From: Subject: GNU Coding Standards Date: Mon, 28 May 2007 20:22:34 +0530 MIME-Version: 1.0 Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Content-Location: http://www.gnu.org/prep/standards/standards.html X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3028 GNU Coding Standards

GNU Coding Standards

Table of Contents

Next: Preface, Previous: (dir), Up: (dir)=20

Version

The GNU coding standards, last updated January 22, 2007.=20

Copyright =A9 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, = 2001, 2002,=20 2003, 2004, 2005, 2006, 2007 Free Software Foundation, Inc.=20

Permission is granted to copy, distribute and/or modify this document = under=20 the terms of the GNU Free Documentation License, Version 1.2 or any = later=20 version published by the Free Software Foundation; with no Invariant = Sections,=20 with no Front-Cover Texts, and with no Back-Cover Texts. A copy of the = license=20 is included in the section entitled =93GNU Free Documentation = License=94.=20

Next: Legal Issues, Previous: Top, Up: Top=20

1 About the GNU Coding Standards

The GNU Coding Standards were written by Richard Stallman and other = GNU=20 Project volunteers. Their purpose is to make the GNU system clean, = consistent,=20 and easy to install. This document can also be read as a guide to = writing=20 portable, robust and reliable programs. It focuses on programs written = in C, but=20 many of the rules and principles are useful even if you write in another = programming language. The rules often state reasons for writing in a = certain=20 way.=20

This release of the GNU Coding Standards was last updated January 22, = 2007.=20

= If you did not obtain this = file=20 directly from the GNU project and recently, please check for a newer = version.=20 You can get the GNU Coding Standards from the GNU web server in many = different=20 formats, including the Texinfo source, PDF, HTML, DVI, plain text, and = more, at:=20 http://www.gnu.org/prep/stand= ards/.=20

Corrections or suggestions for this document should be sent to bug-standards@gnu.org. If you = make a=20 suggestion, please include a suggested new wording for it; our time is = limited.=20 We prefer a context diff to the standards.texi or make-stds.texi files, but if you don't have = those=20 files, please mail your suggestion anyway.=20

These standards cover the minimum of what is important when writing a = GNU=20 package. Likely, the needs for additional standards will come up. = Sometimes, you=20 might suggest that such standards be added to this document. If you = think your=20 standards would be generally useful, please do suggest them.=20

You should also set standards for your package on many questions not=20 addressed or not firmly specified here. The most important point is to = be=20 self-consistent=97try to stick to the conventions you pick, and try to = document=20 them as much as possible. That way, your program will be more = maintainable by=20 others.=20

The GNU Hello program serves as an example of how to follow the GNU = coding=20 standards for a trivial program. http://www.gnu.org/= software/hello/hello.html.=20

Next: Design Advice, Previous: Preface, Up: Top=20

2 Keeping Free Software Free

This chapter discusses how you = can make=20 sure that GNU software avoids legal difficulties, and other related = issues.=20

Next: Contributions, Up: Legal Issues

2.1 Referring to Proprietary Programs

Don't in any circumstances = refer to=20 Unix source code for or during your work on GNU! (Or to any other = proprietary=20 programs.)=20

If you have a vague recollection of the internals of a Unix program, = this=20 does not absolutely mean you can't write an imitation of it, but do try = to=20 organize the imitation internally along different lines, because this is = likely=20 to make the details of the Unix version irrelevant and dissimilar to = your=20 results.=20

For example, Unix utilities were generally optimized to minimize = memory use;=20 if you go for speed instead, your program will be very different. You = could keep=20 the entire input file in memory and scan it there instead of using = stdio. Use a=20 smarter algorithm discovered more recently than the Unix program. = Eliminate use=20 of temporary files. Do it in one pass instead of two (we did this in the = assembler).=20

Or, on the contrary, emphasize simplicity instead of speed. For some=20 applications, the speed of today's computers makes simpler algorithms = adequate.=20

Or go for generality. For example, Unix programs often have static = tables or=20 fixed-size strings, which make for arbitrary limits; use dynamic = allocation=20 instead. Make sure your program handles NULs and other funny characters = in the=20 input files. Add a programming language for extensibility and write part = of the=20 program in that language.=20

Or turn some parts of the program into independently usable = libraries. Or use=20 a simple garbage collector instead of tracking precisely when to free = memory, or=20 use a new GNU facility such as obstacks.=20

Next: Trademarks, Previous: Reading Non-Free Code, Up: Legal Issues

2.2 Accepting Contributions

If the program you are = working on is=20 copyrighted by the Free Software Foundation, then when someone else = sends you a=20 piece of code to add to the program, we need legal papers to use = it=97just as we=20 asked you to sign papers initially. Each person who makes a = nontrivial=20 contribution to a program must sign some sort of legal papers in order = for us to=20 have clear title to the program; the main author alone is not enough.=20

So, before adding in any contributions from other people, please tell = us, so=20 we can arrange to get the papers. Then wait until we tell you that we = have=20 received the signed papers, before you actually use the contribution.=20

This applies both before you release the program and afterward. If = you=20 receive diffs to fix a bug, and they make significant changes, we need = legal=20 papers for that change.=20

This also applies to comments and documentation files. For copyright = law,=20 comments and code are just text. Copyright applies to all kinds of text, = so we=20 need legal papers for all kinds.=20

We know it is frustrating to ask for legal papers; it's frustrating = for us as=20 well. But if you don't wait, you are going out on a limb=97for example, = what if=20 the contributor's employer won't sign a disclaimer? You might have to = take that=20 code out again!=20

You don't need papers for changes of a few lines here or there, since = they=20 are not significant for copyright purposes. Also, you don't need papers = if all=20 you get from the suggestion is some ideas, not actual code which you = use. For=20 example, if someone sent you one implementation, but you write a = different=20 implementation of the same idea, you don't need to get papers.=20

The very worst thing is if you forget to tell us about the other = contributor.=20 We could be very embarrassed in court some day as a result.=20

We have more detailed advice for maintainers of programs; if you have = reached=20 the stage of actually maintaining a program for GNU (whether released or = not),=20 please ask us for a copy. It is also available online for your perusal: = http://www.gnu.org/prep/mainta= in/.=20

Previous: Contributions, Up: Legal Issues

2.3 Trademarks

Please do not include any trademark=20 acknowledgements in GNU software packages or documentation.=20

Trademark acknowledgements are the statements that such-and-such is a = trademark of so-and-so. The GNU Project has no objection to the basic = idea of=20 trademarks, but these acknowledgements feel like kowtowing, and there is = no=20 legal requirement for them, so we don't use them.=20

What is legally required, as regards other people's trademarks, is to = avoid=20 using them in ways which a reader might reasonably understand as naming = or=20 labeling our own programs or activities. For example, since =93Objective = C=94 is (or=20 at least was) a trademark, we made sure to say that we provide a = =93compiler for=20 the Objective C language=94 rather than an =93Objective C compiler=94. = The latter=20 would have been meant as a shorter way of saying the former, but it does = not=20 explicitly state the relationship, so it could be misinterpreted as = using=20 =93Objective C=94 as a label for the compiler rather than for the = language.=20

Please don't use =93win=94 as an abbreviation for Microsoft Windows = in GNU=20 software or documentation. In hacker terminology, calling something a = =93win=94 is a=20 form of praise. If you wish to praise Microsoft Windows when speaking on = your=20 own, by all means do so, but not in GNU software. Usually we write the = name=20 =93Windows=94 in full, but when brevity is very important (as in file = names and=20 sometimes symbol names), we abbreviate it to =93w=94. For instance, the = files and=20 functions in Emacs that deal with Windows start with `w32'.=20

Next: Program Behavior, Previous: Legal Issues, Up: Top=20

3 General Program Design

This chapter discusses some of = the issues=20 you should take into account when designing your program.

Next: Compatibility, Up: Design Advice

3.1 Which Languages to Use

When you want to use a = language=20 that gets compiled and runs at high speed, the best language to use is = C. Using=20 another language is like using a non-standard feature: it will cause = trouble for=20 users. Even if GCC supports the other language, users may find it = inconvenient=20 to have to install the compiler for that other language in order to = build your=20 program. For example, if you write your program in C++, people will have = to=20 install the GNU C++ compiler in order to compile your program.=20

C has one other advantage over C++ and other compiled languages: more = people=20 know C, so more people will find it easy to read and modify the program = if it is=20 written in C.=20

So in general it is much better to use C, rather than the comparable=20 alternatives.=20

But there are two exceptions to that conclusion:=20

Many programs are designed to be extensible: they include an = interpreter for=20 a language that is higher level than C. Often much of the program is = written in=20 that language, too. The Emacs editor pioneered this technique.=20

The standard extensibility interpreter = for GNU=20 software is GUILE (http://www.gnu.org/software/g= uile/),=20 which implements the language Scheme (an especially clean and simple = dialect of=20 Lisp). We don't reject programs written in other =93scripting = languages=94 such as=20 Perl and Python, but using GUILE is very important for the overall = consistency=20 of the GNU system.=20

Next: Using Extensions, Previous: Source Language, Up: Design Advice

3.2 Compatibility with Other Implementations

With = occasional=20 exceptions, utility programs and libraries for GNU should be upward = compatible=20 with those in Berkeley Unix, and upward compatible with Standard C if = Standard C=20 specifies their behavior, and upward compatible with posix=20 if posix specifies their behavior.=20

When these standards conflict, it is useful to offer compatibility = modes for=20 each of them.=20

Standard C and posix prohibit many kinds of extensions. Feel free to = make the=20 extensions anyway, and include a `--ansi',=20 `--posix', or `--compatible' option to turn them off. = However, if the=20 extension has a significant chance of breaking any real programs or = scripts,=20 then it is not really upward compatible. So you should try to redesign = its=20 interface to make it upward compatible.=20

Many=20 GNU programs suppress extensions that conflict with posix=20 if the environment variable POSIXLY_CORRECT is defined = (even if it=20 is defined with a null value). Please make your program recognize this = variable=20 if appropriate.=20

When a feature is used only by users (not by programs or command = files), and=20 it is done poorly in Unix, feel free to replace it completely with = something=20 totally different and better. (For example, vi is replaced = with=20 Emacs.) But it is nice to offer a compatible feature as well. (There is = a free=20 vi clone, so we offer it.)=20

Additional useful features are welcome regardless of whether there is = any=20 precedent for them.=20

Next: Standard C, Previous: Compatibility, Up: Design Advice

3.3 Using Non-standard Features

Many GNU = facilities that=20 already exist support a number of convenient extensions over the = comparable Unix=20 facilities. Whether to use these extensions in implementing your program = is a=20 difficult question.=20

On the one hand, using the extensions can make a cleaner program. On = the=20 other hand, people will not be able to build the program unless the = other GNU=20 tools are available. This might cause the program to work on fewer kinds = of=20 machines.=20

With some extensions, it might be easy to provide both alternatives. = For=20 example, you can define functions with a =93keyword=94 = INLINE and=20 define that as a macro to expand into either inline or = nothing,=20 depending on the compiler.=20

In general, perhaps it is best not to use the extensions if you can=20 straightforwardly do without them, but to use the extensions if they are = a big=20 improvement.=20

An exception to this rule are the large, established programs (such = as Emacs)=20 which run on a great variety of systems. Using GNU extensions in such = programs=20 would make many users unhappy, so we don't do that.=20

Another exception is for programs that are used as part of = compilation:=20 anything that must be compiled with other compilers in order to = bootstrap the=20 GNU compilation facilities. If these require the GNU compiler, then no = one can=20 compile them without having them installed already. That would be = extremely=20 troublesome in certain cases.=20

Next: Conditional Compilation, Previous: Using Extensions, Up: Design Advice

3.4 Standard C and Pre-Standard C

1989 = Standard C is=20 widespread enough now that it is ok to use its features in new programs. = There=20 is one exception: do not ever use the =93trigraph=94 feature of Standard = C.=20

1999 Standard C is not widespread yet, so please do not require its = features=20 in programs. It is ok to use its features if they are present.=20

However, it is easy to support pre-standard compilers in most = programs, so if=20 you know how to do that, feel free. If a program you are maintaining has = such=20 support, you should try to keep it working.=20

To support pre-standard C, = instead=20 of writing function definitions in standard prototype form, 

     int
     foo (int x, int y)
     ...

write the definition in pre-standard style like = this, 

     int
     foo (x, y)
          int x, y;
     ...

and use a separate declaration to specify the = argument=20 prototype: 

     int foo (int, int);

You need such a declaration anyway, in a header file, to get the = benefit of=20 prototypes in all the files where the function is called. And once you = have the=20 declaration, you normally lose nothing by writing the function = definition in the=20 pre-standard style.=20

This technique does not work for integer types narrower than=20 int. If you think of an argument as being of a type = narrower than=20 int, declare it as int instead.=20

There are a few special cases where this technique is hard to use. = For=20 example, if a function argument needs to hold the system type=20 dev_t, you run into trouble, because dev_t is = shorter=20 than int on some machines; but you cannot use = int=20 instead, because dev_t is wider than int on = some=20 machines. There is no type you can safely use on all machines in a = non-standard=20 definition. The only way to support non-standard C and pass such an = argument is=20 to check the width of dev_t using Autoconf and choose the = argument=20 type accordingly. This may not be worth the trouble.=20

In order to support pre-standard compilers that do not recognize = prototypes,=20 you may want to use a preprocessor macro like this: 

     /* Declare the prototype for a general external =
function.  */
     #if defined (__STDC__) || defined (WINDOWSNT)
     #define P_(proto) proto
     #else
     #define P_(proto) ()
     #endif

Previous: Standard C, Up: Design Advice

3.5 Conditional Compilation

When supporting configuration options already known when building = your=20 program we prefer using if (... ) over conditional = compilation, as=20 in the former case the compiler is able to perform more extensive = checking of=20 all possible code paths.=20

For example, please write 

       if =
(HAS_FOO)
         ...
       else
         ...

instead of: 

       #ifdef =
HAS_FOO
         ...
       #else
         ...
       #endif

A modern compiler such as GCC will generate exactly the same code in = both=20 cases, and we have been using similar techniques with good success in = several=20 projects. Of course, the former method assumes that HAS_FOO = is=20 defined as either 0 or 1.=20

While this is not a silver bullet solving all portability problems, = and is=20 not always appropriate, following this policy would have saved GCC = developers=20 many hours, or even days, per year.=20

In the case of function-like macros like = REVERSIBLE_CC_MODE in=20 GCC which cannot be simply used in if( ...) statements, = there is an=20 easy workaround. Simply introduce another macro=20 HAS_REVERSIBLE_CC_MODE as in the following example: 

       #ifdef REVERSIBLE_CC_MODE
       #define HAS_REVERSIBLE_CC_MODE 1
       #else
       #define HAS_REVERSIBLE_CC_MODE 0
       #endif

Next: Writing C, Previous: Design Advice, Up: Top=20

4 Program Behavior for All Programs

This chapter describes conventions for writing robust software. It = also=20 describes general standards for error messages, the command line = interface, and=20 how libraries should behave.=20

Next: Semantics, Up: Program Behavior

4.1 Non-GNU Standards

The GNU Project regards standards published by other organizations as = suggestions, not orders. We consider those standards, but we do not = =93obey=94 them.=20 In developing a GNU program, you should implement an outside standard's=20 specifications when that makes the GNU system better overall in an = objective=20 sense. When it doesn't, you shouldn't.=20

In most cases, following published standards is convenient for = users=97it means=20 that their programs or scripts will work more portably. For instance, = GCC=20 implements nearly all the features of Standard C as specified by that = standard.=20 C program developers would be unhappy if it did not. And GNU utilities = mostly=20 follow specifications of POSIX.2; shell script writers and users would = be=20 unhappy if our programs were incompatible.=20

But we do not follow either of these specifications rigidly, and = there are=20 specific points on which we decided not to follow them, so as to make = the GNU=20 system better for users.=20

For instance, Standard C says that nearly all extensions to C are = prohibited.=20 How silly! GCC implements many extensions, some of which were later = adopted as=20 part of the standard. If you want these constructs to give an error = message as=20 =93required=94 by the standard, you must specify `--pedantic', which was implemented only so = that we can=20 say =93GCC is a 100% implementation of the standard,=94 not because = there is any=20 reason to actually use it.=20

POSIX.2 specifies that `df' = and=20 `du' must output sizes by default = in units=20 of 512 bytes. What users want is units of 1k, so that is what we do by = default.=20 If you want the ridiculous behavior =93required=94 by POSIX, you must = set the=20 environment variable `POSIXLY_CORRECT'=20 (which was originally going to be named `POSIX_ME_HARDER').=20

GNU utilities also depart from the letter of the POSIX.2 = specification when=20 they support long-named command-line options, and intermixing options = with=20 ordinary arguments. This minor incompatibility with POSIX is never a = problem in=20 practice, and it is very useful.=20

In particular, don't reject a new feature, or remove an old one, = merely=20 because a standard says it is =93forbidden=94 or =93deprecated.=94=20

Next: Libraries, Previous: Non-GNU Standards, Up: Program Behavior

4.2 Writing Robust Programs

Avoid arbitrary = limits on the=20 length or number of any data structure, including file names, = lines,=20 files, and symbols, by allocating all data structures dynamically. In = most Unix=20 utilities, =93long lines are silently truncated=94. This is not = acceptable in a GNU=20 utility.=20

Utilities = reading=20 files should not drop NUL characters, or any other nonprinting = characters=20 including those with codes above 0177. The only sensible = exceptions=20 would be utilities specifically intended for interface to certain types = of=20 terminals or printers that can't handle those characters. Whenever = possible, try=20 to make programs work properly with sequences of bytes that represent = multibyte=20 characters, using encodings such as UTF-8 and others.=20

Check every system call for an = error=20 return, unless you know you wish to ignore errors. Include the system = error text=20 (from perror or equivalent) in every error message = resulting from a failing system call, as well as the name of the file if = any and=20 the name of the utility. Just =93cannot open foo.c=94 or =93stat = failed=94 is not=20 sufficient.=20

Check every call to=20 malloc or realloc to see if it returned zero. = Check=20 realloc even if you are making the block smaller; in a = system that=20 rounds block sizes to a power of 2, realloc may get a = different=20 block if you ask for less space.=20

In Unix, realloc can destroy the storage block if it = returns=20 zero. GNU realloc does not have this bug: if it fails, the = original=20 block is unchanged. Feel free to assume the bug is fixed. If you wish to = run=20 your program on Unix, and wish to avoid lossage in this case, you can = use the=20 GNU malloc.=20

You must expect free to alter the contents of the block = that was=20 freed. Anything you want to fetch from the block, you must fetch before = calling=20 free.=20

If malloc fails in a noninteractive program, make that a = fatal=20 error. In an interactive program (one that reads commands from the = user), it is=20 better to abort the command and return to the command reader loop. This = allows=20 the user to kill other processes to free up virtual memory, and then try = the=20 command again.=20

Use=20 getopt_long to decode arguments, unless the argument syntax = makes=20 this unreasonable.=20

When static storage is to be written in during program execution, use = explicit C code to initialize it. Reserve C initialized declarations for = data=20 that will not be changed.

