} in @code{cp}, @code{ctags}, @code{mv}, @code{tar}.
2239 Used in @code{gawk}; same as @samp{--help}.
2242 @samp{-B} in @code{shar}.
2244 @item vanilla-operation
2245 @samp{-V} in @code{shar}.
2248 Print more information about progress. Many programs support this.
2251 @samp{-W} in @code{tar}.
2254 Print the version number.
2256 @item version-control
2257 @samp{-V} in @code{cp}, @code{ln}, @code{mv}.
2260 @samp{-v} in @code{ctags}.
2263 @samp{-V} in @code{tar}.
2266 @samp{-W} in @code{make}.
2268 @item whole-size-limit
2269 @samp{-l} in @code{shar}.
2272 @samp{-w} in @code{ls} and @code{ptx}.
2275 @samp{-W} in @code{ptx}.
2278 @samp{-T} in @code{who}.
2281 @samp{-z} in @code{gprof}.
2284 @node OID Allocations
2285 @section OID Allocations
2286 @cindex OID allocations for GNU
2291 The OID (object identifier) 1.3.6.1.4.1.11591 has been assigned to the
2292 GNU Project (thanks to Werner Koch). These are used for SNMP, LDAP,
2293 X.509 certificates, and so on. The web site
2294 @url{http://www.alvestrand.no/objectid} has a (voluntary) listing of
2295 many OID assignments.
2297 If you need a new slot for your GNU package, write
2298 @email{maintainers@@gnu.org}. Here is a list of arcs currently
2302 @include gnu-oids.texi
2307 @section Memory Usage
2308 @cindex memory usage
2310 If a program typically uses just a few meg of memory, don't bother making any
2311 effort to reduce memory usage. For example, if it is impractical for
2312 other reasons to operate on files more than a few meg long, it is
2313 reasonable to read entire input files into memory to operate on them.
2315 However, for programs such as @code{cat} or @code{tail}, that can
2316 usefully operate on very large files, it is important to avoid using a
2317 technique that would artificially limit the size of files it can handle.
2318 If a program works by lines and could be applied to arbitrary
2319 user-supplied input files, it should keep only a line in memory, because
2320 this is not very hard and users will want to be able to operate on input
2321 files that are bigger than will fit in memory all at once.
2323 If your program creates complicated data structures, just make them in
2324 memory and give a fatal error if @code{malloc} returns zero.
2328 Memory analysis tools such as @command{valgrind} can be useful, but
2329 don't complicate a program merely to avoid their false alarms. For
2330 example, if memory is used until just before a process exits, don't
2331 free it simply to silence such a tool.
2338 Programs should be prepared to operate when @file{/usr} and @file{/etc}
2339 are read-only file systems. Thus, if the program manages log files,
2340 lock files, backup files, score files, or any other files which are
2341 modified for internal purposes, these files should not be stored in
2342 @file{/usr} or @file{/etc}.
2344 There are two exceptions. @file{/etc} is used to store system
2345 configuration information; it is reasonable for a program to modify
2346 files in @file{/etc} when its job is to update the system configuration.
2347 Also, if the user explicitly asks to modify one file in a directory, it
2348 is reasonable for the program to store other files in the same
2352 @chapter Making The Best Use of C
2354 This chapter provides advice on how best to use the C language
2355 when writing GNU software.
2358 * Formatting:: Formatting your source code.
2359 * Comments:: Commenting your work.
2360 * Syntactic Conventions:: Clean use of C constructs.
2361 * Names:: Naming variables, functions, and files.
2362 * System Portability:: Portability among different operating systems.
2363 * CPU Portability:: Supporting the range of CPU types.
2364 * System Functions:: Portability and ``standard'' library functions.
2365 * Internationalization:: Techniques for internationalization.
2366 * Character Set:: Use ASCII by default.
2367 * Quote Characters:: Use "..." or '...' in the C locale.
2368 * Mmap:: How you can safely use @code{mmap}.
2372 @section Formatting Your Source Code
2373 @cindex formatting source code
2376 @cindex braces, in C source
2377 @cindex function definitions, formatting
2378 It is important to put the open-brace that starts the body of a C
2379 function in column one, so that they will start a defun. Several
2380 tools look for open-braces in column one to find the beginnings of C
2381 functions. These tools will not work on code not formatted that way.
2383 Avoid putting open-brace, open-parenthesis or open-bracket in column
2384 one when they are inside a function, so that they won't start a defun.
2385 The open-brace that starts a @code{struct} body can go in column one
2386 if you find it useful to treat that definition as a defun.
2388 It is also important for function definitions to start the name of the
2389 function in column one. This helps people to search for function
2390 definitions, and may also help certain tools recognize them. Thus,
2391 using Standard C syntax, the format is this:
2395 concat (char *s1, char *s2)
2402 or, if you want to use traditional C syntax, format the definition like
2407 concat (s1, s2) /* Name starts in column one here */
2409 @{ /* Open brace in column one here */
2414 In Standard C, if the arguments don't fit nicely on one line,
2419 lots_of_args (int an_integer, long a_long, short a_short,
2420 double a_double, float a_float)
2424 @cindex @code{struct} types, formatting
2425 @cindex @code{enum} types, formatting
2426 For @code{struct} and @code{enum} types, likewise put the braces in
2427 column one, unless the whole contents fits on one line:
2435 struct foo @{ int a, b; @}
2438 The rest of this section gives our recommendations for other aspects of
2439 C formatting style, which is also the default style of the @code{indent}
2440 program in version 1.2 and newer. It corresponds to the options
2443 -nbad -bap -nbc -bbo -bl -bli2 -bls -ncdb -nce -cp1 -cs -di2
2444 -ndj -nfc1 -nfca -hnl -i2 -ip5 -lp -pcs -psl -nsc -nsob
2447 We don't think of these recommendations as requirements, because it
2448 causes no problems for users if two different programs have different
2451 But whatever style you use, please use it consistently, since a mixture
2452 of styles within one program tends to look ugly. If you are
2453 contributing changes to an existing program, please follow the style of
2456 For the body of the function, our recommended style looks like this:
2468 return ++x + bar ();
2472 @cindex spaces before open-paren
2473 We find it easier to read a program when it has spaces before the
2474 open-parentheses and after the commas. Especially after the commas.
2476 When you split an expression into multiple lines, split it
2477 before an operator, not after one. Here is the right way:
2479 @cindex expressions, splitting
2481 if (foo_this_is_long && bar > win (x, y, z)
2482 && remaining_condition)
2485 Try to avoid having two operators of different precedence at the same
2486 level of indentation. For example, don't write this:
2489 mode = (inmode[j] == VOIDmode
2490 || GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])
2491 ? outmode[j] : inmode[j]);
2494 Instead, use extra parentheses so that the indentation shows the nesting:
2497 mode = ((inmode[j] == VOIDmode
2498 || (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])))
2499 ? outmode[j] : inmode[j]);
2502 Insert extra parentheses so that Emacs will indent the code properly.
2503 For example, the following indentation looks nice if you do it by hand,
2506 v = rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000
2507 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000;
2511 but Emacs would alter it. Adding a set of parentheses produces
2512 something that looks equally nice, and which Emacs will preserve:
2515 v = (rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000
2516 + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000);
2519 Format do-while statements like this:
2531 Please use formfeed characters (control-L) to divide the program into
2532 pages at logical places (but not within a function). It does not matter
2533 just how long the pages are, since they do not have to fit on a printed
2534 page. The formfeeds should appear alone on lines by themselves.
2537 @section Commenting Your Work
2540 Every program should start with a comment saying briefly what it is for.
2541 Example: @samp{fmt - filter for simple filling of text}. This comment
2542 should be at the top of the source file containing the @samp{main}
2543 function of the program.
2545 Also, please write a brief comment at the start of each source file,
2546 with the file name and a line or two about the overall purpose of the
2549 Please write the comments in a GNU program in English, because English
2550 is the one language that nearly all programmers in all countries can
2551 read. If you do not write English well, please write comments in
2552 English as well as you can, then ask other people to help rewrite them.
2553 If you can't write comments in English, please find someone to work with
2554 you and translate your comments into English.
2556 Please put a comment on each function saying what the function does,
2557 what sorts of arguments it gets, and what the possible values of
2558 arguments mean and are used for. It is not necessary to duplicate in
2559 words the meaning of the C argument declarations, if a C type is being
2560 used in its customary fashion. If there is anything nonstandard about
2561 its use (such as an argument of type @code{char *} which is really the
2562 address of the second character of a string, not the first), or any
2563 possible values that would not work the way one would expect (such as,
2564 that strings containing newlines are not guaranteed to work), be sure
2567 Also explain the significance of the return value, if there is one.
2569 Please put two spaces after the end of a sentence in your comments, so
2570 that the Emacs sentence commands will work. Also, please write
2571 complete sentences and capitalize the first word. If a lower-case
2572 identifier comes at the beginning of a sentence, don't capitalize it!
2573 Changing the spelling makes it a different identifier. If you don't
2574 like starting a sentence with a lower case letter, write the sentence
2575 differently (e.g., ``The identifier lower-case is @dots{}'').
2577 The comment on a function is much clearer if you use the argument
2578 names to speak about the argument values. The variable name itself
2579 should be lower case, but write it in upper case when you are speaking
2580 about the value rather than the variable itself. Thus, ``the inode
2581 number NODE_NUM'' rather than ``an inode''.
2583 There is usually no purpose in restating the name of the function in
2584 the comment before it, because readers can see that for themselves.
2585 There might be an exception when the comment is so long that the function
2586 itself would be off the bottom of the screen.
2588 There should be a comment on each static variable as well, like this:
2591 /* Nonzero means truncate lines in the display;
2592 zero means continue them. */
2596 @cindex conditionals, comments for
2597 @cindex @code{#endif}, commenting
2598 Every @samp{#endif} should have a comment, except in the case of short
2599 conditionals (just a few lines) that are not nested. The comment should
2600 state the condition of the conditional that is ending, @emph{including
2601 its sense}. @samp{#else} should have a comment describing the condition
2602 @emph{and sense} of the code that follows. For example:
2610 #endif /* not foo */
2620 but, by contrast, write the comments this way for a @samp{#ifndef}:
2633 #endif /* not foo */
2637 @node Syntactic Conventions
2638 @section Clean Use of C Constructs
2639 @cindex syntactic conventions
2641 @cindex implicit @code{int}
2642 @cindex function argument, declaring
2643 Please explicitly declare the types of all objects. For example, you
2644 should explicitly declare all arguments to functions, and you should
2645 declare functions to return @code{int} rather than omitting the
2648 @cindex compiler warnings
2649 @cindex @samp{-Wall} compiler option
2650 Some programmers like to use the GCC @samp{-Wall} option, and change the
2651 code whenever it issues a warning. If you want to do this, then do.
2652 Other programmers prefer not to use @samp{-Wall}, because it gives
2653 warnings for valid and legitimate code which they do not want to change.
2654 If you want to do this, then do. The compiler should be your servant,
2659 Don't make the program ugly just to placate static analysis tools such
2660 as @command{lint}, @command{clang}, and GCC with extra warnings
2661 options such as @option{-Wconversion} and @option{-Wundef}. These
2662 tools can help find bugs and unclear code, but they can also generate
2663 so many false alarms that it hurts readability to silence them with
2664 unnecessary casts, wrappers, and other complications. For example,
2665 please don't insert casts to @code{void} or calls to do-nothing
2666 functions merely to pacify a lint checker.
2668 Declarations of external functions and functions to appear later in the
2669 source file should all go in one place near the beginning of the file
2670 (somewhere before the first function definition in the file), or else
2671 should go in a header file. Don't put @code{extern} declarations inside
2674 @cindex temporary variables
2675 It used to be common practice to use the same local variables (with
2676 names like @code{tem}) over and over for different values within one
2677 function. Instead of doing this, it is better to declare a separate local
2678 variable for each distinct purpose, and give it a name which is
2679 meaningful. This not only makes programs easier to understand, it also
2680 facilitates optimization by good compilers. You can also move the
2681 declaration of each local variable into the smallest scope that includes
2682 all its uses. This makes the program even cleaner.
2684 Don't use local variables or parameters that shadow global identifiers.
2685 GCC's @samp{-Wshadow} option can detect this problem.
2687 @cindex multiple variables in a line
2688 Don't declare multiple variables in one declaration that spans lines.
2689 Start a new declaration on each line, instead. For example, instead
2715 (If they are global variables, each should have a comment preceding it
2718 When you have an @code{if}-@code{else} statement nested in another
2719 @code{if} statement, always put braces around the @code{if}-@code{else}.
2720 Thus, never write like this:
2743 If you have an @code{if} statement nested inside of an @code{else}
2744 statement, either write @code{else if} on one line, like this,
2754 with its @code{then}-part indented like the preceding @code{then}-part,
2755 or write the nested @code{if} within braces like this:
2767 Don't declare both a structure tag and variables or typedefs in the
2768 same declaration. Instead, declare the structure tag separately
2769 and then use it to declare the variables or typedefs.
2771 Try to avoid assignments inside @code{if}-conditions (assignments
2772 inside @code{while}-conditions are ok). For example, don't write
2776 if ((foo = (char *) malloc (sizeof *foo)) == 0)
2777 fatal ("virtual memory exhausted");
2781 instead, write this:
2784 foo = (char *) malloc (sizeof *foo);
2786 fatal ("virtual memory exhausted");
2789 This example uses zero without a cast as a null pointer constant.
2790 This is perfectly fine, except that a cast is needed when calling a
2791 varargs function or when using @code{sizeof}.
2794 @section Naming Variables, Functions, and Files
2796 @cindex names of variables, functions, and files
2797 The names of global variables and functions in a program serve as
2798 comments of a sort. So don't choose terse names---instead, look for
2799 names that give useful information about the meaning of the variable or
2800 function. In a GNU program, names should be English, like other
2803 Local variable names can be shorter, because they are used only within
2804 one context, where (presumably) comments explain their purpose.
2806 Try to limit your use of abbreviations in symbol names. It is ok to
2807 make a few abbreviations, explain what they mean, and then use them
2808 frequently, but don't use lots of obscure abbreviations.
2810 Please use underscores to separate words in a name, so that the Emacs
2811 word commands can be useful within them. Stick to lower case; reserve
2812 upper case for macros and @code{enum} constants, and for name-prefixes
2813 that follow a uniform convention.
2815 For example, you should use names like @code{ignore_space_change_flag};
2816 don't use names like @code{iCantReadThis}.
2818 Variables that indicate whether command-line options have been
2819 specified should be named after the meaning of the option, not after
2820 the option-letter. A comment should state both the exact meaning of
2821 the option and its letter. For example,
2825 /* Ignore changes in horizontal whitespace (-b). */
2826 int ignore_space_change_flag;
2830 When you want to define names with constant integer values, use
2831 @code{enum} rather than @samp{#define}. GDB knows about enumeration
2834 @cindex file-name limitations
2836 You might want to make sure that none of the file names would conflict
2837 if the files were loaded onto an MS-DOS file system which shortens the
2838 names. You can use the program @code{doschk} to test for this.
2840 Some GNU programs were designed to limit themselves to file names of 14
2841 characters or less, to avoid file name conflicts if they are read into
2842 older System V systems. Please preserve this feature in the existing
2843 GNU programs that have it, but there is no need to do this in new GNU
2844 programs. @code{doschk} also reports file names longer than 14
2848 @node System Portability
2849 @section Portability between System Types
2850 @cindex portability, between system types
2852 In the Unix world, ``portability'' refers to porting to different Unix
2853 versions. For a GNU program, this kind of portability is desirable, but
2856 The primary purpose of GNU software is to run on top of the GNU kernel,
2857 compiled with the GNU C compiler, on various types of CPU. So the
2858 kinds of portability that are absolutely necessary are quite limited.
2859 But it is important to support Linux-based GNU systems, since they
2860 are the form of GNU that is popular.
2862 Beyond that, it is good to support the other free operating systems
2863 (*BSD), and it is nice to support other Unix-like systems if you want
2864 to. Supporting a variety of Unix-like systems is desirable, although
2865 not paramount. It is usually not too hard, so you may as well do it.
2866 But you don't have to consider it an obligation, if it does turn out to
2870 The easiest way to achieve portability to most Unix-like systems is to
2871 use Autoconf. It's unlikely that your program needs to know more
2872 information about the host platform than Autoconf can provide, simply
2873 because most of the programs that need such knowledge have already been
2876 Avoid using the format of semi-internal data bases (e.g., directories)
2877 when there is a higher-level alternative (@code{readdir}).
2879 @cindex non-POSIX systems, and portability
2880 As for systems that are not like Unix, such as MSDOS, Windows, VMS, MVS,
2881 and older Macintosh systems, supporting them is often a lot of work.
2882 When that is the case, it is better to spend your time adding features
2883 that will be useful on GNU and GNU/Linux, rather than on supporting
2884 other incompatible systems.
2886 If you do support Windows, please do not abbreviate it as ``win''. In
2887 hacker terminology, calling something a ``win'' is a form of praise.