Try to avoid low-level interfaces to obscure Unix data structures = (such as=20 file directories, utmp, or the layout of kernel memory), since these are = less=20 likely to work compatibly. If you need to find all the files in a = directory, use=20 readdir or some other high-level interface. These are = supported=20 compatibly by GNU.=20

The preferred signal handling = facilities=20 are the BSD variant of signal, and the posix=20 sigaction function; the alternative USG signal = interface is an inferior design.=20

Nowadays, using the posix signal functions = may be the=20 easiest way to make a program portable. If you use signal, = then on=20 GNU/Linux systems running GNU libc version 1, you should include = bsd/signal.h instead of signal.h, so as to get BSD behavior. It is up = to you=20 whether to support systems where signal has only the USG = behavior,=20 or give up on them.=20

In error checks that = detect=20 =93impossible=94 conditions, just abort. There is usually no point in = printing any=20 message. These checks indicate the existence of bugs. Whoever wants to = fix the=20 bugs will have to read the source code and run a debugger. So explain = the=20 problem with comments in the source. The relevant data will be in = variables,=20 which are easy to examine with the debugger, so there is no point moving = them=20 elsewhere.=20

Do not use a count of errors as the exit status for a program. = That does=20 not work, because exit status values are limited to 8 bits (0 = through 255).=20 A single run of the program might have 256 errors; if you try to return = 256 as=20 the exit status, the parent process will see 0 as the status, and it = will appear=20 that the program succeeded.=20

If = you make=20 temporary files, check the TMPDIR environment variable; if = that=20 variable is defined, use the specified directory instead of /tmp.=20

In addition, be aware that there is a possible security problem when = creating=20 temporary files in world-writable directories. In C, you can avoid this = problem=20 by creating temporary files in this manner: 

     fd =
=3D open(filename, O_WRONLY | O_CREAT | O_EXCL, 0600);

or by using the mkstemps function from = libiberty.=20

In bash, use set -C to avoid this problem.=20

Next: Errors, Previous: Semantics, Up: Program Behavior

4.3 Library Behavior

Try to make library functions = reentrant. If=20 they need to do dynamic storage allocation, at least try to avoid any=20 nonreentrancy aside from that of malloc itself.=20

Here are certain name conventions for libraries, to avoid name = conflicts.=20

Choose a name prefix for the library, more than two characters long. = All=20 external function and variable names should start with this prefix. In = addition,=20 there should only be one of these in any given library member. This = usually=20 means putting each one in a separate source file.=20

An exception can be made when two external symbols are always used = together,=20 so that no reasonable program could use one without the other; then they = can=20 both go in the same file.=20

External symbols that are not documented entry points for the user = should=20 have names beginning with `_'. = The=20 `_' should be followed by the = chosen name=20 prefix for the library, to prevent collisions with other libraries. = These can go=20 in the same files with user entry points if you like.=20

Static functions and variables can be used as you like and need not = fit any=20 naming convention.=20

Next: User Interfaces, Previous: Libraries, Up: Program Behavior

4.4 Formatting Error Messages

Error messages from = compilers=20 should look like this: 

     =
source-file-name:lineno: message

If you want to mention the column number, use one of = these=20 formats: 

     =
source-file-name:lineno:column: =
message
     source-file-name:lineno.column: =
message
    =20

Line numbers should start from 1 at the beginning of = the file,=20 and column numbers should start from 1 at the beginning of the line. = (Both of=20 these conventions are chosen for compatibility.) Calculate column = numbers=20 assuming that space and all ASCII printing characters have equal width, = and=20 assuming tab stops every 8 columns.=20

The error message can also give both the starting and ending = positions of the=20 erroneous text. There are several formats so that you can avoid = redundant=20 information such as a duplicate line number. Here are the possible = formats: 

     =
source-file-name:lineno-1.column-1-=
lineno-2.column-2: message
     =
source-file-name:lineno-1.column-1-=
column-2: message
     =
source-file-name:lineno-1-lineno-2: =
message

When an error is spread over several files, you can = use this=20 format: 

     =
file-1:lineno-1.column-1-file-2:lineno-2.column-2: message

Error messages from other noninteractive programs should look like = this: 

     =
program:source-file-name:lineno: =
message

when there is an appropriate source file, or like = this: 

     program: message

when there is no relevant source file.=20

If you want to mention the column number, use this format: 

     =
program:source-file-name:lineno:col=
umn: message

In an interactive program (one that is reading commands from a = terminal), it=20 is better not to include the program name in an error message. The place = to=20 indicate which program is running is in the prompt or with the screen = layout.=20 (When the same program runs with input from a source other than a = terminal, it=20 is not interactive and would do best to print error messages using the=20 noninteractive style.)=20

The string message should not begin with a capital letter = when it=20 follows a program name and/or file name, because that isn't the = beginning of a=20 sentence. (The sentence conceptually starts at the beginning of the = line.) Also,=20 it should not end with a period.=20

Error messages from interactive programs, and other messages such as = usage=20 messages, should start with a capital letter. But they should not end = with a=20 period.=20

Next: Graphical Interfaces, Previous: Errors, Up: Program Behavior

4.5 Standards for Interfaces Generally

Please = don't=20 make the behavior of a utility depend on the name used to invoke it. It = is=20 useful sometimes to make a link to a utility with a different name, and = that=20 should not change what it does.=20

Instead, use a run time option or a compilation switch or both to = select=20 among the alternate behaviors.=20

Likewise,=20 please don't make the behavior of the program depend on the type of = output=20 device it is used with. Device independence is an important principle of = the=20 system's design; do not compromise it merely to save someone from typing = an=20 option now and then. (Variation in error message syntax when using a = terminal is=20 ok, because that is a side issue that people do not depend on.)=20

If you think one behavior is most useful when the output is to a = terminal,=20 and another is most useful when the output is a file or a pipe, then it = is=20 usually best to make the default behavior the one that is useful with = output to=20 a terminal, and have an option for the other behavior.=20

Compatibility requires certain programs to depend on the type of = output=20 device. It would be disastrous if ls or sh did = not do=20 so in the way all users expect. In some of these cases, we supplement = the=20 program with a preferred alternate version that does not depend on the = output=20 device type. For example, we provide a dir program much = like=20 ls except that its default output format is always = multi-column=20 format.=20

Next: Command-Line Interfaces, Previous: User Interfaces, Up: Program Behavior

4.6 Standards for Graphical Interfaces

When you write a program that provides a = graphical=20 user interface, please make it work with X Windows and the GTK+ toolkit = unless=20 the functionality specifically requires some alternative (for example,=20 =93displaying jpeg images while in console mode=94).=20

In addition, please provide a command-line interface to control the=20 functionality. (In many cases, the graphical user interface can be a = separate=20 program which invokes the command-line program.) This is so that the = same jobs=20 can be done from scripts.=20

Please also = consider=20 providing a CORBA interface (for use from GNOME), a library interface = (for use=20 from C), and perhaps a keyboard-driven console interface (for use by = users from=20 console mode). Once you are doing the work to provide the functionality = and the=20 graphical interface, these won't be much extra work.=20

Next: Option Table, Previous: Graphical Interfaces, Up: Program Behavior

4.7 Standards for Command Line Interfaces

It is a good idea to follow the posix guidelines for the command-line options of a = program. The=20 easiest way to do this is to use getopt to parse them. Note = that=20 the GNU version of getopt will normally permit options = anywhere=20 among the arguments unless the special argument `--' is used. This is not what posix specifies; it is a GNU extension.=20

Please define = long-named options=20 that are equivalent to the single-letter Unix-style options. We hope to = make GNU=20 more user friendly this way. This is easy to do with the GNU function=20 getopt_long.=20

One of the advantages of long-named options is that they can be = consistent=20 from program to program. For example, users should be able to expect the = =93verbose=94 option of any GNU program which has one, to be spelled = precisely=20 `--verbose'. To achieve this = uniformity,=20 look at the table of common long-option names when you choose the option = names=20 for your program (see Op= tion=20 Table).=20

It is usually a good idea for file names given as ordinary arguments = to be=20 input files only; any output files would be specified using options = (preferably=20 `-o' or `--output'). Even if you allow an output file = name as an=20 ordinary argument for compatibility, try to provide an option as another = way to=20 specify it. This will lead to more consistency among GNU utilities, and = fewer=20 idiosyncrasies for users to remember.=20

All= programs=20 should support two standard options: `--version' and `--help'. CGI programs should accept these as=20 command-line options, and also if given as the PATH_INFO; for instance, visiting http://example.org/p.cgi/--help<= /A> in a=20 browser should output the same information as invoking `p.cgi --help' from the command line.

--version
This option should direct the program to print information about = its name,=20 version, origin and legal status, all on standard output, and then = exit=20 successfully. Other options and arguments should be ignored once this = is seen,=20 and the program should not perform its normal function.=20

The first line is = meant to be=20 easy for a program to parse; the version number proper starts after = the last=20 space. In addition, it contains the canonical name for this program, = in this=20 format: 

          GNU Emacs 19.30

The program's name should be a constant string;=20 don't compute it from argv[0]. The idea is to = state the=20 standard or canonical name for the program, not its file name. There = are other=20 ways to find out the precise file name where a command is found in=20 PATH.=20

If the program is a subsidiary part of a larger package, mention = the=20 package name in parentheses, like this: 

          =
emacsserver (GNU Emacs) 19.30

If the package has a version number which is = different from=20 this program's version number, you can mention the package version = number just=20 before the close-parenthesis.=20

If you need to mention the version numbers of = libraries=20 which are distributed separately from the package which contains this = program,=20 you can do so by printing an additional line of version info for each = library=20 you want to mention. Use the same format for these lines as for the = first=20 line.=20

Please do not mention all of the libraries that the program uses = =93just for=20 completeness=94=97that would produce a lot of unhelpful clutter. = Please mention=20 library version numbers only if you find in practice that they are = very=20 important to you in debugging.=20

The following line, after the version number line or lines, should = be a=20 copyright notice. If more than one copyright notice is called for, put = each on=20 a separate line.=20

Next should follow a brief statement that the program is free = software, and=20 that users are free to copy and change it on certain conditions. If = the=20 program is covered by the GNU GPL, say so here. Also mention that = there is no=20 warranty, to the extent permitted by law.=20

It is ok to finish the output with a list of the major authors of = the=20 program, as a way of giving credit.=20

Here's an example of output that follows these rules: 

          GNU Emacs 19.34.5
          Copyright (C) 1996 Free Software Foundation, Inc.
          GNU Emacs comes with NO WARRANTY,
          to the extent permitted by law.
          You may redistribute copies of GNU Emacs
          under the terms of the GNU General Public License.
          For more information about these matters,
          see the files named COPYING.

You should adapt this to your program, of course, filling in the = proper=20 year, copyright holder, name of program, and the references to = distribution=20 terms, and changing the rest of the wording as necessary.=20

This copyright notice only needs to mention the most recent year in = which=20 changes were made=97there's no need to list the years for previous = versions'=20 changes. You don't have to mention the name of the program in these = notices,=20 if that is inconvenient, since it appeared in the first line. (The = rules are=20 different for copyright notices in source files; see Copyright=20 Notices.)=20

Translations of the above lines must preserve the validity of the = copyright=20 notices (see Internationalization).=20 If the translation's character set supports it, the `(C)' should be replaced with the copyright = symbol, as=20 follows:=20

=A9=20

Write the word =93Copyright=94 exactly like that, in English. Do = not translate=20 it into another language. International treaties recognize the English = word=20 =93Copyright=94; translations into other languages do not have legal = significance.=20


--help
This option should output brief documentation for how to invoke = the=20 program, on standard output, then exit successfully. Other options and = arguments should be ignored once this is seen, and the program should = not=20 perform its normal function.=20

Near the end of the `--help' option's output there should be a = line that=20 says where to mail bug reports. It should have this format: 

          Report bugs to mailing-address.

Next: Memory Usage, Previous: Command-Line Interfaces, Up: Program Behavior

4.8 Table of Long Options

Here is a table of long = options used by=20 GNU programs. It is surely incomplete, but we aim to list all the = options that a=20 new program might want to be compatible with. If you use names not = already in=20 the table, please send bug-standards@gnu.org a list = of them,=20 with their meanings, so we can update the table.