2888 You're free to praise Microsoft Windows on your own if you want, but
2889 please don't do this in GNU packages. Instead of abbreviating
2890 ``Windows'' to ``win'', you can write it in full or abbreviate it to
2891 ``woe'' or ``w''. In GNU Emacs, for instance, we use @samp{w32} in
2892 file names of Windows-specific files, but the macro for Windows
2893 conditionals is called @code{WINDOWSNT}.
2895 It is a good idea to define the ``feature test macro''
2896 @code{_GNU_SOURCE} when compiling your C files. When you compile on GNU
2897 or GNU/Linux, this will enable the declarations of GNU library extension
2898 functions, and that will usually give you a compiler error message if
2899 you define the same function names in some other way in your program.
2900 (You don't have to actually @emph{use} these functions, if you prefer
2901 to make the program more portable to other systems.)
2903 But whether or not you use these GNU extensions, you should avoid
2904 using their names for any other meanings. Doing so would make it hard
2905 to move your code into other GNU programs.
2907 @node CPU Portability
2908 @section Portability between CPUs
2910 @cindex data types, and portability
2911 @cindex portability, and data types
2912 Even GNU systems will differ because of differences among CPU
2913 types---for example, difference in byte ordering and alignment
2914 requirements. It is absolutely essential to handle these differences.
2915 However, don't make any effort to cater to the possibility that an
2916 @code{int} will be less than 32 bits. We don't support 16-bit machines
2919 Similarly, don't make any effort to cater to the possibility that
2920 @code{long} will be smaller than predefined types like @code{size_t}.
2921 For example, the following code is ok:
2924 printf ("size = %lu\n", (unsigned long) sizeof array);
2925 printf ("diff = %ld\n", (long) (pointer2 - pointer1));
2928 1989 Standard C requires this to work, and we know of only one
2929 counterexample: 64-bit programs on Microsoft Windows. We will leave
2930 it to those who want to port GNU programs to that environment to
2931 figure out how to do it.
2933 Predefined file-size types like @code{off_t} are an exception: they are
2934 longer than @code{long} on many platforms, so code like the above won't
2935 work with them. One way to print an @code{off_t} value portably is to
2936 print its digits yourself, one by one.
2938 Don't assume that the address of an @code{int} object is also the
2939 address of its least-significant byte. This is false on big-endian
2940 machines. Thus, don't make the following mistake:
2945 while ((c = getchar ()) != EOF)
2946 write (file_descriptor, &c, 1);
2949 @noindent Instead, use @code{unsigned char} as follows. (The @code{unsigned}
2950 is for portability to unusual systems where @code{char} is signed and
2951 where there is integer overflow checking.)
2955 while ((c = getchar ()) != EOF)
2957 unsigned char u = c;
2958 write (file_descriptor, &u, 1);
2962 @cindex casting pointers to integers
2963 Avoid casting pointers to integers if you can. Such casts greatly
2964 reduce portability, and in most programs they are easy to avoid. In the
2965 cases where casting pointers to integers is essential---such as, a Lisp
2966 interpreter which stores type information as well as an address in one
2967 word---you'll have to make explicit provisions to handle different word
2968 sizes. You will also need to make provision for systems in which the
2969 normal range of addresses you can get from @code{malloc} starts far away
2973 @node System Functions
2974 @section Calling System Functions
2976 @cindex C library functions, and portability
2977 @cindex POSIX functions, and portability
2978 @cindex library functions, and portability
2979 @cindex portability, and library functions
2981 Historically, C implementations differed substantially, and many
2982 systems lacked a full implementation of ANSI/ISO C89. Nowadays,
2983 however, all practical systems have a C89 compiler and GNU C supports
2984 almost all of C99 and some of C11. Similarly, most systems implement
2985 POSIX.1-2001 libraries and tools, and many have POSIX.1-2008.
2987 Hence, there is little reason to support old C or non-POSIX systems,
2988 and you may want to take advantage of standard C and POSIX to write
2989 clearer, more portable, or faster code. You should use standard
2990 interfaces where possible; but if GNU extensions make your program
2991 more maintainable, powerful, or otherwise better, don't hesitate to
2992 use them. In any case, don't make your own declaration of system
2993 functions; that's a recipe for conflict.
2995 Despite the standards, nearly every library function has some sort of
2996 portability issue on some system or another. Here are some examples:
3000 Names with trailing @code{/}'s are mishandled on many platforms.
3003 @code{long double} may be unimplemented; floating values Infinity and
3004 NaN are often mishandled; output for large precisions may be
3008 May return @code{int} instead of @code{ssize_t}.
3011 On Windows, @code{errno} is not set on failure.
3015 @uref{http://www.gnu.org/software/gnulib/, Gnulib} is a big help in
3016 this regard. Gnulib provides implementations of standard interfaces
3017 on many of the systems that lack them, including portable
3018 implementations of enhanced GNU interfaces, thereby making their use
3019 portable, and of POSIX-1.2008 interfaces, some of which are missing
3020 even on up-to-date GNU systems.
3022 @findex xmalloc, in Gnulib
3023 @findex error messages, in Gnulib
3024 @findex data structures, in Gnulib
3025 Gnulib also provides many useful non-standard interfaces; for example,
3026 C implementations of standard data structures (hash tables, binary
3027 trees), error-checking type-safe wrappers for memory allocation
3028 functions (@code{xmalloc}, @code{xrealloc}), and output of error
3031 Gnulib integrates with GNU Autoconf and Automake to remove much of the
3032 burden of writing portable code from the programmer: Gnulib makes your
3033 configure script automatically determine what features are missing and
3034 use the Gnulib code to supply the missing pieces.
3036 The Gnulib and Autoconf manuals have extensive sections on
3037 portability: @ref{Top,, Introduction, gnulib, Gnulib} and
3038 @pxref{Portable C and C++,,, autoconf, Autoconf}. Please consult them
3039 for many more details.
3042 @node Internationalization
3043 @section Internationalization
3044 @cindex internationalization
3047 GNU has a library called GNU gettext that makes it easy to translate the
3048 messages in a program into various languages. You should use this
3049 library in every program. Use English for the messages as they appear
3050 in the program, and let gettext provide the way to translate them into
3053 Using GNU gettext involves putting a call to the @code{gettext} macro
3054 around each string that might need translation---like this:
3057 printf (gettext ("Processing file '%s'..."), file);
3061 This permits GNU gettext to replace the string @code{"Processing file
3062 '%s'..."} with a translated version.
3064 Once a program uses gettext, please make a point of writing calls to
3065 @code{gettext} when you add new strings that call for translation.
3067 Using GNU gettext in a package involves specifying a @dfn{text domain
3068 name} for the package. The text domain name is used to separate the
3069 translations for this package from the translations for other packages.
3070 Normally, the text domain name should be the same as the name of the
3071 package---for example, @samp{coreutils} for the GNU core utilities.
3073 @cindex message text, and internationalization
3074 To enable gettext to work well, avoid writing code that makes
3075 assumptions about the structure of words or sentences. When you want
3076 the precise text of a sentence to vary depending on the data, use two or
3077 more alternative string constants each containing a complete sentences,
3078 rather than inserting conditionalized words or phrases into a single
3081 Here is an example of what not to do:
3084 printf ("%s is full", capacity > 5000000 ? "disk" : "floppy disk");
3087 If you apply gettext to all strings, like this,
3090 printf (gettext ("%s is full"),
3091 capacity > 5000000 ? gettext ("disk") : gettext ("floppy disk"));
3095 the translator will hardly know that "disk" and "floppy disk" are meant to
3096 be substituted in the other string. Worse, in some languages (like French)
3097 the construction will not work: the translation of the word "full" depends
3098 on the gender of the first part of the sentence; it happens to be not the
3099 same for "disk" as for "floppy disk".
3101 Complete sentences can be translated without problems:
3104 printf (capacity > 5000000 ? gettext ("disk is full")
3105 : gettext ("floppy disk is full"));
3108 A similar problem appears at the level of sentence structure with this
3112 printf ("# Implicit rule search has%s been done.\n",
3113 f->tried_implicit ? "" : " not");
3117 Adding @code{gettext} calls to this code cannot give correct results for
3118 all languages, because negation in some languages requires adding words
3119 at more than one place in the sentence. By contrast, adding
3120 @code{gettext} calls does the job straightforwardly if the code starts
3124 printf (f->tried_implicit
3125 ? "# Implicit rule search has been done.\n",
3126 : "# Implicit rule search has not been done.\n");
3129 Another example is this one:
3132 printf ("%d file%s processed", nfiles,
3133 nfiles != 1 ? "s" : "");
3137 The problem with this example is that it assumes that plurals are made
3138 by adding `s'. If you apply gettext to the format string, like this,
3141 printf (gettext ("%d file%s processed"), nfiles,
3142 nfiles != 1 ? "s" : "");
3146 the message can use different words, but it will still be forced to use
3147 `s' for the plural. Here is a better way, with gettext being applied to
3148 the two strings independently:
3151 printf ((nfiles != 1 ? gettext ("%d files processed")
3152 : gettext ("%d file processed")),
3157 But this still doesn't work for languages like Polish, which has three
3158 plural forms: one for nfiles == 1, one for nfiles == 2, 3, 4, 22, 23, 24, ...
3159 and one for the rest. The GNU @code{ngettext} function solves this problem:
3162 printf (ngettext ("%d files processed", "%d file processed", nfiles),
3168 @section Character Set
3169 @cindex character set
3171 @cindex ASCII characters
3172 @cindex non-ASCII characters
3174 Sticking to the ASCII character set (plain text, 7-bit characters) is
3175 preferred in GNU source code comments, text documents, and other
3176 contexts, unless there is good reason to do something else because of
3177 the application domain. For example, if source code deals with the
3178 French Revolutionary calendar, it is OK if its literal strings contain
3179 accented characters in month names like ``Flor@'eal''. Also, it is OK
3180 (but not required) to use non-ASCII characters to represent proper
3181 names of contributors in change logs (@pxref{Change Logs}).
3183 If you need to use non-ASCII characters, you should normally stick
3184 with one encoding, certainly within a single file. UTF-8 is likely to
3188 @node Quote Characters
3189 @section Quote Characters
3190 @cindex quote characters
3191 @cindex locale-specific quote characters
3194 @cindex opening quote
3195 @cindex single quote
3196 @cindex double quote
3197 @cindex grave accent
3198 @set txicodequoteundirected
3199 @set txicodequotebacktick
3201 In the C locale, the output of GNU programs should stick to plain
3202 ASCII for quotation characters in messages to users: preferably 0x22
3203 (@samp{"}) or 0x27 (@samp{'}) for both opening and closing quotes.
3204 Although GNU programs traditionally used 0x60 (@samp{`}) for opening
3205 and 0x27 (@samp{'}) for closing quotes, nowadays quotes @samp{`like
3206 this'} are typically rendered asymmetrically, so quoting @samp{"like
3207 this"} or @samp{'like this'} typically looks better.
3209 It is ok, but not required, for GNU programs to generate
3210 locale-specific quotes in non-C locales. For example:
3213 printf (gettext ("Processing file '%s'..."), file);
3217 Here, a French translation might cause @code{gettext} to return the
3218 string @code{"Traitement de fichier
3219 @guilsinglleft{}@tie{}%s@tie{}@guilsinglright{}..."}, yielding quotes
3220 more appropriate for a French locale.
3222 Sometimes a program may need to use opening and closing quotes
3223 directly. By convention, @code{gettext} translates the string
3224 @samp{"`"} to the opening quote and the string @samp{"'"} to the
3225 closing quote, and a program can use these translations. Generally,
3226 though, it is better to translate quote characters in the context of
3229 If the output of your program is ever likely to be parsed by another
3230 program, it is good to provide an option that makes this parsing
3231 reliable. For example, you could escape special characters using
3232 conventions from the C language or the Bourne shell. See for example
3233 the option @option{--quoting-style} of GNU @code{ls}.
3235 @clear txicodequoteundirected
3236 @clear txicodequotebacktick
3243 If you use @code{mmap} to read or write files, don't assume it either
3244 works on all files or fails for all files. It may work on some files
3247 The proper way to use @code{mmap} is to try it on the specific file for
3248 which you want to use it---and if @code{mmap} doesn't work, fall back on
3249 doing the job in another way using @code{read} and @code{write}.
3251 The reason this precaution is needed is that the GNU kernel (the HURD)
3252 provides a user-extensible file system, in which there can be many
3253 different kinds of ``ordinary files''. Many of them support
3254 @code{mmap}, but some do not. It is important to make programs handle
3255 all these kinds of files.
3259 @chapter Documenting Programs
3260 @cindex documentation
3262 A GNU program should ideally come with full free documentation, adequate
3263 for both reference and tutorial purposes. If the package can be
3264 programmed or extended, the documentation should cover programming or
3265 extending it, as well as just using it.
3268 * GNU Manuals:: Writing proper manuals.
3269 * Doc Strings and Manuals:: Compiling doc strings doesn't make a manual.
3270 * Manual Structure Details:: Specific structure conventions.
3271 * License for Manuals:: Writing the distribution terms for a manual.
3272 * Manual Credits:: Giving credit to documentation contributors.
3273 * Printed Manuals:: Mentioning the printed manual.
3274 * NEWS File:: NEWS files supplement manuals.
3275 * Change Logs:: Recording changes.
3276 * Man Pages:: Man pages are secondary.
3277 * Reading other Manuals:: How far you can go in learning
3282 @section GNU Manuals
3284 The preferred document format for the GNU system is the Texinfo
3285 formatting language. Every GNU package should (ideally) have
3286 documentation in Texinfo both for reference and for learners. Texinfo
3287 makes it possible to produce a good quality formatted book, using
3288 @TeX{}, and to generate an Info file. It is also possible to generate
3289 HTML output from Texinfo source. See the Texinfo manual, either the
3290 hardcopy, or the on-line version available through @code{info} or the
3291 Emacs Info subsystem (@kbd{C-h i}).
3293 Nowadays some other formats such as Docbook and Sgmltexi can be
3294 converted automatically into Texinfo. It is ok to produce the Texinfo
3295 documentation by conversion this way, as long as it gives good results.
3297 Make sure your manual is clear to a reader who knows nothing about the
3298 topic and reads it straight through. This means covering basic topics
3299 at the beginning, and advanced topics only later. This also means
3300 defining every specialized term when it is first used.
3302 Programmers tend to carry over the structure of the program as the
3303 structure for its documentation. But this structure is not
3304 necessarily good for explaining how to use the program; it may be
3305 irrelevant and confusing for a user.
3307 Instead, the right way to structure documentation is according to the
3308 concepts and questions that a user will have in mind when reading it.
3309 This principle applies at every level, from the lowest (ordering
3310 sentences in a paragraph) to the highest (ordering of chapter topics
3311 within the manual). Sometimes this structure of ideas matches the
3312 structure of the implementation of the software being documented---but
3313 often they are different. An important part of learning to write good
3314 documentation is to learn to notice when you have unthinkingly
3315 structured the documentation like the implementation, stop yourself,
3316 and look for better alternatives.
3318 For example, each program in the GNU system probably ought to be
3319 documented in one manual; but this does not mean each program should
3320 have its own manual. That would be following the structure of the
3321 implementation, rather than the structure that helps the user
3324 Instead, each manual should cover a coherent @emph{topic}. For example,
3325 instead of a manual for @code{diff} and a manual for @code{diff3}, we
3326 have one manual for ``comparison of files'' which covers both of those
3327 programs, as well as @code{cmp}. By documenting these programs
3328 together, we can make the whole subject clearer.
3330 The manual which discusses a program should certainly document all of
3331 the program's command-line options and all of its commands. It should
3332 give examples of their use. But don't organize the manual as a list
3333 of features. Instead, organize it logically, by subtopics. Address
3334 the questions that a user will ask when thinking about the job that
3335 the program does. Don't just tell the reader what each feature can
3336 do---say what jobs it is good for, and show how to use it for those
3337 jobs. Explain what is recommended usage, and what kinds of usage
3340 In general, a GNU manual should serve both as tutorial and reference.
3341 It should be set up for convenient access to each topic through Info,
3342 and for reading straight through (appendixes aside). A GNU manual
3343 should give a good introduction to a beginner reading through from the
3344 start, and should also provide all the details that hackers want.
3345 The Bison manual is a good example of this---please take a look at it
3346 to see what we mean.
3348 That is not as hard as it first sounds. Arrange each chapter as a
3349 logical breakdown of its topic, but order the sections, and write their
3350 text, so that reading the chapter straight through makes sense. Do
3351 likewise when structuring the book into chapters, and when structuring a
3352 section into paragraphs. The watchword is, @emph{at each point, address
3353 the most fundamental and important issue raised by the preceding text.}
3355 If necessary, add extra chapters at the beginning of the manual which
3356 are purely tutorial and cover the basics of the subject. These provide
3357 the framework for a beginner to understand the rest of the manual. The
3358 Bison manual provides a good example of how to do this.
3360 To serve as a reference, a manual should have an Index that lists all
3361 the functions, variables, options, and important concepts that are
3362 part of the program. One combined Index should do for a short manual,
3363 but sometimes for a complex package it is better to use multiple
3364 indices. The Texinfo manual includes advice on preparing good index
3365 entries, see @ref{Index Entries, , Making Index Entries, texinfo, GNU
3366 Texinfo}, and see @ref{Indexing Commands, , Defining the Entries of an
3367 Index, texinfo, GNU Texinfo}.
3369 Don't use Unix man pages as a model for how to write GNU documentation;
3370 most of them are terse, badly structured, and give inadequate
3371 explanation of the underlying concepts. (There are, of course, some
3372 exceptions.) Also, Unix man pages use a particular format which is
3373 different from what we use in GNU manuals.