`after-date'
`-N' in tar. =
`all'
`-a' in du,=20 ls, nm, stty, = uname, and=20 unexpand.
`all-text'
`-a' in diff. =
`almost-all'
`-A' in ls. =
`append'
`-a' in etags, = tee, time; `-r'=20 in tar.
`archive'
`-a' in cp. =
`archive-name'
`-n' in shar. =
`arglength'
`-l' in m4. =
`ascii'
`-a' in diff. =
`assign'
`-v' in gawk. =
`assume-new'
`-W' in Make.
`assume-old'
`-o' in Make.
`auto-check'
`-a' in = recode.
`auto-pager'
`-a' in wdiff. =
`auto-reference'
`-A' in ptx. =
`avoid-wraps'
`-n' in wdiff. =
`background'
For server programs, run in the background.
`backward-search'
`-B' in ctags. =
`basename'
`-f' in shar. =
`batch'
Used in GDB.
`baud'
Used in GDB.
`before'
`-b' in tac. =
`binary'
`-b' in cpio = and=20 diff.
`bits-per-code'
`-b' in shar. =
`block-size'
Used in cpio and tar.
`blocks'
`-b' in head = and=20 tail.
`break-file'
`-b' in ptx. =
`brief'
Used in various programs to make output shorter.
`bytes'
`-c' in head,=20 split, and tail.
`c++'
`-C' in etags. =
`catenate'
`-A' in tar. =
`cd'
Used in various programs to specify the directory to use.
`changes'
`-c' in chgrp = and=20 chown.
`classify'
`-F' in ls. =
`colons'
`-c' in = recode.
`command'
`-c' in su; = `-x' in GDB.
`compare'
`-d' in tar. =
`compat'
Used in gawk.
`compress'
`-Z' in tar = and=20 shar.
`concatenate'
`-A' in tar. =
`confirmation'
`-w' in tar. =
`context'
Used in diff.
`copyleft'
`-W copyleft' in = gawk.=20
`copyright'
`-C' in ptx,=20 recode, and wdiff; `-W=20 copyright' in gawk.
`core'
Used in GDB.
`count'
`-q' in who. =
`count-links'
`-l' in du. =
`create'
Used in tar and cpio.
`cut-mark'
`-c' in shar. =
`cxref'
`-x' in ctags. =
`date'
`-d' in touch. =
`debug'
`-d' in Make and = m4;=20 `-t' in Bison.
`define'
`-D' in m4. =
`defines'
`-d' in Bison and = ctags.=20
`delete'
`-D' in tar. =
`dereference'
`-L' in chgrp, = chown, cpio, du, = ls, and=20 tar.
`dereference-args'
`-D' in du. =
`device'
Specify an I/O device (special file name).
`diacritics'
`-d' in = recode.
`dictionary-order'
`-d' in look. =
`diff'
`-d' in tar. =
`digits'
`-n' in = csplit.
`directory'
Specify the directory to use, in various programs. In = ls, it=20 means to show directories themselves rather than their contents. In=20 rm and ln, it means to not treat links to=20 directories specially.
`discard-all'
`-x' in strip. =
`discard-locals'
`-X' in strip. =
`dry-run'
`-n' in Make.
`ed'
`-e' in diff. =
`elide-empty-files'
`-z' in = csplit.
`end-delete'
`-x' in wdiff. =
`end-insert'
`-z' in wdiff. =
`entire-new-file'
`-N' in diff. =
`environment-overrides'
`-e' in Make.
`eof'
`-e' in xargs. =
`epoch'
Used in GDB.
`error-limit'
Used in makeinfo.
`error-output'
`-o' in m4. =
`escape'
`-b' in ls. =
`exclude-from'
`-X' in tar. =
`exec'
Used in GDB.
`exit'
`-x' in xargs. =
`exit-0'
`-e' in = unshar.
`expand-tabs'
`-t' in diff. =
`expression'
`-e' in sed. =
`extern-only'
`-g' in nm. =
`extract'
`-i' in cpio;=20 `-x' in tar.
`faces'
`-f' in = finger.
`fast'
`-f' in su. =
`fatal-warnings'
`-E' in m4. =
`file'
`-f' in info,=20 gawk, Make, mt, and tar; = `-n' in sed; `-r' in touch.
`field-separator'
`-F' in gawk. =
`file-prefix'
`-b' in Bison.
`file-type'
`-F' in ls. =
`files-from'
`-T' in tar. =
`fill-column'
Used in makeinfo.
`flag-truncation'
`-F' in ptx. =
`fixed-output-files'
`-y' in Bison.
`follow'
`-f' in tail. =
`footnote-style'
Used in makeinfo.
`force'
`-f' in cp,=20 ln, mv, and rm.
`force-prefix'
`-F' in shar. =
`foreground'
For server programs, run in the foreground; in other words, don't = do=20 anything special to run the server in the background.
`format'
Used in ls, time, and ptx. =
`freeze-state'
`-F' in m4. =
`fullname'
Used in GDB.
`gap-size'
`-g' in ptx. =
`get'
`-x' in tar. =
`graphic'
`-i' in ul. =
`graphics'
`-g' in = recode.
`group'
`-g' in = install.
`gzip'
`-z' in tar = and=20 shar.
`hashsize'
`-H' in m4. =
`header'
`-h' in = objdump and=20 recode
`heading'
`-H' in who. =
`help'
Used to ask for brief usage information.
`here-delimiter'
`-d' in shar. =
`hide-control-chars'
`-q' in ls. =
`html'
In makeinfo, output HTML.
`idle'
`-u' in who. =
`ifdef'
`-D' in diff. =
`ignore'
`-I' in ls; = `-x' in recode.
`ignore-all-space'
`-w' in diff. =
`ignore-backups'
`-B' in ls. =
`ignore-blank-lines'
`-B' in diff. =
`ignore-case'
`-f' in look = and=20 ptx; `-i' in=20 diff and wdiff.
`ignore-errors'
`-i' in Make.
`ignore-file'
`-i' in ptx. =
`ignore-indentation'
`-I' in etags. =
`ignore-init-file'
`-f' in Oleo.
`ignore-interrupts'
`-i' in tee. =
`ignore-matching-lines'
`-I' in diff. =
`ignore-space-change'
`-b' in diff. =
`ignore-zeros'
`-i' in tar. =
`include'
`-i' in etags; = `-I' in m4.
`include-dir'
`-I' in Make.
`incremental'
`-G' in tar. =
`info'
`-i', `-l', and `-m' in=20 Finger.
`init-file'
In some programs, specify the name of the file to read as the = user's init=20 file.
`initial'
`-i' in = expand.
`initial-tab'
`-T' in diff. =
`inode'
`-i' in ls. =
`interactive'
`-i' in cp,=20 ln, mv, rm; `-e' in m4; `-p' in xargs; `-w' in tar.
`intermix-type'
`-p' in shar. =
`iso-8601'
Used in date
`jobs'
`-j' in Make.
`just-print'
`-n' in Make.
`keep-going'
`-k' in Make.
`keep-files'
`-k' in = csplit.
`kilobytes'
`-k' in du and = ls.
`language'
`-l' in etags. =
`less-mode'
`-l' in wdiff. =
`level-for-gzip'
`-g' in shar. =
`line-bytes'
`-C' in split. =
`lines'
Used in split, head, and = tail.
`link'
`-l' in cpio. =
`lint'
`lint-old'
Used in gawk.
`list'
`-t' in cpio;=20 `-l' in recode. =
`list'
`-t' in tar. =
`literal'
`-N' in ls. =
`load-average'
`-l' in Make.
`login'
Used in su.
`machine'
Used in uname.
`macro-name'
`-M' in ptx. =
`mail'
`-m' in hello = and=20 uname.
`make-directories'
`-d' in cpio. =
`makefile'
`-f' in Make.
`mapped'
Used in GDB.
`max-args'
`-n' in xargs. =
`max-chars'
`-n' in xargs. =
`max-lines'
`-l' in xargs. =
`max-load'
`-l' in Make.
`max-procs'
`-P' in xargs. =
`mesg'
`-T' in who. =
`message'
`-T' in who. =
`minimal'
`-d' in diff. =
`mixed-uuencode'
`-M' in shar. =
`mode'
`-m' in = install,=20 mkdir, and mkfifo.
`modification-time'
`-m' in tar. =
`multi-volume'
`-M' in tar. =
`name-prefix'
`-a' in Bison.
`nesting-limit'
`-L' in m4. =
`net-headers'
`-a' in shar. =
`new-file'
`-W' in Make.
`no-builtin-rules'
`-r' in Make.
`no-character-count'
`-w' in shar. =
`no-check-existing'
`-x' in shar. =
`no-common'
`-3' in wdiff. =
`no-create'
`-c' in touch. =
`no-defines'
`-D' in etags. =
`no-deleted'
`-1' in wdiff. =
`no-dereference'
`-d' in cp. =
`no-inserted'
`-2' in wdiff. =
`no-keep-going'
`-S' in Make.
`no-lines'
`-l' in Bison.
`no-piping'
`-P' in shar. =
`no-prof'
`-e' in gprof. =
`no-regex'
`-R' in etags. =
`no-sort'
`-p' in nm. =
`no-splash'
Don't print a startup splash screen.
`no-split'
Used in makeinfo.
`no-static'
`-a' in gprof. =
`no-time'
`-E' in gprof. =
`no-timestamp'
`-m' in shar. =
`no-validate'
Used in makeinfo.
`no-wait'
Used in emacsclient.
`no-warn'
Used in various programs to inhibit warnings.
`node'
`-n' in info. =
`nodename'
`-n' in uname. =
`nonmatching'
`-f' in cpio. =
`nstuff'
`-n' in = objdump.
`null'
`-0' in xargs. =
`number'
`-n' in cat. =
`number-nonblank'
`-b' in cat. =
`numeric-sort'
`-n' in nm. =
`numeric-uid-gid'
`-n' in cpio = and=20 ls.
`nx'
Used in GDB.
`old-archive'
`-o' in tar. =
`old-file'
`-o' in Make.
`one-file-system'
`-l' in tar,=20 cp, and du.
`only-file'
`-o' in ptx. =
`only-prof'
`-f' in gprof. =
`only-time'
`-F' in gprof. =
`options'
`-o' in = getopt,=20 fdlist, fdmount, fdmountd, and=20 fdumount.
`output'
In various programs, specify the output file name.
`output-prefix'
`-o' in shar. =
`override'
`-o' in rm. =
`overwrite'
`-c' in = unshar.
`owner'
`-o' in = install.
`paginate'
`-l' in diff. =
`paragraph-indent'
Used in makeinfo.
`parents'
`-p' in mkdir = and=20 rmdir.
`pass-all'
`-p' in ul. =
`pass-through'
`-p' in cpio. =
`port'
`-P' in = finger.
`portability'
`-c' in cpio = and=20 tar.
`posix'
Used in gawk.
`prefix-builtins'
`-P' in m4. =
`prefix'
`-f' in = csplit.
`preserve'
Used in tar and cp.
`preserve-environment'
`-p' in su. =
`preserve-modification-time'
`-m' in cpio. =
`preserve-order'
`-s' in tar. =
`preserve-permissions'
`-p' in tar. =
`print'
`-l' in diff. =
`print-chars'
`-L' in cmp. =
`print-data-base'
`-p' in Make.
`print-directory'
`-w' in Make.
`print-file-name'
`-o' in nm. =
`print-symdefs'
`-s' in nm. =
`printer'
`-p' in wdiff. =
`prompt'
`-p' in ed. =
`proxy'
Specify an HTTP proxy.
`query-user'
`-X' in shar. =
`question'
`-q' in Make.
`quiet'
Used in many programs to inhibit the usual output. Every program = accepting=20 `--quiet' should accept = `--silent' as a synonym.
`quiet-unshar'
`-Q' in shar =
`quote-name'
`-Q' in ls. =
`rcs'
`-n' in diff. =
`re-interval'
Used in gawk.
`read-full-blocks'
`-B' in tar. =
`readnow'
Used in GDB.
`recon'
`-n' in Make.
`record-number'
`-R' in tar. =
`recursive'
Used in chgrp, chown, cp,=20 ls, diff, and rm.
`reference-limit'
Used in makeinfo.
`references'
`-r' in ptx. =
`regex'
`-r' in tac = and=20 etags.
`release'
`-r' in uname. =
`reload-state'
`-R' in m4. =
`relocation'
`-r' in = objdump.
`rename'
`-r' in cpio. =
`replace'
`-i' in xargs. =
`report-identical-files'
`-s' in diff. =
`reset-access-time'
`-a' in cpio. =
`reverse'
`-r' in ls and = nm.
`reversed-ed'
`-f' in diff. =
`right-side-defs'
`-R' in ptx. =
`same-order'
`-s' in tar. =
`same-permissions'
`-p' in tar. =
`save'
`-g' in stty. =
`se'
Used in GDB.
`sentence-regexp'
`-S' in ptx. =
`separate-dirs'
`-S' in du. =
`separator'
`-s' in tac. =
`sequence'
Used by recode to chose files or pipes for sequencing = passes.=20
`shell'
`-s' in su. =
`show-all'
`-A' in cat. =
`show-c-function'
`-p' in diff. =
`show-ends'
`-E' in cat. =
`show-function-line'
`-F' in diff. =
`show-tabs'
`-T' in cat. =
`silent'
Used in many programs to inhibit the usual output. Every program = accepting=20 `--silent' should accept = `--quiet' as a synonym.
`size'
`-s' in ls. =
`socket'
Specify a file descriptor for a network server to use for its = socket,=20 instead of opening and binding a new socket. This provides a way to = run, in a=20 non-privileged process, a server that normally needs a reserved port = number.=20
`sort'
Used in ls.
`source'
`-W source' in = gawk.=20
`sparse'
`-S' in tar. =
`speed-large-files'
`-H' in diff. =
`split-at'
`-E' in = unshar.
`split-size-limit'
`-L' in shar. =
`squeeze-blank'
`-s' in cat. =
`start-delete'
`-w' in wdiff. =
`start-insert'
`-y' in wdiff. =
`starting-file'
Used in tar and diff to specify which = file=20 within a directory to start processing with.
`statistics'
`-s' in wdiff. =
`stdin-file-list'
`-S' in shar. =
`stop'
`-S' in Make.
`strict'
`-s' in = recode.
`strip'
`-s' in = install.
`strip-all'
`-s' in strip. =
`strip-debug'
`-S' in strip. =
`submitter'
`-s' in shar. =
`suffix'
`-S' in cp,=20 ln, mv.
`suffix-format'
`-b' in = csplit.
`sum'
`-s' in gprof. =
`summarize'
`-s' in du. =
`symbolic'
`-s' in ln. =
`symbols'
Used in GDB and objdump.
`synclines'
`-s' in m4. =
`sysname'
`-s' in uname. =
`tabs'
`-t' in expand = and=20 unexpand.
`tabsize'
`-T' in ls. =
`terminal'
`-T' in tput = and=20 ul. `-t' in=20 wdiff.
`text'
`-a' in diff. =
`text-files'
`-T' in shar. =
`time'
Used in ls and touch.
`timeout'
Specify how long to wait before giving up on some operation.
`to-stdout'
`-O' in tar. =
`total'
`-c' in du. =
`touch'
`-t' in Make, = ranlib,=20 and recode.
`trace'
`-t' in m4. =
`traditional'
`-t' in hello; = `-W traditional' in = gawk;=20 `-G' in ed, = m4,=20 and ptx.
`tty'
Used in GDB.
`typedefs'
`-t' in ctags. =
`typedefs-and-c++'
`-T' in ctags. =
`typeset-mode'
`-t' in ptx. =
`uncompress'
`-z' in tar. =
`unconditional'
`-u' in cpio. =
`undefine'
`-U' in m4. =
`undefined-only'
`-u' in nm. =
`update'
`-u' in cp,=20 ctags, mv, tar.
`usage'
Used in gawk; same as `--help'.
`uuencode'
`-B' in shar. =
`vanilla-operation'
`-V' in shar. =
`verbose'
Print more information about progress. Many programs support this. =
`verify'
`-W' in tar. =
`version'
Print the version number.
`version-control'
`-V' in cp,=20 ln, mv.
`vgrind'
`-v' in ctags. =
`volume'
`-V' in tar. =
`what-if'
`-W' in Make.
`whole-size-limit'
`-l' in shar. =
`width'
`-w' in ls and = ptx.
`word-regexp'
`-W' in ptx. =
`writable'
`-T' in who. =
`zeros'
`-z' in gprof. =

Next: File=20 Usage, Previous: Option Table, Up: Program Behavior

4.9 Memory Usage

If a program typically uses just = a few meg=20 of memory, don't bother making any effort to reduce memory usage. For = example,=20 if it is impractical for other reasons to operate on files more than a = few meg=20 long, it is reasonable to read entire input files into memory to operate = on=20 them.=20

However, for programs such as cat or tail, = that can=20 usefully operate on very large files, it is important to avoid using a = technique=20 that would artificially limit the size of files it can handle. If a = program=20 works by lines and could be applied to arbitrary user-supplied input = files, it=20 should keep only a line in memory, because this is not very hard and = users will=20 want to be able to operate on input files that are bigger than will fit = in=20 memory all at once.=20

If your program creates complicated data structures, just make them = in memory=20 and give a fatal error if malloc returns zero.=20

Previous: Memory Usage, Up: Program Behavior

4.10 File Usage

Programs should be prepared to = operate when=20 /usr and /etc are read-only file systems. Thus, if the = program=20 manages log files, lock files, backup files, score files, or any other = files=20 which are modified for internal purposes, these files should not be = stored in=20 /usr or /etc.=20

There are two exceptions. /etc = is used=20 to store system configuration information; it is reasonable for a = program to=20 modify files in /etc when its job = is to=20 update the system configuration. Also, if the user explicitly asks to = modify one=20 file in a directory, it is reasonable for the program to store other = files in=20 the same directory.=20

Next: Documentation, Previous: Program Behavior, Up: Top=20

5 Making The Best Use of C

This chapter provides advice on how best to use the C language when = writing=20 GNU software.=20

Next: Comments, Up: Writing=20 C

5.1 Formatting Your Source Code

It is=20 important to put the open-brace that starts the body of a C function in = column=20 one, so that they will start a defun. Several tools look for open-braces = in=20 column one to find the beginnings of C functions. These tools will not = work on=20 code not formatted that way.=20

Avoid putting open-brace, open-parenthesis or open-bracket in column = one when=20 they are inside a function, so that they won't start a defun. The = open-brace=20 that starts a struct body can go in column one if you find = it=20 useful to treat that definition as a defun.=20

It is also important for function definitions to start the name of = the=20 function in column one. This helps people to search for function = definitions,=20 and may also help certain tools recognize them. Thus, using Standard C = syntax,=20 the format is this: 

     static char *
     concat (char *s1, char *s2)
     {
       ...
     }

or, if you want to use traditional C syntax, format = the=20 definition like this: 

     static char *
     concat (s1, s2)        /* Name starts in column one here */
          char *s1, *s2;
     {                     /* Open brace in column one here */
       ...
     }

In Standard C, if the arguments don't fit nicely on one line, split = it like=20 this: 

     int
     lots_of_args (int an_integer, long a_long, short a_short,
                   double a_double, float a_float)
     ...