3375 Please include an email address in the manual for where to report
3376 bugs @emph{in the text of the manual}.
3378 Please do not use the term ``pathname'' that is used in Unix
3379 documentation; use ``file name'' (two words) instead. We use the term
3380 ``path'' only for search paths, which are lists of directory names.
3382 Please do not use the term ``illegal'' to refer to erroneous input to
3383 a computer program. Please use ``invalid'' for this, and reserve the
3384 term ``illegal'' for activities prohibited by law.
3386 Please do not write @samp{()} after a function name just to indicate
3387 it is a function. @code{foo ()} is not a function, it is a function
3388 call with no arguments.
3390 @node Doc Strings and Manuals
3391 @section Doc Strings and Manuals
3393 Some programming systems, such as Emacs, provide a documentation string
3394 for each function, command or variable. You may be tempted to write a
3395 reference manual by compiling the documentation strings and writing a
3396 little additional text to go around them---but you must not do it. That
3397 approach is a fundamental mistake. The text of well-written
3398 documentation strings will be entirely wrong for a manual.
3400 A documentation string needs to stand alone---when it appears on the
3401 screen, there will be no other text to introduce or explain it.
3402 Meanwhile, it can be rather informal in style.
3404 The text describing a function or variable in a manual must not stand
3405 alone; it appears in the context of a section or subsection. Other text
3406 at the beginning of the section should explain some of the concepts, and
3407 should often make some general points that apply to several functions or
3408 variables. The previous descriptions of functions and variables in the
3409 section will also have given information about the topic. A description
3410 written to stand alone would repeat some of that information; this
3411 redundancy looks bad. Meanwhile, the informality that is acceptable in
3412 a documentation string is totally unacceptable in a manual.
3414 The only good way to use documentation strings in writing a good manual
3415 is to use them as a source of information for writing good text.
3417 @node Manual Structure Details
3418 @section Manual Structure Details
3419 @cindex manual structure
3421 The title page of the manual should state the version of the programs or
3422 packages documented in the manual. The Top node of the manual should
3423 also contain this information. If the manual is changing more
3424 frequently than or independent of the program, also state a version
3425 number for the manual in both of these places.
3427 Each program documented in the manual should have a node named
3428 @samp{@var{program} Invocation} or @samp{Invoking @var{program}}. This
3429 node (together with its subnodes, if any) should describe the program's
3430 command line arguments and how to run it (the sort of information people
3431 would look for in a man page). Start with an @samp{@@example}
3432 containing a template for all the options and arguments that the program
3435 Alternatively, put a menu item in some menu whose item name fits one of
3436 the above patterns. This identifies the node which that item points to
3437 as the node for this purpose, regardless of the node's actual name.
3439 The @samp{--usage} feature of the Info reader looks for such a node
3440 or menu item in order to find the relevant text, so it is essential
3441 for every Texinfo file to have one.
3443 If one manual describes several programs, it should have such a node for
3444 each program described in the manual.
3446 @node License for Manuals
3447 @section License for Manuals
3448 @cindex license for manuals
3450 Please use the GNU Free Documentation License for all GNU manuals that
3451 are more than a few pages long. Likewise for a collection of short
3452 documents---you only need one copy of the GNU FDL for the whole
3453 collection. For a single short document, you can use a very permissive
3454 non-copyleft license, to avoid taking up space with a long license.
3456 See @uref{http://www.gnu.org/copyleft/fdl-howto.html} for more explanation
3457 of how to employ the GFDL.
3459 Note that it is not obligatory to include a copy of the GNU GPL or GNU
3460 LGPL in a manual whose license is neither the GPL nor the LGPL. It can
3461 be a good idea to include the program's license in a large manual; in a
3462 short manual, whose size would be increased considerably by including
3463 the program's license, it is probably better not to include it.
3465 @node Manual Credits
3466 @section Manual Credits
3467 @cindex credits for manuals
3469 Please credit the principal human writers of the manual as the authors,
3470 on the title page of the manual. If a company sponsored the work, thank
3471 the company in a suitable place in the manual, but do not cite the
3472 company as an author.
3474 @node Printed Manuals
3475 @section Printed Manuals
3477 The FSF publishes some GNU manuals in printed form. To encourage sales
3478 of these manuals, the on-line versions of the manual should mention at
3479 the very start that the printed manual is available and should point at
3480 information for getting it---for instance, with a link to the page
3481 @url{http://www.gnu.org/order/order.html}. This should not be included
3482 in the printed manual, though, because there it is redundant.
3484 It is also useful to explain in the on-line forms of the manual how the
3485 user can print out the manual from the sources.
3488 @section The NEWS File
3489 @cindex @file{NEWS} file
3491 In addition to its manual, the package should have a file named
3492 @file{NEWS} which contains a list of user-visible changes worth
3493 mentioning. In each new release, add items to the front of the file and
3494 identify the version they pertain to. Don't discard old items; leave
3495 them in the file after the newer items. This way, a user upgrading from
3496 any previous version can see what is new.
3498 If the @file{NEWS} file gets very long, move some of the older items
3499 into a file named @file{ONEWS} and put a note at the end referring the
3503 @section Change Logs
3506 Keep a change log to describe all the changes made to program source
3507 files. The purpose of this is so that people investigating bugs in the
3508 future will know about the changes that might have introduced the bug.
3509 Often a new bug can be found by looking at what was recently changed.
3510 More importantly, change logs can help you eliminate conceptual
3511 inconsistencies between different parts of a program, by giving you a
3512 history of how the conflicting concepts arose and who they came from.
3515 * Change Log Concepts::
3516 * Style of Change Logs::
3518 * Conditional Changes::
3519 * Indicating the Part Changed::
3522 @node Change Log Concepts
3523 @subsection Change Log Concepts
3526 @cindex batch of changes
3527 You can think of the change log as a conceptual ``undo list'' which
3528 explains how earlier versions were different from the current version.
3529 People can see the current version; they don't need the change log to
3530 tell them what is in it. What they want from a change log is a clear
3531 explanation of how the earlier version differed. Each @dfn{entry} in
3532 a change log describes either an individual change or the smallest
3533 batch of changes that belong together, also known as a @dfn{change
3535 @cindex title, change log entry
3536 @cindex description, change log entry
3537 For later reference or for summarizing, sometimes it is useful to
3538 start the entry with a one-line description (sometimes called a
3539 @dfn{title}) to describe its overall purpose.
3541 In the past, we recommended not mentioning changes in non-software
3542 files (manuals, help files, media files, etc.)@: in change logs.
3543 However, we've been advised that it is a good idea to include them,
3544 for the sake of copyright records.
3546 The change log file is normally called @file{ChangeLog} and covers an
3547 entire directory. Each directory can have its own change log, or a
3548 directory can use the change log of its parent directory---it's up to
3551 Another alternative is to record change log information with a version
3552 control system such as RCS or CVS. This can be converted automatically
3553 to a @file{ChangeLog} file using @code{rcs2log}; in Emacs, the command
3554 @kbd{C-x v a} (@code{vc-update-change-log}) does the job.
3556 For changes to code, there's no need to describe the full purpose of
3557 the changes or how they work together. If you think that a change
3558 calls for explanation, you're probably right. Please do explain
3559 it---but please put the full explanation in comments in the code,
3560 where people will see it whenever they see the code. For example,
3561 ``New function'' is enough for the change log when you add a function,
3562 because there should be a comment before the function definition to
3563 explain what it does.
3565 For changes to files that do not support a comment syntax (e.g., media
3566 files), it is ok to include the full explanation in the change log file,
3567 after the title and before the list of individual changes.
3569 The easiest way to add an entry to @file{ChangeLog} is with the Emacs
3570 command @kbd{M-x add-change-log-entry}. An individual change should
3571 have an asterisk, the name of the changed file, and then in
3572 parentheses the name of the changed functions, variables or whatever,
3573 followed by a colon. Then describe the changes you made to that
3574 function or variable.
3577 @node Style of Change Logs
3578 @subsection Style of Change Logs
3579 @cindex change logs, style
3581 Here are some simple examples of change log entries, starting with the
3582 header line that says who made the change and when it was installed,
3583 followed by descriptions of specific changes. (These examples are
3584 drawn from Emacs and GCC.)
3587 1998-08-17 Richard Stallman <rms@@gnu.org>
3589 * register.el (insert-register): Return nil.
3590 (jump-to-register): Likewise.
3592 * sort.el (sort-subr): Return nil.
3594 * tex-mode.el (tex-bibtex-file, tex-file, tex-region):
3595 Restart the tex shell if process is gone or stopped.
3596 (tex-shell-running): New function.
3598 * expr.c (store_one_arg): Round size up for move_block_to_reg.
3599 (expand_call): Round up when emitting USE insns.
3600 * stmt.c (assign_parms): Round size up for move_block_from_reg.
3603 It's important to name the changed function or variable in full. Don't
3604 abbreviate function or variable names, and don't combine them.
3605 Subsequent maintainers will often search for a function name to find all
3606 the change log entries that pertain to it; if you abbreviate the name,
3607 they won't find it when they search.
3609 For example, some people are tempted to abbreviate groups of function
3610 names by writing @samp{* register.el (@{insert,jump-to@}-register)};
3611 this is not a good idea, since searching for @code{jump-to-register} or
3612 @code{insert-register} would not find that entry.
3614 Separate unrelated change log entries with blank lines. Don't put
3615 blank lines between individual changes of an entry. You can omit the
3616 file name and the asterisk when successive individual changes are in
3619 Break long lists of function names by closing continued lines with
3620 @samp{)}, rather than @samp{,}, and opening the continuation with
3621 @samp{(} as in this example:
3624 * keyboard.c (menu_bar_items, tool_bar_items)
3625 (Fexecute_extended_command): Deal with 'keymap' property.
3628 When you install someone else's changes, put the contributor's name in
3629 the change log entry rather than in the text of the entry. In other
3633 2002-07-14 John Doe <jdoe@@gnu.org>
3635 * sewing.c: Make it sew.
3642 2002-07-14 Usual Maintainer <usual@@gnu.org>
3644 * sewing.c: Make it sew. Patch by jdoe@@gnu.org.
3647 As for the date, that should be the date you applied the change.
3649 @node Simple Changes
3650 @subsection Simple Changes
3652 Certain simple kinds of changes don't need much detail in the change
3655 When you change the calling sequence of a function in a simple fashion,
3656 and you change all the callers of the function to use the new calling
3657 sequence, there is no need to make individual entries for all the
3658 callers that you changed. Just write in the entry for the function
3659 being called, ``All callers changed''---like this:
3662 * keyboard.c (Fcommand_execute): New arg SPECIAL.
3663 All callers changed.
3666 When you change just comments or doc strings, it is enough to write an
3667 entry for the file, without mentioning the functions. Just ``Doc
3668 fixes'' is enough for the change log.
3670 There's no technical need to make change log entries for documentation
3671 files. This is because documentation is not susceptible to bugs that
3672 are hard to fix. Documentation does not consist of parts that must
3673 interact in a precisely engineered fashion. To correct an error, you
3674 need not know the history of the erroneous passage; it is enough to
3675 compare what the documentation says with the way the program actually
3678 However, you should keep change logs for documentation files when the
3679 project gets copyright assignments from its contributors, so as to
3680 make the records of authorship more accurate.
3682 @node Conditional Changes
3683 @subsection Conditional Changes
3684 @cindex conditional changes, and change logs
3685 @cindex change logs, conditional changes
3687 Source files can often contain code that is conditional to build-time
3688 or static conditions. For example, C programs can contain
3689 compile-time @code{#if} conditionals; programs implemented in
3690 interpreted languages can contain module imports of function
3691 definitions that are only performed for certain versions of the
3692 interpreter; and Automake @file{Makefile.am} files can contain
3693 variable definitions or target declarations that are only to be
3694 considered if a configure-time Automake conditional is true.
3696 Many changes are conditional as well: sometimes you add a new variable,
3697 or function, or even a new program or library, which is entirely
3698 dependent on a build-time condition. It is useful to indicate
3699 in the change log the conditions for which a change applies.
3701 Our convention for indicating conditional changes is to use
3702 @emph{square brackets around the name of the condition}.
3704 Conditional changes can happen in numerous scenarios and with many
3705 variations, so here are some examples to help clarify. This first
3706 example describes changes in C, Perl, and Python files which are
3707 conditional but do not have an associated function or entity name:
3710 * xterm.c [SOLARIS2]: Include <string.h>.
3711 * FilePath.pm [$^O eq 'VMS']: Import the VMS::Feature module.
3712 * framework.py [sys.version_info < (2, 6)]: Make "with" statement
3713 available by importing it from __future__,
3714 to support also python 2.5.
3717 Our other examples will for simplicity be limited to C, as the minor
3718 changes necessary to adapt them to other languages should be
3721 Next, here is an entry describing a new definition which is entirely
3722 conditional: the C macro @code{FRAME_WINDOW_P} is defined (and used)
3723 only when the macro @code{HAVE_X_WINDOWS} is defined:
3726 * frame.h [HAVE_X_WINDOWS] (FRAME_WINDOW_P): Macro defined.
3729 Next, an entry for a change within the function @code{init_display},
3730 whose definition as a whole is unconditional, but the changes
3731 themselves are contained in a @samp{#ifdef HAVE_LIBNCURSES}
3735 * dispnew.c (init_display) [HAVE_LIBNCURSES]: If X, call tgetent.
3738 Finally, here is an entry for a change that takes effect only when
3739 a certain macro is @emph{not} defined:
3742 * host.c (gethostname) [!HAVE_SOCKETS]: Replace with winsock version.
3746 @node Indicating the Part Changed
3747 @subsection Indicating the Part Changed
3749 Indicate the part of a function which changed by using angle brackets
3750 enclosing an indication of what the changed part does. Here is an entry
3751 for a change in the part of the function @code{sh-while-getopts} that
3752 deals with @code{sh} commands:
3755 * progmodes/sh-script.el (sh-while-getopts) <sh>: Handle case that
3756 user-specified option string is empty.
3764 In the GNU project, man pages are secondary. It is not necessary or
3765 expected for every GNU program to have a man page, but some of them do.
3766 It's your choice whether to include a man page in your program.
3768 When you make this decision, consider that supporting a man page
3769 requires continual effort each time the program is changed. The time
3770 you spend on the man page is time taken away from more useful work.
3772 For a simple program which changes little, updating the man page may be
3773 a small job. Then there is little reason not to include a man page, if
3776 For a large program that changes a great deal, updating a man page may
3777 be a substantial burden. If a user offers to donate a man page, you may
3778 find this gift costly to accept. It may be better to refuse the man
3779 page unless the same person agrees to take full responsibility for
3780 maintaining it---so that you can wash your hands of it entirely. If
3781 this volunteer later ceases to do the job, then don't feel obliged to
3782 pick it up yourself; it may be better to withdraw the man page from the
3783 distribution until someone else agrees to update it.
3785 When a program changes only a little, you may feel that the
3786 discrepancies are small enough that the man page remains useful without
3787 updating. If so, put a prominent note near the beginning of the man
3788 page explaining that you don't maintain it and that the Texinfo manual
3789 is more authoritative. The note should say how to access the Texinfo
3792 Be sure that man pages include a copyright statement and free license.
3793 The simple all-permissive license is appropriate for simple man pages
3794 (@pxref{License Notices for Other Files,,,maintain,Information for GNU
3797 For long man pages, with enough explanation and documentation that
3798 they can be considered true manuals, use the GFDL (@pxref{License for
3801 Finally, the GNU help2man program
3802 (@uref{http://www.gnu.org/software/help2man/}) is one way to automate
3803 generation of a man page, in this case from @option{--help} output.
3804 This is sufficient in many cases.
3806 @node Reading other Manuals
3807 @section Reading other Manuals
3809 There may be non-free books or documentation files that describe the
3810 program you are documenting.
3812 It is ok to use these documents for reference, just as the author of a
3813 new algebra textbook can read other books on algebra. A large portion
3814 of any non-fiction book consists of facts, in this case facts about how
3815 a certain program works, and these facts are necessarily the same for
3816 everyone who writes about the subject. But be careful not to copy your
3817 outline structure, wording, tables or examples from preexisting non-free
3818 documentation. Copying from free documentation may be ok; please check
3819 with the FSF about the individual case.
3821 @node Managing Releases
3822 @chapter The Release Process
3825 Making a release is more than just bundling up your source files in a
3826 tar file and putting it up for FTP. You should set up your software so
3827 that it can be configured to run on a variety of systems. Your Makefile
3828 should conform to the GNU standards described below, and your directory
3829 layout should also conform to the standards discussed below. Doing so
3830 makes it easy to include your package into the larger framework of
3834 * Configuration:: How configuration of GNU packages should work.
3835 * Makefile Conventions:: Makefile conventions.