The rest of this section gives our recommendations for other aspects = of C=20 formatting style, which is also the default style of the = indent=20 program in version 1.2 and newer. It corresponds to the options 

     -nbad -bap -nbc -bbo -bl -bli2 -bls -ncdb -nce =
-cp1 -cs -di2
     -ndj -nfc1 -nfca -hnl -i2 -ip5 -lp -pcs -psl -nsc -nsob

We don't think of these recommendations as requirements, because it = causes no=20 problems for users if two different programs have different formatting = styles.=20

But whatever style you use, please use it consistently, since a = mixture of=20 styles within one program tends to look ugly. If you are contributing = changes to=20 an existing program, please follow the style of that program.=20

For the body of the function, our recommended style looks like this: = 

     if (x < foo (y, z))
       haha =3D bar[4] + 5;
     else
       {
         while (z)
           {
             haha +=3D foo (z, z);
             z--;
           }
         return ++x + bar ();
       }

We find it easier = to read a=20 program when it has spaces before the open-parentheses and after the = commas.=20 Especially after the commas.=20

When you split an expression into multiple lines, split it before an=20 operator, not after one. Here is the right way:=20 

     if (foo_this_is_long && bar > win (x, y, =
z)
         && remaining_condition)

Try to avoid having two operators of different precedence at the same = level=20 of indentation. For example, don't write this: 

     =
mode =3D (inmode[j] =3D=3D VOIDmode
             || GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE =
(inmode[j])
             ? outmode[j] : inmode[j]);

Instead, use extra parentheses so that the indentation shows the = nesting: 

     mode =3D ((inmode[j] =3D=3D VOIDmode
              || (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE =
(inmode[j])))
             ? outmode[j] : inmode[j]);

Insert extra parentheses so that Emacs will indent the code properly. = For=20 example, the following indentation looks nice if you do it by hand, 

     v =3D rup->ru_utime.tv_sec*1000 + =
rup->ru_utime.tv_usec/1000
         + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000;

but Emacs would alter it. Adding a set of = parentheses produces=20 something that looks equally nice, and which Emacs will preserve: 

     v =3D (rup->ru_utime.tv_sec*1000 + =
rup->ru_utime.tv_usec/1000
          + rup->ru_stime.tv_sec*1000 + =
rup->ru_stime.tv_usec/1000);

Format do-while statements like this: 

     do
       {
         a =3D foo (a);
       }
     while (a > 0);

Please use=20 formfeed characters (control-L) to divide the program into pages at = logical=20 places (but not within a function). It does not matter just how long the = pages=20 are, since they do not have to fit on a printed page. The formfeeds = should=20 appear alone on lines by themselves.=20

Next: Syntactic Conventions, Previous: Formatting, Up: Writing=20 C

5.2 Commenting Your Work

Every program should start with a = comment=20 saying briefly what it is for. Example: `fmt - = filter for=20 simple filling of text'. This comment should be at the top = of the=20 source file containing the `main' = function=20 of the program.=20

Also, please write a brief comment at the start of each source file, = with the=20 file name and a line or two about the overall purpose of the file.=20

Please write the comments in a GNU program in English, because = English is the=20 one language that nearly all programmers in all countries can read. If = you do=20 not write English well, please write comments in English as well as you = can,=20 then ask other people to help rewrite them. If you can't write comments = in=20 English, please find someone to work with you and translate your = comments into=20 English.=20

Please put a comment on each function saying what the function does, = what=20 sorts of arguments it gets, and what the possible values of arguments = mean and=20 are used for. It is not necessary to duplicate in words the meaning of = the C=20 argument declarations, if a C type is being used in its customary = fashion. If=20 there is anything nonstandard about its use (such as an argument of type = char * which is really the address of the second character = of a=20 string, not the first), or any possible values that would not work the = way one=20 would expect (such as, that strings containing newlines are not = guaranteed to=20 work), be sure to say so.=20

Also explain the significance of the return value, if there is one.=20

Please put two spaces after the end of a sentence in your comments, = so that=20 the Emacs sentence commands will work. Also, please write complete = sentences and=20 capitalize the first word. If a lower-case identifier comes at the = beginning of=20 a sentence, don't capitalize it! Changing the spelling makes it a = different=20 identifier. If you don't like starting a sentence with a lower case = letter,=20 write the sentence differently (e.g., =93The identifier lower-case is = ...=94).=20

The comment on a function is much clearer if you use the argument = names to=20 speak about the argument values. The variable name itself should be = lower case,=20 but write it in upper case when you are speaking about the value rather = than the=20 variable itself. Thus, =93the inode number NODE_NUM=94 rather than =93an = inode=94.=20

There is usually no purpose in restating the name of the function in = the=20 comment before it, because the reader can see that for himself. There = might be=20 an exception when the comment is so long that the function itself would = be off=20 the bottom of the screen.=20

There should be a comment on each static variable as well, like this: = 

     /* Nonzero means truncate lines in the =
display;
        zero means continue them.  */
     int truncate_lines;

Ever= y=20 `#endif' should have a comment, = except in=20 the case of short conditionals (just a few lines) that are not nested. = The=20 comment should state the condition of the conditional that is ending,=20 including its sense. `#else'=20 should have a comment describing the condition and sense of the = code=20 that follows. For example: 

     #ifdef foo
       ...
     #else /* not foo */
       ...
     #endif /* not foo */
     #ifdef foo
       ...
     #endif /* foo */

but, by contrast, write the comments this way for a=20 `#ifndef': 

  =
   #ifndef foo
       ...
     #else /* foo */
       ...
     #endif /* foo */
     #ifndef foo
       ...
     #endif /* not foo */

Next: Names, Previous: Comments, Up: Writing=20 C

5.3 Clean Use of C Constructs

Please explicitly = declare the=20 types of all objects. For example, you should explicitly declare all = arguments=20 to functions, and you should declare functions to return = int rather=20 than omitting the int.=20

Some = programmers like to use the GCC `-Wall'=20 option, and change the code whenever it issues a warning. If you want to = do=20 this, then do. Other programmers prefer not to use `-Wall', because it gives warnings for valid = and=20 legitimate code which they do not want to change. If you want to do = this, then=20 do. The compiler should be your servant, not your master.=20

Declarations of external functions and functions to appear later in = the=20 source file should all go in one place near the beginning of the file = (somewhere=20 before the first function definition in the file), or else should go in = a header=20 file. Don't put extern declarations inside functions.=20

It used to be common = practice to use=20 the same local variables (with names like tem) over and = over for=20 different values within one function. Instead of doing this, it is = better to=20 declare a separate local variable for each distinct purpose, and give it = a name=20 which is meaningful. This not only makes programs easier to understand, = it also=20 facilitates optimization by good compilers. You can also move the = declaration of=20 each local variable into the smallest scope that includes all its uses. = This=20 makes the program even cleaner.=20

Don't use local variables or parameters that shadow global = identifiers.=20

Don't declare = multiple=20 variables in one declaration that spans lines. Start a new declaration = on each=20 line, instead. For example, instead of this: 

     =
int    foo,
            bar;

write either this: 

     int =
foo, bar;

or this: 

     int foo;
     int bar;

(If they are global variables, each should have a = comment=20 preceding it anyway.)=20

When you have an if-else statement nested = in=20 another if statement, always put braces around the=20 if-else. Thus, never write like this: 

     if (foo)
       if (bar)
         win ();
       else
         lose ();

always like this: 

     if (foo)
       {
         if (bar)
           win ();
         else
           lose ();
       }

If you have an if statement nested inside of an=20 else statement, either write else if on one = line, like=20 this, 

     if (foo)
       ...
     else if (bar)
       ...

with its then-part indented like the = preceding=20 then-part, or write the nested if within = braces like=20 this: 

     if (foo)
       ...
     else
       {
         if (bar)
           ...
       }

Don't declare both a structure tag and variables or typedefs in the = same=20 declaration. Instead, declare the structure tag separately and then use = it to=20 declare the variables or typedefs.=20

Try to avoid assignments inside if-conditions = (assignments=20 inside while-conditions are ok). For example, don't write = this: 

     if ((foo =3D (char *) malloc (sizeof =
*foo)) =3D=3D 0)
       fatal ("virtual memory exhausted");

instead, write this: 

     foo =
=3D (char *) malloc (sizeof *foo);
     if (foo =3D=3D 0)
       fatal ("virtual memory exhausted");

Don't make the program ugly to placate=20 lint. Please don't insert any casts to void. = Zero=20 without a cast is perfectly fine as a null pointer constant, except when = calling=20 a varargs function.=20

Next: System Portability, Previous: Syntactic Conventions, Up: Writing=20 C

5.4 Naming Variables, Functions, and Files

The = names of global variables and functions in a program serve as comments = of a=20 sort. So don't choose terse names=97instead, look for names that give = useful=20 information about the meaning of the variable or function. In a GNU = program,=20 names should be English, like other comments.=20

Local variable names can be shorter, because they are used only = within one=20 context, where (presumably) comments explain their purpose.=20

Try to limit your use of abbreviations in symbol names. It is ok to = make a=20 few abbreviations, explain what they mean, and then use them frequently, = but=20 don't use lots of obscure abbreviations.=20

Please use underscores to separate words in a name, so that the Emacs = word=20 commands can be useful within them. Stick to lower case; reserve upper = case for=20 macros and enum constants, and for name-prefixes that = follow a=20 uniform convention.=20

For example, you should use names like = ignore_space_change_flag;=20 don't use names like iCantReadThis.=20

Variables that indicate whether command-line options have been = specified=20 should be named after the meaning of the option, not after the = option-letter. A=20 comment should state both the exact meaning of the option and its = letter. For=20 example, 

     /* Ignore changes in horizontal =
whitespace (-b).  */
     int ignore_space_change_flag;

When you want to define names with constant integer values, use=20 enum rather than `#define'.=20 GDB knows about enumeration constants.=20

You might want to make sure that none of the = file names=20 would conflict if the files were loaded onto an MS-DOS file system which = shortens the names. You can use the program doschk to test = for=20 this.=20

Some GNU programs were designed to limit themselves to file names of = 14=20 characters or less, to avoid file name conflicts if they are read into = older=20 System V systems. Please preserve this feature in the existing GNU = programs that=20 have it, but there is no need to do this in new GNU programs.=20 doschk also reports file names longer than 14 characters.=20

Next: CPU Portability, Previous: Names, Up: Writing=20 C

5.5 Portability between System Types

In the = Unix world,=20 =93portability=94 refers to porting to different Unix versions. For a = GNU program,=20 this kind of portability is desirable, but not paramount.=20

The primary purpose of GNU software is to run on top of the GNU = kernel,=20 compiled with the GNU C compiler, on various types of cpu.=20 So the kinds of portability that are absolutely necessary are quite = limited. But=20 it is important to support Linux-based GNU systems, since they are the = form of=20 GNU that is popular.=20

Beyond that, it is good to support the other free operating systems = (*BSD),=20 and it is nice to support other Unix-like systems if you want to. = Supporting a=20 variety of Unix-like systems is desirable, although not paramount. It is = usually=20 not too hard, so you may as well do it. But you don't have to consider = it an=20 obligation, if it does turn out to be hard.=20

The easiest way to achieve = portability to most=20 Unix-like systems is to use Autoconf. It's unlikely that your program = needs to=20 know more information about the host platform than Autoconf can provide, = simply=20 because most of the programs that need such knowledge have already been = written.=20

Avoid using the format of semi-internal data bases (e.g., = directories) when=20 there is a higher-level alternative (readdir).=20

As=20 for systems that are not like Unix, such as MSDOS, Windows, VMS, MVS, = and older=20 Macintosh systems, supporting them is often a lot of work. When that is = the=20 case, it is better to spend your time adding features that will be = useful on GNU=20 and GNU/Linux, rather than on supporting other incompatible systems.=20

If you do support Windows, please do not abbreviate it as =93win=94. = In hacker=20 terminology, calling something a =93win=94 is a form of praise. You're = free to=20 praise Microsoft Windows on your own if you want, but please don't do = this in=20 GNU packages. Instead of abbreviating =93Windows=94 to =93un=94, you can = write it in=20 full or abbreviate it to =93woe=94 or =93w=94. In GNU Emacs, for = instance, we use=20 `w32' in file names of = Windows-specific=20 files, but the macro for Windows conditionals is called = WINDOWSNT.=20

It is a good idea to define the =93feature test macro=94 = _GNU_SOURCE=20 when compiling your C files. When you compile on GNU or GNU/Linux, this = will=20 enable the declarations of GNU library extension functions, and that = will=20 usually give you a compiler error message if you define the same = function names=20 in some other way in your program. (You don't have to actually = use=20 these functions, if you prefer to make the program more portable to = other=20 systems.)=20

But whether or not you use these GNU extensions, you should avoid = using their=20 names for any other meanings. Doing so would make it hard to move your = code into=20 other GNU programs.=20

Next: System Functions, Previous: System Portability, Up: Writing=20 C

5.6 Portability between cpus

Even GNU systems = will differ=20 because of differences among cpu types=97for = example,=20 difference in byte ordering and alignment requirements. It is absolutely = essential to handle these differences. However, don't make any effort to = cater=20 to the possibility that an int will be less than 32 bits. = We don't=20 support 16-bit machines in GNU.=20

Similarly, don't make any effort to cater to the possibility that=20 long will be smaller than predefined types like=20 size_t. For example, the following code is ok: 

     printf ("size =3D %lu\n", (unsigned long) sizeof =
array);
     printf ("diff =3D %ld\n", (long) (pointer2 - pointer1));

1989 Standard C requires this to work, and we know of only one=20 counterexample: 64-bit programs on Microsoft Windows. We will leave it = to those=20 who want to port GNU programs to that environment to figure out how to = do it.=20

Predefined file-size types like off_t are an exception: = they are=20 longer than long on many platforms, so code like the above = won't=20 work with them. One way to print an off_t value portably is = to=20 print its digits yourself, one by one.=20

Don't assume that the address of an int object is also = the=20 address of its least-significant byte. This is false on big-endian = machines.=20 Thus, don't make the following mistake: 

     int c;
     ...
     while ((c =3D getchar ()) !=3D EOF)
       write (file_descriptor, &c, 1);

Instead, use unsigned char as follows. = (The=20 unsigned is for portability to unusual systems where=20 char is signed and where there is integer overflow = checking.) 

     int c;
     while ((c =3D getchar ()) !=3D EOF)
       {
         unsigned char u =3D c;
         write (file_descriptor, &u, 1);
       }

It used to be ok to not worry about the difference between pointers = and=20 integers when passing arguments to functions. However, on most modern = 64-bit=20 machines pointers are wider than int. Conversely, integer = types=20 like long long int and off_t are wider than = pointers=20 on most modern 32-bit machines. Hence it's often better nowadays to use=20 prototypes to define functions whose argument types are not trivial.=20

In particular, if functions accept varying argument counts or types = they=20 should be declared using prototypes containing `...' and defined using stdarg.h. For an example of this, please see = the Gnulib error module, = which=20 declares and defines the following function: 

     =
/* Print a message with `fprintf (stderr, FORMAT, ...)';
        if ERRNUM is nonzero, follow it with ": " and strerror (ERRNUM).
        If STATUS is nonzero, terminate the program with `exit =
(STATUS)'.  */
    =20
     void error (int status, int errnum, const char *format, ...);

A simple way to use the Gnulib error module is to obtain the two = source files=20 error.c and error.h from the Gnulib library source code = repository=20 at http:= //savannah.gnu.org/cgi-bin/viewcvs/gnulib/gnulib/lib/.=20 Here's a sample use: 

     #include "error.h"
     #include <errno.h>
     #include <stdio.h>
    =20
     char *program_name =3D "myprogram";
    =20
     FILE *
     xfopen (char const *name)
     {
       FILE *fp =3D fopen (name, "r");
       if (! fp)
         error (1, errno, "cannot read %s", name);
       return fp;
     }

Avoid casting = pointers to=20 integers if you can. Such casts greatly reduce portability, and in most = programs=20 they are easy to avoid. In the cases where casting pointers to integers = is=20 essential=97such as, a Lisp interpreter which stores type information as = well as=20 an address in one word=97you'll have to make explicit provisions to = handle=20 different word sizes. You will also need to make provision for systems = in which=20 the normal range of addresses you can get from malloc = starts far=20 away from zero.=20

Next: Internationalization, Previous: CPU Portability, Up: Writing=20 C

5.7 Calling System Functions

C = implementations=20 differ substantially. Standard C reduces but does not eliminate the=20 incompatibilities; meanwhile, many GNU packages still support = pre-standard=20 compilers because this is not hard to do. This chapter gives = recommendations for=20 how to use the more-or-less standard C library functions to avoid = unnecessary=20 loss of portability.=20

Here we assume that HAVE_STRCHR and = HAVE_STRRCHR=20 are macros defined in systems where the corresponding functions exist. = One way=20 to get them properly defined is to use Autoconf.=20

Next: Character Set, Previous: System Functions, Up: Writing=20 C

5.8 Internationalization

GNU=20 has a library called GNU gettext that makes it easy to translate the = messages in=20 a program into various languages. You should use this library in every = program.=20 Use English for the messages as they appear in the program, and let = gettext=20 provide the way to translate them into other languages.=20

Using GNU gettext involves putting a call to the gettext = macro=20 around each string that might need translation=97like this: 

     printf (gettext ("Processing file `%s'..."));

This permits GNU gettext to replace the string=20 "Processing file `%s'..." with a translated version.=20

Once a program uses gettext, please make a point of writing calls to=20 gettext when you add new strings that call for translation. =

Using GNU gettext in a package involves specifying a text domain = name for the package. The text domain name is used to separate the = translations for this package from the translations for other packages.=20 Normally, the text domain name should be the same as the name of the = package=97for=20 example, `coreutils' for the GNU = core=20 utilities.=20

To = enable=20 gettext to work well, avoid writing code that makes assumptions about = the=20 structure of words or sentences. When you want the precise text of a = sentence to=20 vary depending on the data, use two or more alternative string constants = each=20 containing a complete sentences, rather than inserting conditionalized = words or=20 phrases into a single sentence framework.=20

Here is an example of what not to do: 

     =
printf ("%s is full", capacity > 5000000 ? "disk" : "floppy disk");

If you apply gettext to all strings, like this, 

     printf (gettext ("%s is full"),
             capacity > 5000000 ? gettext ("disk") : gettext ("floppy =
disk"));

the translator will hardly know that "disk" and = "floppy disk"=20 are meant to be substituted in the other string. Worse, in some = languages (like=20 French) the construction will not work: the translation of the word = "full"=20 depends on the gender of the first part of the sentence; it happens to = be not=20 the same for "disk" as for "floppy disk".=20

Complete sentences can be translated without problems: 

     printf (capacity > 5000000 ? gettext ("disk is =
full")
             : gettext ("floppy disk is full"));

A similar problem appears at the level of sentence structure with = this code: 

     printf ("#  Implicit rule search =
has%s been done.\n",
             f->tried_implicit ? "" : " not");

Adding gettext calls to this code = cannot give=20 correct results for all languages, because negation in some languages = requires=20 adding words at more than one place in the sentence. By contrast, adding = gettext calls does the job straightforwardly if the code = starts out=20 like this: 

     printf (f->tried_implicit
             ? "#  Implicit rule search has been done.\n",
             : "#  Implicit rule search has not been done.\n");

Another example is this one: 

     printf ("%d =
file%s processed", nfiles,
             nfiles !=3D 1 ? "s" : "");

The problem with this example is that it assumes = that plurals=20 are made by adding `s'. If you apply gettext to the format string, like = this, 

     printf (gettext ("%d file%s processed"), =
nfiles,
             nfiles !=3D 1 ? "s" : "");

the message can use different words, but it will = still be=20 forced to use `s' for the plural. Here is a better way, with gettext = being=20 applied to the two strings independently: 

     =
printf ((nfiles !=3D 1 ? gettext ("%d files processed")
              : gettext ("%d file processed")),
             nfiles);

But this still doesn't work for languages like = Polish, which=20 has three plural forms: one for nfiles =3D=3D 1, one for nfiles =3D=3D = 2, 3, 4, 22, 23,=20 24, ... and one for the rest. The GNU ngettext function = solves this=20 problem: 

     printf (ngettext ("%d files =
processed", "%d file processed", nfiles),
             nfiles);

Next: Quote Characters, Previous: Internationalization, Up: Writing=20 C

5.9 Character Set

Sticking to the ASCII = character set=20 (plain text, 7-bit characters) is preferred in GNU source code comments, = text=20 documents, and other contexts, unless there is good reason to do = something else=20 because of the application domain. For example, if source code deals = with the=20 French Revolutionary calendar, it is OK if its literal strings contain = accented=20 characters in month names like =93Flor=E9al=94. Also, it is OK to use = non-ASCII=20 characters to represent proper names of contributors in change logs (see = Cha= nge=20 Logs).=20

If you need to use non-ASCII characters, you should normally stick = with one=20 encoding, as one cannot in general mix encodings reliably.=20

Next: Mmap,=20 Previous: Character Set, Up: Writing=20 C

5.10 Quote Characters

In = the C locale,=20 GNU programs should stick to plain ASCII for quotation characters in = messages to=20 users: preferably 0x60 (``') for = left=20 quotes and 0x27 (`'') for right = quotes. It=20 is ok, but not required, to use locale-specific quotes in other locales. =

The Gnulib=20 quote and quotearg modules provide a = reasonably=20 straightforward way to support locale-specific quote characters, as well = as=20 taking care of other issues, such as quoting a filename that itself = contains a=20 quote character. See the Gnulib documentation for usage details.=20

In any case, the documentation for your program should clearly = specify how it=20 does quoting, if different than the preferred method of ``' and `''. This=20 is especially important if the output of your program is ever likely to = be=20 parsed by another program.=20

Quotation characters are a difficult area in the computing world at = this=20 time: there are no true left or right quote characters in Latin1; the=20 ``' character we use was = standardized there=20 as a grave accent. Moreover, Latin1 is still not universally usable.=20

Unicode contains the unambiguous quote characters required, and its = common=20 encoding UTF-8 is upward compatible with Latin1. However, Unicode and = UTF-8 are=20 not universally well-supported, either.=20

This may change over the next few years, and then we will revisit = this.=20

Previous: Quote Characters, Up: Writing=20 C

5.11 Mmap

Don't assume that mmap = either works on=20 all files or fails for all files. It may work on some files and fail on = others.=20

The proper way to use mmap is to try it on the specific = file for=20 which you want to use it=97and if mmap doesn't work, fall = back on=20 doing the job in another way using read and = write.=20

The reason this precaution is needed is that the GNU kernel (the = HURD)=20 provides a user-extensible file system, in which there can be many = different=20 kinds of =93ordinary files.=94 Many of them support mmap, = but some do=20 not. It is important to make programs handle all these kinds of files.=20

Next: Managing Releases, Previous: Writing C, Up: Top=20

6 Documenting Programs

A GNU program should ideally = come with=20 full free documentation, adequate for both reference and tutorial = purposes. If=20 the package can be programmed or extended, the documentation should = cover=20 programming or extending it, as well as just using it.=20

Next: Doc Strings and Manuals, Up: Documentation

6.1 GNU Manuals

The preferred document format for the GNU system is the Texinfo = formatting=20 language. Every GNU package should (ideally) have documentation in = Texinfo both=20 for reference and for learners. Texinfo makes it possible to produce a = good=20 quality formatted book, using TeX, and to generate an Info file. It is = also=20 possible to generate HTML output from Texinfo source. See the Texinfo = manual,=20 either the hardcopy, or the on-line version available through = info=20 or the Emacs Info subsystem (C-h i).=20

Nowadays some other formats such as Docbook and Sgmltexi can be = converted=20 automatically into Texinfo. It is ok to produce the Texinfo = documentation by=20 conversion this way, as long as it gives good results.=20

Make sure your manual is clear to a reader who knows nothing about = the topic=20 and reads it straight through. This means covering basic topics at the=20 beginning, and advanced topics only later. This also means defining = every=20 specialized term when it is first used.=20

Programmers tend to carry over the structure of the program as the = structure=20 for its documentation. But this structure is not necessarily good for = explaining=20 how to use the program; it may be irrelevant and confusing for a user.=20

Instead, the right way to structure documentation is according to the = concepts and questions that a user will have in mind when reading it. = This=20 principle applies at every level, from the lowest (ordering sentences in = a=20 paragraph) to the highest (ordering of chapter topics within the = manual).=20 Sometimes this structure of ideas matches the structure of the = implementation of=20 the software being documented=97but often they are different. An = important part of=20 learning to write good documentation is to learn to notice when you have = unthinkingly structured the documentation like the implementation, stop=20 yourself, and look for better alternatives.=20

For example, each program in the GNU system probably ought to be = documented=20 in one manual; but this does not mean each program should have its own = manual.=20 That would be following the structure of the implementation, rather than = the=20 structure that helps the user understand.=20

Instead, each manual should cover a coherent topic. For = example,=20 instead of a manual for diff and a manual for = diff3,=20 we have one manual for =93comparison of files=94 which covers both of = those=20 programs, as well as cmp. By documenting these programs = together,=20 we can make the whole subject clearer.=20

The manual which discusses a program should certainly document all of = the=20 program's command-line options and all of its commands. It should give = examples=20 of their use. But don't organize the manual as a list of features. = Instead,=20 organize it logically, by subtopics. Address the questions that a user = will ask=20 when thinking about the job that the program does. Don't just tell the = reader=20 what each feature can do=97say what jobs it is good for, and show how to = use it=20 for those jobs. Explain what is recommended usage, and what kinds of = usage users=20 should avoid.=20

In general, a GNU manual should serve both as tutorial and reference. = It=20 should be set up for convenient access to each topic through Info, and = for=20 reading straight through (appendixes aside). A GNU manual should give a = good=20 introduction to a beginner reading through from the start, and should = also=20 provide all the details that hackers want. The Bison manual is a good = example of=20 this=97please take a look at it to see what we mean.=20

That is not as hard as it first sounds. Arrange each chapter as a = logical=20 breakdown of its topic, but order the sections, and write their text, so = that=20 reading the chapter straight through makes sense. Do likewise when = structuring=20 the book into chapters, and when structuring a section into paragraphs. = The=20 watchword is, at each point, address the most fundamental and = important=20 issue raised by the preceding text.=20

If necessary, add extra chapters at the beginning of the manual which = are=20 purely tutorial and cover the basics of the subject. These provide the = framework=20 for a beginner to understand the rest of the manual. The Bison manual = provides a=20 good example of how to do this.=20

To serve as a reference, a manual should have an Index that list all = the=20 functions, variables, options, and important concepts that are part of = the=20 program. One combined Index should do for a short manual, but sometimes = for a=20 complex package it is better to use multiple indices. The Texinfo manual = includes advice on preparing good index entries, see Mak= ing Index=20 Entries, and see Defining=20 the Entries of an Index.=20

Don't use Unix man pages as a model for how to write GNU = documentation; most=20 of them are terse, badly structured, and give inadequate explanation of = the=20 underlying concepts. (There are, of course, some exceptions.) Also, Unix = man=20 pages use a particular format which is different from what we use in GNU = manuals.=20

Please include an email address in the manual for where to report = bugs in=20 the text of the manual.=20

Please do not use the term =93pathname=94 that is used in Unix = documentation; use=20 =93file name=94 (two words) instead. We use the term =93path=94 only for = search paths,=20 which are lists of directory names.=20

Please do not use the term =93illegal=94 to refer to erroneous input = to a=20 computer program. Please use =93invalid=94 for this, and reserve the = term =93illegal=94=20 for activities prohibited by law.=20

Please do not write `()' after = a=20 function name just to indicate it is a function. foo () is = not a=20 function, it is a function call with no arguments.=20

Next: Manual Structure Details, Previous: GNU Manuals, Up: Documentation

6.2 Doc Strings and Manuals

Some programming systems, such as Emacs, provide a documentation = string for=20 each function, command or variable. You may be tempted to write a = reference=20 manual by compiling the documentation strings and writing a little = additional=20 text to go around them=97but you must not do it. That approach is a = fundamental=20 mistake. The text of well-written documentation strings will be entirely = wrong=20 for a manual.=20

A documentation string needs to stand alone=97when it appears on the = screen,=20 there will be no other text to introduce or explain it. Meanwhile, it = can be=20 rather informal in style.=20

The text describing a function or variable in a manual must not stand = alone;=20 it appears in the context of a section or subsection. Other text at the=20 beginning of the section should explain some of the concepts, and should = often=20 make some general points that apply to several functions or variables. = The=20 previous descriptions of functions and variables in the section will = also have=20 given information about the topic. A description written to stand alone = would=20 repeat some of that information; this redundancy looks bad. Meanwhile, = the=20 informality that is acceptable in a documentation string is totally = unacceptable=20 in a manual.=20

The only good way to use documentation strings in writing a good = manual is to=20 use them as a source of information for writing good text.=20

Next: License for Manuals, Previous: Doc Strings and Manuals, Up: Documentation

6.3 Manual Structure Details

The title page of the manual = should=20 state the version of the programs or packages documented in the manual. = The Top=20 node of the manual should also contain this information. If the manual = is=20 changing more frequently than or independent of the program, also state = a=20 version number for the manual in both of these places.=20

Each program documented in the manual should have a node named=20 `program Invocation' = or=20 `Invoking program'. = This node=20 (together with its subnodes, if any) should describe the program's = command line=20 arguments and how to run it (the sort of information people would look = for in a=20 man page). Start with an `@example'=20 containing a template for all the options and arguments that the program = uses.=20

Alternatively, put a menu item in some menu whose item name fits one = of the=20 above patterns. This identifies the node which that item points to as = the node=20 for this purpose, regardless of the node's actual name.=20

The `--usage' feature of the = Info reader=20 looks for such a node or menu item in order to find the relevant text, = so it is=20 essential for every Texinfo file to have one.=20

If one manual describes several programs, it should have such a node = for each=20 program described in the manual.=20

Next: Manual Credits, Previous: Manual Structure Details, Up: Documentation

6.4 License for Manuals

Please use the GNU Free=20 Documentation License for all GNU manuals that are more than a few pages = long.=20 Likewise for a collection of short documents=97you only need one copy of = the GNU=20 FDL for the whole collection. For a single short document, you can use a = very=20 permissive non-copyleft license, to avoid taking up space with a long = license.=20

See http://www.gnu.org/co= pyleft/fdl-howto.html=20 for more explanation of how to employ the GFDL.=20

Note that it is not obligatory to include a copy of the GNU GPL or = GNU LGPL=20 in a manual whose license is neither the GPL nor the LGPL. It can be a = good idea=20 to include the program's license in a large manual; in a short manual, = whose=20 size would be increased considerably by including the program's license, = it is=20 probably better not to include it.=20

Next: Printed Manuals, Previous: License for Manuals, Up: Documentation

6.5 Manual Credits

Please credit the = principal human=20 writers of the manual as the authors, on the title page of the manual. = If a=20 company sponsored the work, thank the company in a suitable place in the = manual,=20 but do not cite the company as an author.=20

Next: NEWS=20 File, Previous: Manual Credits, Up: Documentation

6.6 Printed Manuals

The FSF publishes some GNU manuals in printed form. To encourage = sales of=20 these manuals, the on-line versions of the manual should mention at the = very=20 start that the printed manual is available and should point at = information for=20 getting it=97for instance, with a link to the page http://www.gnu.org/order/ord= er.html.=20 This should not be included in the printed manual, though, because there = it is=20 redundant.=20

It is also useful to explain in the on-line forms of the manual how = the user=20 can print out the manual from the sources.=20

Next: Change Logs, Previous: Printed Manuals, Up: Documentation

6.7 The NEWS File

In addition = to its=20 manual, the package should have a file named NEWS which contains a list of user-visible = changes=20 worth mentioning. In each new release, add items to the front of the = file and=20 identify the version they pertain to. Don't discard old items; leave = them in the=20 file after the newer items. This way, a user upgrading from any previous = version=20 can see what is new.=20

If the NEWS file gets very = long, move=20 some of the older items into a file named ONEWS and put a note at the end referring the = user to=20 that file.=20

Next: Man=20 Pages, Previous: NEWS File, Up: Documentation

6.8 Change Logs

Keep a change log to describe all = the=20 changes made to program source files. The purpose of this is so that = people=20 investigating bugs in the future will know about the changes that might = have=20 introduced the bug. Often a new bug can be found by looking at what was = recently=20 changed. More importantly, change logs can help you eliminate conceptual = inconsistencies between different parts of a program, by giving you a = history of=20 how the conflicting concepts arose and who they came from.=20

Next: Style of Change Logs, Up: Change Logs

6.8.1 Change Log Concepts

You can think of the change log as a conceptual =93undo list=94 which = explains=20 how earlier versions were different from the current version. People can = see the=20 current version; they don't need the change log to tell them what is in = it. What=20 they want from a change log is a clear explanation of how the earlier = version=20 differed.=20

The change log file is normally called ChangeLog and covers an entire directory. = Each=20 directory can have its own change log, or a directory can use the change = log of=20 its parent directory=96it's up to you.=20

Another alternative is to record change log information with a = version=20 control system such as RCS or CVS. This can be converted automatically = to a=20 ChangeLog file using = rcs2log;=20 in Emacs, the command C-x v a = (vc-update-change-log)=20 does the job.=20

There's no need to describe the full purpose of the changes or how = they work=20 together. If you think that a change calls for explanation, you're = probably=20 right. Please do explain it=97but please put the explanation in comments = in the=20 code, where people will see it whenever they see the code. For example, = =93New=20 function=94 is enough for the change log when you add a function, = because there=20 should be a comment before the function definition to explain what it = does.=20

In the past, we recommended not mentioning changes in non-software = files=20 (manuals, help files, etc.) in change logs. However, we've been advised = that it=20 is a good idea to include them, for the sake of copyright records.=20

However, sometimes it is useful to write one line to describe the = overall=20 purpose of a batch of changes.=20

The easiest way to add an entry to ChangeLog is with the Emacs command M-x=20 add-change-log-entry. An entry should have an asterisk, the name = of the=20 changed file, and then in parentheses the name of the changed functions, = variables or whatever, followed by a colon. Then describe the changes = you made=20 to that function or variable.=20

Next: Simple Changes, Previous: Change Log Concepts, Up: Change Logs

6.8.2 Style of Change Logs

Here are some simple = examples of=20 change log entries, starting with the header line that says who made the = change=20 and when it was installed, followed by descriptions of specific changes. = (These=20 examples are drawn from Emacs and GCC.) 

     =
1998-08-17  Richard Stallman  <rms@gnu.org>
    =20
     * register.el (insert-register): Return nil.
     (jump-to-register): Likewise.
    =20
     * sort.el (sort-subr): Return nil.
    =20
     * tex-mode.el (tex-bibtex-file, tex-file, tex-region):
     Restart the tex shell if process is gone or stopped.
     (tex-shell-running): New function.
    =20
     * expr.c (store_one_arg): Round size up for move_block_to_reg.
     (expand_call): Round up when emitting USE insns.
     * stmt.c (assign_parms): Round size up for move_block_from_reg.

It's important to name the changed function or variable in full. = Don't=20 abbreviate function or variable names, and don't combine them. = Subsequent=20 maintainers will often search for a function name to find all the change = log=20 entries that pertain to it; if you abbreviate the name, they won't find = it when=20 they search.=20

For example, some people are tempted to abbreviate groups of function = names=20 by writing `* register.el=20 ({insert,jump-to}-register)'; this is not a good idea, = since=20 searching for jump-to-register or = insert-register=20 would not find that entry.=20

Separate unrelated change log entries with blank lines. When two = entries=20 represent parts of the same change, so that they work together, then = don't put=20 blank lines between them. Then you can omit the file name and the = asterisk when=20 successive entries are in the same file.=20

Break long lists of function names by closing continued lines with=20 `)', rather than `,', and opening the continuation with = `(' as in this example: 

  =
   * keyboard.c (menu_bar_items, tool_bar_items)
     (Fexecute_extended_command): Deal with `keymap' property.

When you install someone else's changes, put the contributor's name = in the=20 change log entry rather than in the text of the entry. In other words, = write=20 this: 

     2002-07-14  John Doe  =
<jdoe@gnu.org>
    =20
             * sewing.c: Make it sew.

rather than this: 

     =
2002-07-14  Usual Maintainer  <usual@gnu.org>
    =20
             * sewing.c: Make it sew.  Patch by jdoe@gnu.org.