3836 * Releases:: Making releases
3840 @section How Configuration Should Work
3841 @cindex program configuration
3844 Each GNU distribution should come with a shell script named
3845 @code{configure}. This script is given arguments which describe the
3846 kind of machine and system you want to compile the program for.
3847 The @code{configure} script must record the configuration options so
3848 that they affect compilation.
3850 The description here is the specification of the interface for the
3851 @code{configure} script in GNU packages. Many packages implement it
3852 using GNU Autoconf (@pxref{Top,, Introduction, autoconf, Autoconf})
3853 and/or GNU Automake (@pxref{Top,, Introduction, automake, Automake}),
3854 but you do not have to use these tools. You can implement it any way
3855 you like; for instance, by making @code{configure} be a wrapper around
3856 a completely different configuration system.
3858 Another way for the @code{configure} script to operate is to make a
3859 link from a standard name such as @file{config.h} to the proper
3860 configuration file for the chosen system. If you use this technique,
3861 the distribution should @emph{not} contain a file named
3862 @file{config.h}. This is so that people won't be able to build the
3863 program without configuring it first.
3865 Another thing that @code{configure} can do is to edit the Makefile. If
3866 you do this, the distribution should @emph{not} contain a file named
3867 @file{Makefile}. Instead, it should include a file @file{Makefile.in} which
3868 contains the input used for editing. Once again, this is so that people
3869 won't be able to build the program without configuring it first.
3871 If @code{configure} does write the @file{Makefile}, then @file{Makefile}
3872 should have a target named @file{Makefile} which causes @code{configure}
3873 to be rerun, setting up the same configuration that was set up last
3874 time. The files that @code{configure} reads should be listed as
3875 dependencies of @file{Makefile}.
3877 All the files which are output from the @code{configure} script should
3878 have comments at the beginning explaining that they were generated
3879 automatically using @code{configure}. This is so that users won't think
3880 of trying to edit them by hand.
3882 The @code{configure} script should write a file named @file{config.status}
3883 which describes which configuration options were specified when the
3884 program was last configured. This file should be a shell script which,
3885 if run, will recreate the same configuration.
3887 The @code{configure} script should accept an option of the form
3888 @samp{--srcdir=@var{dirname}} to specify the directory where sources are found
3889 (if it is not the current directory). This makes it possible to build
3890 the program in a separate directory, so that the actual source directory
3893 If the user does not specify @samp{--srcdir}, then @code{configure} should
3894 check both @file{.} and @file{..} to see if it can find the sources. If
3895 it finds the sources in one of these places, it should use them from
3896 there. Otherwise, it should report that it cannot find the sources, and
3897 should exit with nonzero status.
3899 Usually the easy way to support @samp{--srcdir} is by editing a
3900 definition of @code{VPATH} into the Makefile. Some rules may need to
3901 refer explicitly to the specified source directory. To make this
3902 possible, @code{configure} can add to the Makefile a variable named
3903 @code{srcdir} whose value is precisely the specified directory.
3905 In addition, the @samp{configure} script should take options
3906 corresponding to most of the standard directory variables
3907 (@pxref{Directory Variables}). Here is the list:
3910 --prefix --exec-prefix --bindir --sbindir --libexecdir --sysconfdir
3911 --sharedstatedir --localstatedir --runstatedir
3912 --libdir --includedir --oldincludedir
3913 --datarootdir --datadir --infodir --localedir --mandir --docdir
3914 --htmldir --dvidir --pdfdir --psdir
3917 The @code{configure} script should also take an argument which specifies the
3918 type of system to build the program for. This argument should look like
3922 @var{cpu}-@var{company}-@var{system}
3925 For example, an Athlon-based GNU/Linux system might be
3926 @samp{i686-pc-linux-gnu}.
3928 The @code{configure} script needs to be able to decode all plausible
3929 alternatives for how to describe a machine. Thus,
3930 @samp{athlon-pc-gnu/linux} would be a valid alias. There is a shell
3932 @uref{http://git.savannah.gnu.org/@/gitweb/@/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD,
3933 @file{config.sub}} that you can use as a subroutine to validate system
3934 types and canonicalize aliases.
3936 The @code{configure} script should also take the option
3937 @option{--build=@var{buildtype}}, which should be equivalent to a
3938 plain @var{buildtype} argument. For example, @samp{configure
3939 --build=i686-pc-linux-gnu} is equivalent to @samp{configure
3940 i686-pc-linux-gnu}. When the build type is not specified by an option
3941 or argument, the @code{configure} script should normally guess it using
3943 @uref{http://git.savannah.gnu.org/@/gitweb/@/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD,
3944 @file{config.guess}}.
3946 @cindex optional features, configure-time
3947 Other options are permitted to specify in more detail the software
3948 or hardware present on the machine, to include or exclude optional parts
3949 of the package, or to adjust the name of some tools or arguments to them:
3952 @item --enable-@var{feature}@r{[}=@var{parameter}@r{]}
3953 Configure the package to build and install an optional user-level
3954 facility called @var{feature}. This allows users to choose which
3955 optional features to include. Giving an optional @var{parameter} of
3956 @samp{no} should omit @var{feature}, if it is built by default.
3958 No @samp{--enable} option should @strong{ever} cause one feature to
3959 replace another. No @samp{--enable} option should ever substitute one
3960 useful behavior for another useful behavior. The only proper use for
3961 @samp{--enable} is for questions of whether to build part of the program
3964 @item --with-@var{package}
3965 @c @r{[}=@var{parameter}@r{]}
3966 The package @var{package} will be installed, so configure this package
3967 to work with @var{package}.
3969 @c Giving an optional @var{parameter} of
3970 @c @samp{no} should omit @var{package}, if it is used by default.
3972 Possible values of @var{package} include
3973 @samp{gnu-as} (or @samp{gas}), @samp{gnu-ld}, @samp{gnu-libc},
3979 Do not use a @samp{--with} option to specify the file name to use to
3980 find certain files. That is outside the scope of what @samp{--with}
3983 @item @var{variable}=@var{value}
3984 Set the value of the variable @var{variable} to @var{value}. This is
3985 used to override the default values of commands or arguments in the
3986 build process. For example, the user could issue @samp{configure
3987 CFLAGS=-g CXXFLAGS=-g} to build with debugging information and without
3988 the default optimization.
3990 Specifying variables as arguments to @code{configure}, like this:
3994 is preferable to setting them in environment variables:
3998 as it helps to recreate the same configuration later with
3999 @file{config.status}. However, both methods should be supported.
4002 All @code{configure} scripts should accept all of the ``detail''
4003 options and the variable settings, whether or not they make any
4004 difference to the particular package at hand. In particular, they
4005 should accept any option that starts with @samp{--with-} or
4006 @samp{--enable-}. This is so users will be able to configure an
4007 entire GNU source tree at once with a single set of options.
4009 You will note that the categories @samp{--with-} and @samp{--enable-}
4010 are narrow: they @strong{do not} provide a place for any sort of option
4011 you might think of. That is deliberate. We want to limit the possible
4012 configuration options in GNU software. We do not want GNU programs to
4013 have idiosyncratic configuration options.
4015 Packages that perform part of the compilation process may support
4016 cross-compilation. In such a case, the host and target machines for the
4017 program may be different.
4019 The @code{configure} script should normally treat the specified type of
4020 system as both the host and the target, thus producing a program which
4021 works for the same type of machine that it runs on.
4023 To compile a program to run on a host type that differs from the build
4024 type, use the configure option @option{--host=@var{hosttype}}, where
4025 @var{hosttype} uses the same syntax as @var{buildtype}. The host type
4026 normally defaults to the build type.
4028 To configure a cross-compiler, cross-assembler, or what have you, you
4029 should specify a target different from the host, using the configure
4030 option @samp{--target=@var{targettype}}. The syntax for
4031 @var{targettype} is the same as for the host type. So the command would
4035 ./configure --host=@var{hosttype} --target=@var{targettype}
4038 The target type normally defaults to the host type.
4039 Programs for which cross-operation is not meaningful need not accept the
4040 @samp{--target} option, because configuring an entire operating system for
4041 cross-operation is not a meaningful operation.
4043 Some programs have ways of configuring themselves automatically. If
4044 your program is set up to do this, your @code{configure} script can simply
4045 ignore most of its arguments.
4047 @comment The makefile standards are in a separate file that is also
4048 @comment included by make.texinfo. Done by roland@gnu.ai.mit.edu on 1/6/93.
4049 @comment For this document, turn chapters into sections, etc.
4051 @include make-stds.texi
4055 @section Making Releases
4058 @cindex version numbers, for releases
4059 You should identify each release with a pair of version numbers, a
4060 major version and a minor. We have no objection to using more than
4061 two numbers, but it is very unlikely that you really need them.