As for the date, that should be the date you applied the change.=20

Next: Conditional Changes, Previous: Style of Change Logs, Up: Change Logs

6.8.3 Simple Changes

Certain simple kinds of changes don't need much detail in the change = log.=20

When you change the calling sequence of a function in a simple = fashion, and=20 you change all the callers of the function to use the new calling = sequence,=20 there is no need to make individual entries for all the callers that you = changed. Just write in the entry for the function being called, =93All = callers=20 changed=94=97like this: 

     * keyboard.c =
(Fcommand_execute): New arg SPECIAL.
     All callers changed.

When you change just comments or doc strings, it is enough to write = an entry=20 for the file, without mentioning the functions. Just =93Doc fixes=94 is = enough for=20 the change log.=20

There's no technical need to make change log entries for = documentation files.=20 This is because documentation is not susceptible to bugs that are hard = to fix.=20 Documentation does not consist of parts that must interact in a = precisely=20 engineered fashion. To correct an error, you need not know the history = of the=20 erroneous passage; it is enough to compare what the documentation says = with the=20 way the program actually works.=20

However, you should keep change logs for documentation files when the = project=20 gets copyright assignments from its contributors, so as to make the = records of=20 authorship more accurate.=20

Next: Indicating the Part Changed, Previous: Simple Changes, Up: Change Logs

6.8.4 Conditional Changes

C programs = often contain=20 compile-time #if conditionals. Many changes are = conditional;=20 sometimes you add a new definition which is entirely contained in a = conditional.=20 It is very useful to indicate in the change log the conditions for which = the=20 change applies.=20

Our convention for indicating conditional changes is to use square = brackets=20 around the name of the condition.=20

Here is a simple example, describing a change which is conditional = but does=20 not have a function or entity name associated with it: 

     * xterm.c [SOLARIS2]: Include string.h.

Here is an entry describing a new definition which is entirely = conditional.=20 This new definition for the macro FRAME_WINDOW_P is used = only when=20 HAVE_X_WINDOWS is defined: 

     * =
frame.h [HAVE_X_WINDOWS] (FRAME_WINDOW_P): Macro defined.

Here is an entry for a change within the function = init_display,=20 whose definition as a whole is unconditional, but the changes themselves = are=20 contained in a `#ifdef = HAVE_LIBNCURSES'=20 conditional: 

     * dispnew.c (init_display) =
[HAVE_LIBNCURSES]: If X, call tgetent.

Here is an entry for a change that takes affect only when a certain = macro is=20 not defined: 

     (gethostname) =
[!HAVE_SOCKETS]: Replace with winsock version.

Previous: Conditional Changes, Up: Change Logs

6.8.5 Indicating the Part Changed

Indicate the part of a function which changed by using angle brackets = enclosing an indication of what the changed part does. Here is an entry = for a=20 change in the part of the function sh-while-getopts that = deals with=20 sh commands: 

     * =
progmodes/sh-script.el (sh-while-getopts) <sh>: Handle case that
     user-specified option string is empty.

Next: Reading other Manuals, Previous: Change Logs, Up: Documentation

6.9 Man Pages

In the GNU project, man pages are = secondary.=20 It is not necessary or expected for every GNU program to have a man = page, but=20 some of them do. It's your choice whether to include a man page in your = program.=20

When you make this decision, consider that supporting a man page = requires=20 continual effort each time the program is changed. The time you spend on = the man=20 page is time taken away from more useful work.=20

For a simple program which changes little, updating the man page may = be a=20 small job. Then there is little reason not to include a man page, if you = have=20 one.=20

For a large program that changes a great deal, updating a man page = may be a=20 substantial burden. If a user offers to donate a man page, you may find = this=20 gift costly to accept. It may be better to refuse the man page unless = the same=20 person agrees to take full responsibility for maintaining it=97so that = you can=20 wash your hands of it entirely. If this volunteer later ceases to do the = job,=20 then don't feel obliged to pick it up yourself; it may be better to = withdraw the=20 man page from the distribution until someone else agrees to update it.=20

When a program changes only a little, you may feel that the = discrepancies are=20 small enough that the man page remains useful without updating. If so, = put a=20 prominent note near the beginning of the man page explaining that you = don't=20 maintain it and that the Texinfo manual is more authoritative. The note = should=20 say how to access the Texinfo documentation.=20

Be sure that man pages include a copyright statement and free = license. The=20 simple all-permissive license is appropriate for simple man pages: 

     Copying and distribution of this file, with or =
without modification,
     are permitted in any medium without royalty provided the copyright
     notice and this notice are preserved.

For long man pages, with enough explanation and documentation that = they can=20 be considered true manuals, use the GFDL (see License=20 for Manuals).=20

Finally, the GNU help2man program (http://www.gnu.org/softwar= e/help2man/)=20 is one way to automate generation of a man page, in this case from = --help output. This is sufficient in many = cases.=20

Previous: Man Pages, Up: Documentation

6.10 Reading other Manuals

There may be non-free books or documentation files that describe the = program=20 you are documenting.=20

It is ok to use these documents for reference, just as the author of = a new=20 algebra textbook can read other books on algebra. A large portion of any = non-fiction book consists of facts, in this case facts about how a = certain=20 program works, and these facts are necessarily the same for everyone who = writes=20 about the subject. But be careful not to copy your outline structure, = wording,=20 tables or examples from preexisting non-free documentation. Copying from = free=20 documentation may be ok; please check with the FSF about the individual = case.=20

Next: References, Previous: Documentation, Up: Top=20

7 The Release Process

Making a release is more than just = bundling=20 up your source files in a tar file and putting it up for FTP. You should = set up=20 your software so that it can be configured to run on a variety of = systems. Your=20 Makefile should conform to the GNU standards described below, and your = directory=20 layout should also conform to the standards discussed below. Doing so = makes it=20 easy to include your package into the larger framework of all GNU = software.=20

Next: Makefile Conventions, Up: Managing Releases

7.1 How Configuration Should Work

Each GNU distribution should come with a = shell=20 script named configure. This script is given arguments = which=20 describe the kind of machine and system you want to compile the program = for.=20

The configure script must record the configuration = options so=20 that they affect compilation.=20

One way to do this is to make a link from a standard name such as = config.h to the proper configuration file for = the=20 chosen system. If you use this technique, the distribution should = not=20 contain a file named config.h. = This is so=20 that people won't be able to build the program without configuring it = first.=20

Another thing that configure can do is to edit the = Makefile. If=20 you do this, the distribution should not contain a file named=20 Makefile. Instead, it should = include a file=20 Makefile.in which contains the = input used=20 for editing. Once again, this is so that people won't be able to build = the=20 program without configuring it first.=20

If configure does write the Makefile, then Makefile should have a target named = Makefile which causes configure = to be=20 rerun, setting up the same configuration that was set up last time. The = files=20 that configure reads should be listed as dependencies of=20 Makefile.=20

All the files which are output from the configure script = should=20 have comments at the beginning explaining that they were generated = automatically=20 using configure. This is so that users won't think of = trying to=20 edit them by hand.=20

The configure script should write a file named = config.status which describes which = configuration=20 options were specified when the program was last configured. This file = should be=20 a shell script which, if run, will recreate the same configuration.=20

The configure script should accept an option of the form = `--srcdir=3Ddirname' = to specify=20 the directory where sources are found (if it is not the current = directory). This=20 makes it possible to build the program in a separate directory, so that = the=20 actual source directory is not modified.=20

If the user does not specify `--srcdir',=20 then configure should check both . and .. to see if=20 it can find the sources. If it finds the sources in one of these places, = it=20 should use them from there. Otherwise, it should report that it cannot = find the=20 sources, and should exit with nonzero status.=20

Usually the easy way to support `--srcdir' is by editing a definition of=20 VPATH into the Makefile. Some rules may need to refer = explicitly to=20 the specified source directory. To make this possible, = configure=20 can add to the Makefile a variable named srcdir whose value = is=20 precisely the specified directory.=20

The configure script should also take an argument which=20 specifies the type of system to build the program for. This argument = should look=20 like this: 

     =
cpu-company-system

For example, an Athlon-based GNU/Linux system might be `i686-pc-linux-gnu'.=20

The configure script needs to be able to decode all = plausible=20 alternatives for how to describe a machine. Thus, `athlon-pc-gnu/linux' would be a valid alias. = There is a=20 shell script called config.sub that you can use as a = subroutine to=20 validate system types and canonicalize aliases.=20

The configure script should also take the option = --build=3Dbuildtype, which = should be=20 equivalent to a plain buildtype argument. For example, = `configure --build=3Di686-pc-linux-gnu' is = equivalent to=20 `configure i686-pc-linux-gnu'. = When the=20 build type is not specified by an option or argument, the = configure=20 script should normally guess it using the shell script config.guess.=20

Other = options=20 are permitted to specify in more detail the software or hardware present = on the=20 machine, to include or exclude optional parts of the package, or to = adjust the=20 name of some tools or arguments to them:=20

`--enable-feature[=3Dparameter]'
Configure the package to build and install an optional user-level = facility=20 called feature. This allows users to choose which optional = features=20 to include. Giving an optional parameter of `no' should omit feature, if it = is built by=20 default.=20

No `--enable' option should=20 ever cause one feature to replace another. No = `--enable' option should ever substitute one = useful=20 behavior for another useful behavior. The only proper use for = `--enable' is for questions of whether to = build part=20 of the program or exclude it.

`--with-package'
The package = package will be=20 installed, so configure this package to work with package. =

Possible values of package include `gnu-as' (or `gas'), `gnu-ld',=20 `gnu-libc', `gdb', `x', and=20 `x-toolkit'.=20

Do not use a `--with' option = to=20 specify the file name to use to find certain files. That is outside = the scope=20 of what `--with' options are = for.=20

`variable=3Dvalue'
Set the value of the variable variable to = value.=20 This is used to override the default values of commands or arguments = in the=20 build process. For example, the user could issue `configure CFLAGS=3D-g CXXFLAGS=3D-g' to = build with=20 debugging information and without the default optimization.=20

Specifying variables as arguments to configure, like = this: 

          ./configure CC=3Dgcc

is preferable to setting them in environment variables: 

          CC=3Dgcc ./configure

as it helps to recreate the same configuration later with = config.status.

All configure scripts should accept all of the = =93detail=94 options=20 and the variable settings, whether or not they make any difference to = the=20 particular package at hand. In particular, they should accept any option = that=20 starts with `--with-' or = `--enable-'. This is so users will be able to = configure=20 an entire GNU source tree at once with a single set of options.=20

You will note that the categories `--with-' and `--enable-' are narrow: they do = not=20 provide a place for any sort of option you might think of. That is = deliberate.=20 We want to limit the possible configuration options in GNU software. We = do not=20 want GNU programs to have idiosyncratic configuration options.=20

Packages that perform part of the compilation process may support=20 cross-compilation. In such a case, the host and target machines for the = program=20 may be different.=20

The configure script should normally treat the specified = type of=20 system as both the host and the target, thus producing a program which = works for=20 the same type of machine that it runs on.=20

To compile a program to run on a host type that differs from the = build type,=20 use the configure option --host=3Dhosttype, where = hosttype=20 uses the same syntax as buildtype. The host type normally = defaults to=20 the build type.=20

To configure a cross-compiler, cross-assembler, or what have you, you = should=20 specify a target different from the host, using the configure option=20 `--target=3Dtargettype'. The=20 syntax for targettype is the same as for the host type. So = the=20 command would look like this: 

     ./configure =
--host=3Dhosttype --target=3Dtargettype

The target type normally defaults to the host type. Programs for = which=20 cross-operation is not meaningful need not accept the `--target' option, because configuring an = entire=20 operating system for cross-operation is not a meaningful operation.=20

Some programs have ways of configuring themselves automatically. If = your=20 program is set up to do this, your configure script can = simply=20 ignore most of its arguments.

Next: Releases, Previous: Configuration, Up: Managing Releases

7.2 Makefile Conventions

This describes conventions for writing the Makefiles for GNU = programs. Using=20 Automake will help you write a Makefile that follows these conventions.=20

Next: Utilities in Makefiles, Up: Makefile Conventions

7.2.1 General Conventions for Makefiles

Every Makefile should contain this line: 

     =
SHELL =3D /bin/sh

to avoid trouble on systems where the = SHELL=20 variable might be inherited from the environment. (This is never a = problem with=20 GNU make.)=20

Different make programs have incompatible suffix lists = and=20 implicit rules, and this sometimes creates confusion or misbehavior. So = it is a=20 good idea to set the suffix list explicitly using only the suffixes you = need in=20 the particular Makefile, like this: 

     .SUFFIXES:
     .SUFFIXES: .c .o

The first line clears out the suffix list, the = second=20 introduces all suffixes which may be subject to implicit rules in this = Makefile.=20

Don't assume that . is in the = path for=20 command execution. When you need to run programs that are a part of your = package=20 during the make, please make sure that it uses ./ if the program is built as part of the = make or=20 $(srcdir)/ if the file is an = unchanging=20 part of the source code. Without one of these prefixes, the current = search path=20 is used.=20

The distinction between ./ = (the=20 build directory) and $(srcdir)/=20 (the source directory) is important because users can build = in a=20 separate directory using the `--srcdir'=20 option to configure. A rule of = the form: 

     foo.1 : foo.man sedscript
             sed -e sedscript foo.man > foo.1

will fail when the build directory is not the source = directory, because foo.man and = sedscript are in the source directory.=20

When using GNU make, relying on `VPATH' to find the source file will work in = the case=20 where there is a single dependency file, since the make = automatic=20 variable `$<' will represent = the source=20 file wherever it is. (Many versions of make set = `$<' only in implicit rules.) A Makefile = target like 

     foo.o : bar.c
             $(CC) -I. -I$(srcdir) $(CFLAGS) -c bar.c -o foo.o

should instead be written as 

     foo.o : bar.c
             $(CC) -I. -I$(srcdir) $(CFLAGS) -c $< -o $@

in order to allow `VPATH'=20 to work correctly. When the target has multiple dependencies, using an = explicit=20 `$(srcdir)' is the easiest way to = make the=20 rule work well. For example, the target above for foo.1 is best written as: 

     foo.1 : foo.man sedscript
             sed -e $(srcdir)/sedscript $(srcdir)/foo.man > $@

GNU distributions usually contain some files which are not source = files=97for=20 example, Info files, and the output from Autoconf, Automake, Bison or = Flex.=20 Since these files normally appear in the source directory, they should = always=20 appear in the source directory, not in the build directory. So Makefile = rules to=20 update them should put the updated files in the source directory.=20

However, if a file does not appear in the distribution, then the = Makefile=20 should not put it in the source directory, because building a program in = ordinary circumstances should not modify the source directory in any = way.=20

Try to make the build and installation targets, at least (and all = their=20 subtargets) work correctly with a parallel make.=20

Next: Command Variables, Previous: Makefile Basics, Up: Makefile Conventions

7.2.2 Utilities in Makefiles

Write the Makefile commands (and any shell scripts, such as=20 configure) to run in sh, not in = csh.=20 Don't use any special features of ksh or bash. =

The configure script and the Makefile rules for building = and=20 installation should not use any utilities directly except these: 

     cat cmp cp diff echo egrep expr false grep =
install-info
     ln ls mkdir mv pwd rm rmdir sed sleep sort tar test touch true

The compression program gzip can be used in the=20 dist rule.=20

Stick to the generally supported options for these programs. For = example,=20 don't use `mkdir -p', convenient = as it may=20 be, because most systems don't support it.=20

It is a good idea to avoid creating symbolic links in makefiles, = since a few=20 systems don't support them.=20

The Makefile rules for building and installation can also use = compilers and=20 related programs, but should do so via make variables so = that the=20 user can substitute alternatives. Here are some of the programs we mean: = 

     ar bison cc flex install ld ldconfig lex
     make makeinfo ranlib texi2dvi yacc

Use the following make variables to run those programs: = 

     $(AR) $(BISON) $(CC) $(FLEX) $(INSTALL) $(LD) =
$(LDCONFIG) $(LEX)
     $(MAKE) $(MAKEINFO) $(RANLIB) $(TEXI2DVI) $(YACC)

When you use ranlib or ldconfig, you should = make=20 sure nothing bad happens if the system does not have the program in = question.=20 Arrange to ignore an error from that command, and print a message before = the=20 command to tell the user that failure of this command does not mean a = problem.=20 (The Autoconf `AC_PROG_RANLIB' = macro can=20 help with this.)=20

If you use symbolic links, you should implement a fallback for = systems that=20 don't have symbolic links.=20

Additional utilities that can be used via Make variables are: 

     chgrp chmod chown mknod

It is ok to use other utilities in Makefile portions (or scripts) = intended=20 only for particular systems where you know those utilities exist.=20

Next: DESTDIR, Previous: Utilities in Makefiles, Up: Makefile Conventions

7.2.3 Variables for Specifying Commands

Makefiles should provide variables for overriding certain commands, = options,=20 and so on.=20

In particular, you should run most utility programs via variables. = Thus, if=20 you use Bison, have a variable named BISON whose default = value is=20 set with `BISON =3D bison', and = refer to it=20 with $(BISON) whenever you need to use Bison.=20

File management utilities such as ln, rm,=20 mv, and so on, need not be referred to through variables in = this=20 way, since users don't need to replace them with other programs.=20

Each program-name variable should come with an options variable that = is used=20 to supply options to the program. Append `FLAGS' to the program-name variable name to = get the=20 options variable name=97for example, BISONFLAGS. (The names = CFLAGS for the C compiler, YFLAGS for yacc, = and=20 LFLAGS for lex, are exceptions to this rule, but we keep = them=20 because they are standard.) Use CPPFLAGS in any compilation = command=20 that runs the preprocessor, and use LDFLAGS in any = compilation=20 command that does linking as well as in any direct use of = ld.=20

If there are C compiler options that must be used for proper = compilation of certain files, do not include them in = CFLAGS. Users=20 expect to be able to specify CFLAGS freely themselves. = Instead,=20 arrange to pass the necessary options to the C compiler independently of = CFLAGS, by writing them explicitly in the compilation = commands or=20 by defining an implicit rule, like this: 

     =
CFLAGS =3D -g
     ALL_CFLAGS =3D -I. $(CFLAGS)
     .c.o:
             $(CC) -c $(CPPFLAGS) $(ALL_CFLAGS) $<

Do include the `-g' option in=20 CFLAGS, because that is not required for proper=20 compilation. You can consider it a default that is only recommended. If = the=20 package is set up so that it is compiled with GCC by default, then you = might as=20 well include `-O' in the default = value of=20 CFLAGS as well.=20

Put CFLAGS last in the compilation command, after other=20 variables containing compiler options, so the user can use = CFLAGS=20 to override the others.=20

CFLAGS should be used in every invocation of the C = compiler,=20 both those which do compilation and those which do linking.=20

Every Makefile should define the variable INSTALL, which = is the=20 basic command for installing a file into the system.=20

Every Makefile should also define the variables = INSTALL_PROGRAM=20 and INSTALL_DATA. (The default for = INSTALL_PROGRAM=20 should be $(INSTALL); the default for = INSTALL_DATA=20 should be ${INSTALL} -m 644.) Then it should use those = variables as=20 the commands for actual installation, for executables and = non-executables=20 respectively. Minimal use of these variables is as follows: 

     $(INSTALL_PROGRAM) foo $(bindir)/foo
     $(INSTALL_DATA) libfoo.a $(libdir)/libfoo.a

However, it is preferable to support a DESTDIR prefix on = the=20 target files, as explained in the next section.=20

Always use a file name, not a directory name, as the = second=20 argument of the installation commands. Use a separate command for each = file to=20 be installed.=20

Next: Directory Variables, Previous: Command Variables, Up: Makefile Conventions

7.2.4 DESTDIR: support for staged=20 installs

DESTDIR is a = variable=20 prepended to each installed target file, like this: 

     $(INSTALL_PROGRAM) foo $(DESTDIR)$(bindir)/foo
     $(INSTALL_DATA) libfoo.a $(DESTDIR)$(libdir)/libfoo.a

The DESTDIR variable is specified by the user on the=20 make command line. For example: 

     =
make DESTDIR=3D/tmp/stage install

DESTDIR should be supported only in the = install* and uninstall* targets, as those are = the only=20 targets where it is useful.=20

If your installation step would normally install /usr/local/bin/foo and /usr/local/lib/libfoo.a, then an installation = invoked=20 as in the example above would install /tmp/stage/usr/local/bin/foo and /tmp/stage/usr/local/lib/libfoo.a instead.=20

Prepending the variable DESTDIR to each target in this = way=20 provides for staged installs, where the installed files are = not=20 placed directly into their expected location but are instead copied into = a=20 temporary location (DESTDIR). However, installed files = maintain=20 their relative directory structure and any embedded file names will not = be=20 modified.=20

You should not set the value of DESTDIR in your = Makefile at all; then the files are installed = into=20 their expected locations by default. Also, specifying = DESTDIR=20 should not change the operation of the software in any way, so its value = should=20 not be included in any file contents.=20

DESTDIR support is commonly used in package creation. It = is also=20 helpful to users who want to understand what a given package will = install where,=20 and to allow users who don't normally have permissions to install into = protected=20 areas to build and install before gaining those permissions. Finally, it = can be=20 useful with tools such as stow, where code is installed in = one=20 place but made to appear to be installed somewhere else using symbolic = links or=20 special mount operations. So, we strongly recommend GNU packages support = DESTDIR, though it is not an absolute requirement.=20

Next: Standard Targets, Previous: DESTDIR, Up: Makefile Conventions

7.2.5 Variables for Installation Directories

Installation directories should always be named by variables, so it = is easy=20 to install in a nonstandard place. The standard names for these = variables and=20 the values they should have in GNU packages are described below. They = are based=20 on a standard file system layout; variants of it are used in GNU/Linux = and other=20 modern operating systems.=20

Installers are expected to override these values when calling = make (e.g., make prefix=3D/usr = install or=20 configure (e.g., = configure=20 --prefix=3D/usr). GNU packages should not try to guess which value = should be=20 appropriate for these variables on the system they are being installed = onto: use=20 the default settings specified here so that all GNU packages behave = identically,=20 allowing the installer to achieve any desired layout.=20

These first two variables set the root for the installation. All the = other=20 installation directories should be subdirectories of one of these two, = and=20 nothing should be directly installed into these two directories.=20

prefix
A prefix used in constructing the = default=20 values of the variables listed below. The default value of = prefix=20 should be /usr/local. When = building the=20 complete GNU system, the prefix will be empty and /usr will be a symbolic link to /. (If you are using Autoconf, write it as=20 `@prefix@'.)=20

Running `make install' with = a=20 different value of prefix from the one used to build the = program=20 should not recompile the program.

exec_prefix
A prefix used in = constructing the=20 default values of some of the variables listed below. The default = value of=20 exec_prefix should be $(prefix). (If you are = using=20 Autoconf, write it as `@exec_prefix@'.)=20

Generally, $(exec_prefix) is used for directories that = contain=20 machine-specific files (such as executables and subroutine libraries), = while=20 $(prefix) is used directly for other directories.=20

Running `make install' with = a=20 different value of exec_prefix from the one used to build = the=20 program should not recompile the program.

Executable programs are installed in one of the following = directories.=20

bindir
The directory for installing = executable=20 programs that users can run. This should normally be /usr/local/bin, but write it as $(exec_prefix)/bin. (If you are using = Autoconf, write=20 it as `@bindir@'.)
sbindir
The directory for installing = executable=20 programs that can be run from the shell, but are only generally useful = to=20 system administrators. This should normally be /usr/local/sbin, but write it as = $(exec_prefix)/sbin. (If you are using = Autoconf,=20 write it as `@sbindir@'.)
libexecdir
The=20 directory for installing executable programs to be run by other = programs=20 rather than by users. This directory should normally be /usr/local/libexec, but write it as = $(exec_prefix)/libexec. (If you are using = Autoconf,=20 write it as `@libexecdir@'.)=20

The definition of `libexecdir' is the=20 same for all packages, so you should install your data in a = subdirectory=20 thereof. Most packages install their data under $(libexecdir)/package-name/, possibly within additional = subdirectories thereof,=20 such as $(libexecdir)/package-name/machine/version.

Data files used by the program during its execution are divided into=20 categories in two ways.=20

This makes for six different possibilities. However, we want to = discourage=20 the use of architecture-dependent files, aside from object files and = libraries.=20 It is much cleaner to make other data files architecture-independent, = and it is=20 generally not hard.=20

Here are the variables Makefiles should use to specify directories to = put=20 these various kinds of files in:=20

`datarootdir'
The root of the directory tree for read-only = architecture-independent data=20 files. This should normally be /usr/local/share, but write it as = $(prefix)/share. (If you are using = Autoconf, write it=20 as `@datarootdir@'.) = `datadir''s default value is based on this = variable;=20 so are `infodir', `mandir', and others.
`datadir'
The directory for installing idiosyncratic read-only=20 architecture-independent data files for this program. This is usually = the same=20 place as `datarootdir', but we = use the=20 two separate variables so that you can move these program-specific = files=20 without altering the location for Info files, man pages, etc.=20

This should normally be /usr/local/share, but write it as = $(datarootdir). (If you are using Autoconf, = write it=20 as `@datadir@'.)=20

The definition of `datadir' = is the=20 same for all packages, so you should install your data in a = subdirectory=20 thereof. Most packages install their data under $(datadir)/package-name/.

`sysconfdir'
The directory for installing read-only data files that pertain to = a single=20 machine=96that is to say, files for configuring a host. Mailer and = network=20 configuration files, /etc/passwd, and so=20 forth belong here. All the files in this directory should be ordinary = ASCII=20 text files. This directory should normally be /usr/local/etc, but write it as $(prefix)/etc. (If you are using Autoconf, = write it=20 as `@sysconfdir@'.)=20

Do not install executables here in this directory (they probably = belong in=20 $(libexecdir) or $(sbindir)). Also do not install files that = are=20 modified in the normal course of their use (programs whose purpose is = to=20 change the configuration of the system excluded). Those probably = belong in=20 $(localstatedir).

`sharedstatedir'
The directory for installing architecture-independent data files = which the=20 programs modify while they run. This should normally be /usr/local/com, but write it as $(prefix)/com. (If you are using Autoconf, = write it=20 as `@sharedstatedir@'.)
`localstatedir'
The directory for installing data files which the programs modify = while=20 they run, and that pertain to one specific machine. Users should never = need to=20 modify files in this directory to configure the package's operation; = put such=20 configuration information in separate files that go in $(datadir) or $(sysconfdir). $(localstatedir) should normally be = /usr/local/var, but write it as $(prefix)/var. (If you are using Autoconf, = write it=20 as `@localstatedir@'.) =

These variables specify the directory for installing certain specific = types=20 of files, if your program has them. Every GNU package should have Info = files, so=20 every program needs `infodir', = but not all=20 need `libdir' or `lispdir'.=20

`includedir'
The directory for = installing header files to be included by user programs with the C=20 `#include' preprocessor = directive. This=20 should normally be /usr/local/include,=20 but write it as $(prefix)/include. (If=20 you are using Autoconf, write it as `@includedir@'.)=20

Most compilers other than GCC do not look for header files in = directory=20 /usr/local/include. So = installing the=20 header files this way is only useful with GCC. Sometimes this is not a = problem=20 because some libraries are only really intended to work with GCC. But = some=20 libraries are intended to work with other compilers. They should = install their=20 header files in two places, one specified by includedir = and one=20 specified by oldincludedir.

`oldincludedir'
The directory for installing `#include' header files for use with = compilers other=20 than GCC. This should normally be /usr/include. (If you are using Autoconf, = you can=20 write it as `@oldincludedir@'.) =

The Makefile commands should check whether the value of=20 oldincludedir is empty. If it is, they should not try to = use it;=20 they should cancel the second installation of the header files.=20

A package should not replace an existing header in this directory = unless=20 the header came from the same package. Thus, if your Foo package = provides a=20 header file foo.h, then it = should install=20 the header file in the oldincludedir directory if either = (1)=20 there is no foo.h there or (2) = the=20 foo.h that exists came from the = Foo=20 package.=20

To tell whether foo.h came = from the=20 Foo package, put a magic string in the file=97part of a comment=97and=20 grep for that string.

`docdir'
The directory for installing documentation files (other than Info) = for=20 this package. By default, it should be /usr/local/share/doc/yourpkg, = but it=20 should be written as $(datarootdir)/doc/yourpkg. (If = you are=20 using Autoconf, write it as `@docdir@'.)=20 The yourpkg subdirectory, which may include a version = number,=20 prevents collisions among files with common names, such as README.
`infodir'
The directory for installing the Info files for this package. By = default,=20 it should be /usr/local/share/info, but=20 it should be written as $(datarootdir)/info. (If you are using = Autoconf,=20 write it as `@infodir@'.)=20 infodir is separate from docdir for = compatibility=20 with existing practice.
`htmldir'
`dvidir'
`pdfdir'
`psdir'
Directories for installing documentation files in the particular = format.=20 They should all be set to $(docdir) by default. (If you = are using=20 Autoconf, write them as `@htmldir@',=20 `@dvidir@', etc.) Packages = which supply=20 several translations of their documentation should install them in=20 `$(htmldir)/'ll, = `$(pdfdir)/'ll, etc. where = ll is=20 a locale abbreviation such as `en' or=20 `pt_BR'.
`libdir'
The directory for object files and libraries of object code. Do = not=20 install executables here, they probably ought to go in $(libexecdir) instead. The value of=20 libdir should normally be /usr/local/lib, but write it as $(exec_prefix)/lib. (If you are using = Autoconf, write=20 it as `@libdir@'.)
`lispdir'
The directory for installing any Emacs Lisp files in this package. = By=20 default, it should be /usr/local/share/emacs/site-lisp, but it = should be=20 written as $(datarootdir)/emacs/site-lisp.=20

If you are using Autoconf, write the default as `@lispdir@'. In order to make `@lispdir@' work, you need the following = lines in your=20 configure.in file: 

          lispdir=3D'${datarootdir}/emacs/site-lisp'
          AC_SUBST(lispdir)

`localedir'
The directory for installing locale-specific message catalogs for = this=20 package. By default, it should be /usr/local/share/locale, but it should be = written as=20 $(datarootdir)/locale. (If you = are using=20 Autoconf, write it as `@localedir@'.)=20 This directory usually has a subdirectory per locale.

Unix-style man pages are installed in one of the following:=20

`mandir'
The top-level directory for installing the man pages (if any) for = this=20 package. It will normally be /usr/local/share/man, but you should write = it as=20 $(datarootdir)/man. (If you are = using=20 Autoconf, write it as `@mandir@'.)
`man1dir'
The directory for installing section 1 man pages. Write it as = $(mandir)/man1.
`man2dir'
The directory for installing section 2 man pages. Write it as = $(mandir)/man2
`...'
Don't make the primary documentation for any GNU software = be a man=20 page. Write a manual in Texinfo instead. Man pages are just for the = sake of=20 people running GNU software on Unix, which is a secondary application=20 only.
`manext'
The file name extension for the installed man page. This should = contain a=20 period followed by the appropriate digit; it should normally be = `.1'.
`man1ext'
The file name extension for installed section 1 man pages.
`man2ext'
The file name extension for installed section 2 man pages.
`...'
Use these names instead of `manext'=20 if the package needs to install man pages in more than one section of = the=20 manual.

And finally, you should set the following variable:=20

`srcdir'
The directory for the sources being compiled. The value of this = variable=20 is normally inserted by the configure shell script. (If = you are=20 using Autoconf, use `srcdir =3D = @srcdir@'.)=20

For example: 

     
     
     # Common prefix for installation directories.
     # NOTE: This directory must exist when you start the install.
     prefix =3D /usr/local
     datarootdir =3D $(prefix)/share
     datadir =3D $(datarootdir)
     exec_prefix =3D $(prefix)
     # Where to put the executable for the command `gcc'.
     bindir =3D $(exec_prefix)/bin
     # Where to put the directories used by the compiler.
     libexecdir =3D $(exec_prefix)/libexec
     # Where to put the Info files.
     infodir =3D $(datarootdir)/info

If your program installs a large number of files into one of the = standard=20 user-specified directories, it might be useful to group them into a = subdirectory=20 particular to that program. If you do this, you should write the=20 install rule to create these subdirectories.=20

Do not expect the user to include the subdirectory name in the value = of any=20 of the variables listed above. The idea of having a uniform set of = variable=20 names for installation directories is to enable the user to specify the = exact=20 same values for several different GNU packages. In order for this to be = useful,=20 all the packages must be designed so that they will work sensibly when = the user=20 does so.=20

At times, not all of these variables may be implemented in the = current=20 release of Autoconf and/or Automake; but as of Autoconf 2.60,=20 we believe all of them are. When any are missing, the descriptions here = serve as=20 specifications for what Autoconf will implement. As a programmer, you = can either=20 use a development version of Autoconf or avoid using these variables = until a=20 stable release is made which supports them.=20

Next: Install Command Categories, Previous: Directory Variables, Up: Makefile Conventions

7.2.6 Standard Targets for Users

All GNU programs should have the following targets in their = Makefiles:=20

`all'
Compile the entire program. This should be the default target. = This target=20 need not rebuild any documentation files; Info files should normally = be=20 included in the distribution, and DVI (and other documentation format) = files=20 should be made only when explicitly asked for.=20

By default, the Make rules should compile and link with = `-g', so that executable programs have = debugging=20 symbols. Users who don't mind being helpless can strip the executables = later=20 if they wish.

`install'
Compile the program and copy the executables, libraries, and so on = to the=20 file names where they should reside for actual use. If there is a = simple test=20 to verify that a program is properly installed, this target should run = that=20 test.=20

Do not strip executables when installing them. Devil-may-care users = can use=20 the install-strip target to do that.=20

If possible, write the install target rule so that it = does not=20 modify anything in the directory where the program was built, provided = `make all' has just been done. = This is=20 convenient for building the program under one user name and installing = it=20 under another.=20

The commands should create all the directories in which files are = to be=20 installed, if they don't already exist. This includes the directories=20 specified as the values of the variables prefix and=20 exec_prefix, as well as all subdirectories that are = needed. One=20 way to do this is by means of an installdirs target as = described=20 below.=20

Use `-' before any command = for=20 installing a man page, so that make will ignore any = errors. This=20 is in case there are systems that don't have the Unix man page = documentation=20 system installed.=20

The way to install Info files is to copy them into $(infodir) with = $(INSTALL_DATA) (see Command=20 Variables), and then run the install-info program if = it is=20 present. install-info is a program that edits the Info=20 dir file to add or update the = menu entry=20 for the given Info file; it is part of the Texinfo package. Here is a = sample=20 rule to install an Info file: 

          =
$(DESTDIR)$(infodir)/foo.info: foo.info
                  $(POST_INSTALL)
          # There may be a newer info file in . than in srcdir.
                  -if test -f foo.info; then d=3D.; \
                   else d=3D$(srcdir); fi; \
                  $(INSTALL_DATA) $$d/foo.info $(DESTDIR)$@; \
          # Run install-info only if it exists.
          # Use `if' instead of just prepending `-' to the
          # line so we notice real errors from install-info.
          # We use `$(SHELL) -c' because some shells do not
          # fail gracefully when there is an unknown command.
                  if $(SHELL) -c 'install-info --version' \
                     >/dev/null 2>&1; then \
                    install-info --dir-file=3D$(DESTDIR)$(infodir)/dir \
                                 $(DESTDIR)$(infodir)/foo.info; \
                  else true; fi

When writing the install target, you must classify all = the=20 commands into three categories: normal ones, = pre-installation=20 commands and post-installation commands. See Install=20 Command Categories.