4063 Package the distribution of @code{Foo version 69.96} up in a gzipped tar
4064 file with the name @file{foo-69.96.tar.gz}. It should unpack into a
4065 subdirectory named @file{foo-69.96}.
4067 Building and installing the program should never modify any of the files
4068 contained in the distribution. This means that all the files that form
4069 part of the program in any way must be classified into @dfn{source
4070 files} and @dfn{non-source files}. Source files are written by humans
4071 and never changed automatically; non-source files are produced from
4072 source files by programs under the control of the Makefile.
4074 @cindex @file{README} file
4075 The distribution should contain a file named @file{README} with a
4076 general overview of the package:
4079 @item the name of the package;
4081 @item the version number of the package, or refer to where in the
4082 package the version can be found;
4084 @item a general description of what the package does;
4086 @item a reference to the file @file{INSTALL}, which
4087 should in turn contain an explanation of the installation procedure;
4089 @item a brief explanation of any unusual top-level directories or
4090 files, or other hints for readers to find their way around the source;
4092 @item a reference to the file which contains the copying conditions.
4093 The GNU GPL, if used, should be in a file called @file{COPYING}. If
4094 the GNU LGPL is used, it should be in a file called
4095 @file{COPYING.LESSER}.
4098 Naturally, all the source files must be in the distribution. It is
4099 okay to include non-source files in the distribution along with the
4100 source files they are generated from, provided they are up-to-date
4101 with the source they are made from, and machine-independent, so that
4102 normal building of the distribution will never modify them. We
4103 commonly include non-source files produced by Autoconf, Automake,
4104 Bison, @code{flex}, @TeX{}, and @code{makeinfo}; this helps avoid
4105 unnecessary dependencies between our distributions, so that users can
4106 install whichever versions of whichever packages they like. Do not
4107 induce new dependencies on other software lightly.
4109 Non-source files that might actually be modified by building and
4110 installing the program should @strong{never} be included in the
4111 distribution. So if you do distribute non-source files, always make
4112 sure they are up to date when you make a new distribution.
4114 Make sure that all the files in the distribution are world-readable, and
4115 that directories are world-readable and world-searchable (octal mode 755).
4116 We used to recommend that all directories in the distribution also be
4117 world-writable (octal mode 777), because ancient versions of @code{tar}
4118 would otherwise not cope when extracting the archive as an unprivileged
4119 user. That can easily lead to security issues when creating the archive,
4120 however, so now we recommend against that.
4122 Don't include any symbolic links in the distribution itself. If the tar
4123 file contains symbolic links, then people cannot even unpack it on
4124 systems that don't support symbolic links. Also, don't use multiple
4125 names for one file in different directories, because certain file
4126 systems cannot handle this and that prevents unpacking the
4129 Try to make sure that all the file names will be unique on MS-DOS. A
4130 name on MS-DOS consists of up to 8 characters, optionally followed by a
4131 period and up to three characters. MS-DOS will truncate extra
4132 characters both before and after the period. Thus,
4133 @file{foobarhacker.c} and @file{foobarhacker.o} are not ambiguous; they
4134 are truncated to @file{foobarha.c} and @file{foobarha.o}, which are
4137 @cindex @file{texinfo.tex}, in a distribution
4138 Include in your distribution a copy of the @file{texinfo.tex} you used
4139 to test print any @file{*.texinfo} or @file{*.texi} files.
4141 Likewise, if your program uses small GNU software packages like regex,
4142 getopt, obstack, or termcap, include them in the distribution file.
4143 Leaving them out would make the distribution file a little smaller at
4144 the expense of possible inconvenience to a user who doesn't know what
4148 @chapter References to Non-Free Software and Documentation
4149 @cindex references to non-free material
4151 A GNU program should not recommend, promote, or grant legitimacy to
4152 the use of any non-free program. Proprietary software is a social and
4153 ethical problem, and our aim is to put an end to that problem. We
4154 can't stop some people from writing proprietary programs, or stop
4155 other people from using them, but we can and should refuse to
4156 advertise them to new potential customers, or to give the public the
4157 idea that their existence is ethical.
4159 The GNU definition of free software is found on the GNU web site at
4160 @url{http://www.gnu.org/@/philosophy/@/free-sw.html}, and the definition
4161 of free documentation is found at
4162 @url{http://www.gnu.org/@/philosophy/@/free-doc.html}. The terms ``free''
4163 and ``non-free'', used in this document, refer to those definitions.
4165 A list of important licenses and whether they qualify as free is in
4166 @url{http://www.gnu.org/@/licenses/@/license-list.html}. If it is not
4167 clear whether a license qualifies as free, please ask the GNU Project
4168 by writing to @email{licensing@@gnu.org}. We will answer, and if the
4169 license is an important one, we will add it to the list.
4171 When a non-free program or system is well known, you can mention it in
4172 passing---that is harmless, since users who might want to use it
4173 probably already know about it. For instance, it is fine to explain
4174 how to build your package on top of some widely used non-free
4175 operating system, or how to use it together with some widely used
4178 However, you should give only the necessary information to help those
4179 who already use the non-free program to use your program with
4180 it---don't give, or refer to, any further information about the
4181 proprietary program, and don't imply that the proprietary program
4182 enhances your program, or that its existence is in any way a good
4183 thing. The goal should be that people already using the proprietary
4184 program will get the advice they need about how to use your free
4185 program with it, while people who don't already use the proprietary
4186 program will not see anything likely to lead them to take an interest
4189 If a non-free program or system is obscure in your program's domain,
4190 your program should not mention or support it at all, since doing so
4191 would tend to popularize the non-free program more than it popularizes
4192 your program. (You cannot hope to find many additional users for your
4193 program among the users of Foobar, if the existence of Foobar is not
4194 generally known among people who might want to use your program.)
4196 Sometimes a program is free software in itself but depends on a
4197 non-free platform in order to run. For instance, many Java programs
4198 depend on some non-free Java libraries. To recommend or promote such
4199 a program is to promote the other programs it needs. This is why we
4200 are careful about listing Java programs in the Free Software
4201 Directory: we don't want to promote the non-free Java libraries.
4203 We hope this particular problem with Java will be gone by and by, as
4204 we replace the remaining non-free standard Java libraries with free
4205 software, but the general principle will remain the same: don't
4206 recommend, promote or legitimize programs that depend on non-free
4209 Some free programs strongly encourage the use of non-free software. A
4210 typical example is @command{mplayer}. It is free software in itself,
4211 and the free code can handle some kinds of files. However,
4212 @command{mplayer} recommends use of non-free codecs for other kinds of
4213 files, and users that install @command{mplayer} are very likely to
4214 install those codecs along with it. To recommend @command{mplayer}
4215 is, in effect, to promote use of the non-free codecs.
4217 Thus, you should not recommend programs that strongly encourage the
4218 use of non-free software. This is why we do not list
4219 @command{mplayer} in the Free Software Directory.
4221 A GNU package should not refer the user to any non-free documentation
4222 for free software. Free documentation that can be included in free
4223 operating systems is essential for completing the GNU system, or any
4224 free operating system, so encouraging it is a priority; to recommend
4225 use of documentation that we are not allowed to include undermines the
4226 impetus for the community to produce documentation that we can
4227 include. So GNU packages should never recommend non-free
4230 By contrast, it is ok to refer to journal articles and textbooks in
4231 the comments of a program for explanation of how it functions, even
4232 though they are non-free. This is because we don't include such
4233 things in the GNU system even if they are free---they are outside the
4234 scope of what a software distribution needs to include.
4236 Referring to a web site that describes or recommends a non-free
4237 program is promoting that program, so please do not make links to (or
4238 mention by name) web sites that contain such material. This policy is
4239 relevant particularly for the web pages for a GNU package.
4241 Following links from nearly any web site can lead eventually to
4242 non-free software; this is inherent in the nature of the web. So it
4243 makes no sense to criticize a site for having such links. As long as
4244 the site does not itself recommend a non-free program, there is no
4245 need to consider the question of the sites that it links to for other
4248 Thus, for example, you should not refer to AT&T's web site if that
4249 recommends AT&T's non-free software packages; you should not refer to
4250 a site that links to AT&T's site presenting it as a place to get some
4251 non-free program, because that link recommends and legitimizes the
4252 non-free program. However, that a site contains a link to AT&T's web
4253 site for some other purpose (such as long-distance telephone service)
4254 is not an objection against it.
4256 @node GNU Free Documentation License
4257 @appendix GNU Free Documentation License
4259 @cindex FDL, GNU Free Documentation License
4269 eval: (add-hook 'write-file-hooks 'time-stamp)
4270 time-stamp-start: "@set lastupdate "
4272 time-stamp-format: "%:b %:d, %:y"
4273 compile-command: "cd work.s && make"