`install-html'
`install-dvi'
`install-pdf'
`install-ps'
These targets install documentation in formats other than Info; = they're=20 intended to be called explicitly by the person installing the package, = if that=20 format is desired. GNU prefers Info files, so these must be installed = by the=20 install target.=20

When you have many documentation files to install, we recommend = that you=20 avoid collisions and clutter by arranging for these targets to install = in=20 subdirectories of the appropriate installation directory, such as=20 htmldir. As one example, if your package has multiple = manuals,=20 and you wish to install HTML documentation with many files (such as = the=20 =93split=94 mode output by makeinfo --html), you'll = certainly want to=20 use subdirectories, or two nodes with the same name in different = manuals will=20 overwrite each other.=20

Please make these install-format targets = invoke the=20 commands for the format target, for example, by making=20 format a dependency.

`uninstall'
Delete all the installed files=97the copies that the `install' and `install-*' targets create.=20

This rule should not modify the directories where compilation is = done, only=20 the directories where files are installed.=20

The uninstallation commands are divided into three categories, just = like=20 the installation commands. See Install=20 Command Categories.

`install-strip'
Like install, but strip the executable files while = installing=20 them. In simple cases, this target can use the install = target in=20 a simple way: 
          install-strip:
                  $(MAKE) INSTALL_PROGRAM=3D'$(INSTALL_PROGRAM) -s' \
                          install

But if the package installs scripts as well as real executables, = the=20 install-strip target can't just refer to the = install=20 target; it has to strip the executables but not the scripts.=20

install-strip should not strip the executables in the = build=20 directory which are being copied for installation. It should only = strip the=20 copies that are installed.=20

Normally we do not recommend stripping an executable unless you are = sure=20 the program has no bugs. However, it can be reasonable to install a = stripped=20 executable for actual execution while saving the unstripped executable = elsewhere in case there is a bug.

`clean'
Delete all files in the current directory that are normally = created by=20 building the program. Also delete files in other directories if they = are=20 created by this makefile. However, don't delete the files that record = the=20 configuration. Also preserve files that could be made by building, but = normally aren't because the distribution comes with them. There is no = need to=20 delete parent directories that were created with `mkdir=20 -p', since they could have existed anyway.=20

Delete .dvi files here if = they are not=20 part of the distribution.

`distclean'
Delete all files in the current directory (or created by this = makefile)=20 that are created by configuring or building the program. If you have = unpacked=20 the source and built the program without creating any other files,=20 `make distclean' should leave = only the=20 files that were in the distribution. However, there is no need to = delete=20 parent directories that were created with `mkdir=20 -p', since they could have existed anyway.
`mostlyclean'
Like `clean', but may = refrain from=20 deleting a few files that people normally don't want to recompile. For = example, the `mostlyclean' = target for GCC=20 does not delete libgcc.a, = because=20 recompiling it is rarely necessary and takes a lot of time.
`maintainer-clean'
Delete almost everything that can be reconstructed with this = Makefile.=20 This typically includes everything deleted by distclean, = plus=20 more: C source files produced by Bison, tags tables, Info files, and = so on.=20

The reason we say =93almost everything=94 is that running the = command=20 `make maintainer-clean' should = not delete=20 configure even if configure can be remade using a rule in the = Makefile.=20 More generally, `make = maintainer-clean'=20 should not delete anything that needs to exist in order to run = configure and then begin to build the = program. Also,=20 there is no need to delete parent directories that were created with=20 `mkdir -p', since they could = have existed=20 anyway. These are the only exceptions; maintainer-clean = should=20 delete everything else that can be rebuilt.=20

The `maintainer-clean' = target is=20 intended to be used by a maintainer of the package, not by ordinary = users. You=20 may need special tools to reconstruct some of the files that = `make maintainer-clean' deletes. Since these = files are=20 normally included in the distribution, we don't take care to make them = easy to=20 reconstruct. If you find you need to unpack the full distribution = again, don't=20 blame us.=20

To help make users aware of this, the commands for the special=20 maintainer-clean target should start with these two: 

          @echo 'This command is intended for =
maintainers to use; it'
          @echo 'deletes files that may need special tools to rebuild.'

`TAGS'
Update a tags table for this program.
`info'
Generate any Info files needed. The best way to write the rules is = as=20 follows: 
          info: foo.info
         =20
          foo.info: foo.texi chap1.texi chap2.texi
                  $(MAKEINFO) $(srcdir)/foo.texi

You must define the variable MAKEINFO = in the=20 Makefile. It should run the makeinfo program, which is = part of=20 the Texinfo distribution.=20

Normally a GNU distribution comes with Info files, and that means = the Info=20 files are present in the source directory. Therefore, the Make rule = for an=20 info file should update it in the source directory. When users build = the=20 package, ordinarily Make will not update the Info files because they = will=20 already be up to date.

`dvi'
`html'
`pdf'
`ps'
Generate documentation files in the given format. These targets = should=20 always exist, but any or all can be a no-op if the given output format = cannot=20 be generated. These targets should not be dependencies of the = all=20 target; the user must manually invoke them.=20

Here's an example rule for generating DVI files from Texinfo: 

          dvi: foo.dvi
         =20
          foo.dvi: foo.texi chap1.texi chap2.texi
                  $(TEXI2DVI) $(srcdir)/foo.texi

You must define the variable TEXI2DVI = in the=20 Makefile. It should run the program texi2dvi, which is = part of=20 the Texinfo distribution.1 Alternatively, write just the = dependencies, and=20 allow GNU make to provide the command.=20

Here's another example, this one for generating HTML from Texinfo: = 

          html: foo.html
         =20
          foo.html: foo.texi chap1.texi chap2.texi
                  $(TEXI2HTML) $(srcdir)/foo.texi

Again, you would define the variable = TEXI2HTML=20 in the Makefile; for example, it might run makeinfo --no-split=20 --html (makeinfo is = part of the=20 Texinfo distribution).

`dist'
Create a distribution tar file for this program. The tar file = should be=20 set up so that the file names in the tar file start with a = subdirectory name=20 which is the name of the package it is a distribution for. This name = can=20 include the version number.=20

For example, the distribution tar file of GCC version 1.40 unpacks = into a=20 subdirectory named gcc-1.40.=20

The easiest way to do this is to create a subdirectory = appropriately named,=20 use ln or cp to install the proper files in = it, and=20 then tar that subdirectory.=20

Compress the tar file with gzip. For example, the = actual=20 distribution file for GCC version 1.40 is called gcc-1.40.tar.gz.=20

The dist target should explicitly depend on all = non-source=20 files that are in the distribution, to make sure they are up to date = in the=20 distribution. See Making= =20 Releases.

`check'
Perform self-tests (if any). The user must build the program = before=20 running the tests, but need not install the program; you should write = the=20 self-tests so that they work when the program is built but not = installed.=20

The following targets are suggested as conventional names, for = programs in=20 which they are useful.=20

installcheck
Perform installation tests (if any). The user must build and = install the=20 program before running the tests. You should not assume that = $(bindir) is in the search path.
installdirs
It's useful to add a target named `installdirs' to create the directories = where files=20 are installed, and their parent directories. There is a script called=20 mkinstalldirs which is = convenient for=20 this; you can find it in the Texinfo package. You can use a rule like this: 
          # Make sure all installation =
directories (e.g. $(bindir))
          # actually exist by making them if necessary.
          installdirs: mkinstalldirs
                  $(srcdir)/mkinstalldirs $(bindir) $(datadir) \
                                          $(libdir) $(infodir) \
                                          $(mandir)

or, if you wish to support DESTDIR, 

          =
# Make sure all installation directories (e.g. $(bindir))
          # actually exist by making them if necessary.
          installdirs: mkinstalldirs
                  $(srcdir)/mkinstalldirs \
                      $(DESTDIR)$(bindir) $(DESTDIR)$(datadir) \
                      $(DESTDIR)$(libdir) $(DESTDIR)$(infodir) \
                      $(DESTDIR)$(mandir)

This rule should not modify the directories where compilation is = done. It=20 should do nothing but create installation directories.

Previous: Standard Targets, Up: Makefile Conventions

7.2.7 Install Command Categories

When writing the=20 install target, you must classify all the commands into = three=20 categories: normal ones, pre-installation commands and=20 post-installation commands.=20

Normal commands move files into their proper places, and set their = modes.=20 They may not alter any files except the ones that come entirely from the = package=20 they belong to.=20

Pre-installation and post-installation commands may alter other = files; in=20 particular, they can edit global configuration files or data bases.=20

Pre-installation commands are typically executed before the normal = commands,=20 and post-installation commands are typically run after the normal = commands.=20

The most common use for a post-installation command is to run=20 install-info. This cannot be done with a normal command, = since it=20 alters a file (the Info directory) which does not come entirely and = solely from=20 the package being installed. It is a post-installation command because = it needs=20 to be done after the normal command which installs the package's Info = files.=20

Most programs don't need any pre-installation commands, but we have = the=20 feature just in case it is needed.=20

To classify the commands in the install rule into these = three=20 categories, insert category lines among them. A category line = specifies the category for the commands that follow.=20

A category line consists of a tab and a reference to a special Make = variable,=20 plus an optional comment at the end. There are three variables you can = use, one=20 for each category; the variable name specifies the category. Category = lines are=20 no-ops in ordinary execution because these three Make variables are = normally=20 undefined (and you should not define them in the makefile).=20

Here are the three possible category lines, each with a comment that = explains=20 what it means: 

             $(PRE_INSTALL)     =
# Pre-install commands follow.
             $(POST_INSTALL)    # Post-install =
commands follow.
             $(NORMAL_INSTALL)  # Normal commands =
follow.

If you don't use a category line at the beginning of the = install=20 rule, all the commands are classified as normal until the first category = line.=20 If you don't use any category lines, all the commands are classified as = normal.=20

These are the category lines for uninstall: 

             $(PRE_UNINSTALL)     # Pre-uninstall commands follow.
             $(POST_UNINSTALL)    # Post-uninstall =
commands follow.
             $(NORMAL_UNINSTALL)  # Normal commands =
follow.

Typically, a pre-uninstall command would be used for deleting entries = from=20 the Info directory.=20

If the install or uninstall target has any=20 dependencies which act as subroutines of installation, then you should = start=20 each dependency's commands with a category line, and start the = main=20 target's commands with a category line also. This way, you can ensure = that each=20 command is placed in the right category regardless of which of the = dependencies=20 actually run.=20

Pre-installation and post-installation commands should not run any = programs=20 except for these: 

     [ basename bash cat chgrp =
chmod chown cmp cp dd diff echo
     egrep expand expr false fgrep find getopt grep gunzip gzip
     hostname install install-info kill ldconfig ln ls md5sum
     mkdir mkfifo mknod mv printenv pwd rm rmdir sed sort tee
     test touch true uname xargs yes

The reason for distinguishing = the=20 commands in this way is for the sake of making binary packages. = Typically a=20 binary package contains all the executables and other files that need to = be=20 installed, and has its own method of installing them=97so it does not = need to run=20 the normal installation commands. But installing the binary package does = need to=20 execute the pre-installation and post-installation commands.=20

Programs to build binary packages work by extracting the = pre-installation and=20 post-installation commands. Here is one way of extracting the = pre-installation=20 commands (the -s option to = make is needed to silence messages about = entering=20 subdirectories): 

     make -s -n install -o =
all \
           PRE_INSTALL=3Dpre-install \
           POST_INSTALL=3Dpost-install \
           NORMAL_INSTALL=3Dnormal-install \
       | gawk -f pre-install.awk

where the file pre-install.awk could contain this: 

     $0 ~ /^(normal-install|post-install)[ \t]*$/ =
{on =3D 0}
     on {print $0}
     $0 ~ /^pre-install[ \t]*$/ {on =3D 1}

Previous: Makefile Conventions, Up: Managing Releases

7.3 Making Releases

You should identify each release = with a pair=20 of version numbers, a major version and a minor. We have no objection to = using=20 more than two numbers, but it is very unlikely that you really need = them.=20

Package the distribution of Foo version 69.96 up in a = gzipped=20 tar file with the name foo-69.96.tar.gz. It=20 should unpack into a subdirectory named foo-69.96.=20

Building and installing the program should never modify any of the = files=20 contained in the distribution. This means that all the files that form = part of=20 the program in any way must be classified into source files = and=20 non-source files. Source files are written by humans and = never=20 changed automatically; non-source files are produced from source files = by=20 programs under the control of the Makefile.=20

The = distribution=20 should contain a file named README which=20 gives the name of the package, and a general description of what it = does. It is=20 also good to explain the purpose of each of the first-level = subdirectories in=20 the package, if there are any. The README=20 file should either state the version number of the package, or refer to = where in=20 the package it can be found.=20

The README file should refer = to the file=20 INSTALL, which should contain an=20 explanation of the installation procedure.=20

The README file should also = refer to the=20 file which contains the copying conditions. The GNU GPL, if used, should = be in a=20 file called COPYING. If the GNU = LGPL is=20 used, it should be in a file called COPYING.LIB.=20

Naturally, all the source files must be in the distribution. It is = okay to=20 include non-source files in the distribution, provided they are = up-to-date and=20 machine-independent, so that building the distribution normally will = never=20 modify them. We commonly include non-source files produced by Bison,=20 lex, TeX, and makeinfo; this helps avoid = unnecessary=20 dependencies between our distributions, so that users can install = whichever=20 packages they want to install.=20

Non-source files that might actually be modified by building and = installing=20 the program should never be included in the = distribution. So if=20 you do distribute non-source files, always make sure they are up to date = when=20 you make a new distribution.=20

Make sure that the directory into which the distribution unpacks (as = well as=20 any subdirectories) are all world-writable (octal mode 777). This is so = that old=20 versions of tar which preserve the ownership and = permissions of the=20 files from the tar archive will be able to extract all the files even if = the=20 user is unprivileged.=20

Make sure that all the files in the distribution are world-readable.=20

Don't include any symbolic links in the distribution itself. If the = tar file=20 contains symbolic links, then people cannot even unpack it on systems = that don't=20 support symbolic links. Also, don't use multiple names for one file in = different=20 directories, because certain file systems cannot handle this and that = prevents=20 unpacking the distribution.=20

Try to make sure that all the file names will be unique on MS-DOS. A = name on=20 MS-DOS consists of up to 8 characters, optionally followed by a period = and up to=20 three characters. MS-DOS will truncate extra characters both before and = after=20 the period. Thus, foobarhacker.c = and=20 foobarhacker.o are not ambiguous; = they are=20 truncated to foobarha.c and = foobarha.o, which are distinct.=20

Include=20 in your distribution a copy of the texinfo.tex you used to test print any = *.texinfo or *.texi files.=20

Likewise, if your program uses small GNU software packages like = regex,=20 getopt, obstack, or termcap, include them in the distribution file. = Leaving them=20 out would make the distribution file a little smaller at the expense of = possible=20 inconvenience to a user who doesn't know what other files to get.=20

Next: Copying This Manual, Previous: Managing Releases, Up: Top=20

8 References to Non-Free Software and = Documentation

A GNU = program=20 should not recommend use of any non-free program. We can't stop some = people from=20 writing proprietary programs, or stop other people from using them, but = we can=20 and should refuse to advertise them to new potential customers. = Proprietary=20 software is a social and ethical problem, and the point of GNU is to = solve that=20 problem.=20

The GNU definition of free software is found on the GNU web site at = http://www.gnu.org/ph= ilosophy/free-sw.html,=20 and the definition of free documentation is found at http://www.gnu.org/p= hilosophy/free-doc.html.=20 A list of important licenses and whether they qualify as free is in http://www.gnu.org= /licenses/license-list.html.=20 The terms =93free=94 and =93non-free=94, used in this document, refer to = that=20 definition. If it is not clear whether a license qualifies as free under = this=20 definition, please ask the GNU Project by writing to licensing@gnu.org. We will answer, = and if=20 the license is an important one, we will add it to the list.=20

When a non-free program or system is well known, you can mention it = in=20 passing=97that is harmless, since users who might want to use it = probably already=20 know about it. For instance, it is fine to explain how to build your = package on=20 top of some widely used non-free operating system, or how to use it = together=20 with some widely used non-free program.=20

However, you should give only the necessary information to help those = who=20 already use the non-free program to use your program with it=97don't = give, or=20 refer to, any further information about the proprietary program, and = don't imply=20 that the proprietary program enhances your program, or that its = existence is in=20 any way a good thing. The goal should be that people already using the=20 proprietary program will get the advice they need about how to use your = free=20 program with it, while people who don't already use the proprietary = program will=20 not see anything to lead them to take an interest in it.=20

If a non-free program or system is obscure in your program's domain, = your=20 program should not mention or support it at all, since doing so would = tend to=20 popularize the non-free program more than it popularizes your program. = (You=20 cannot hope to find many additional users among the users of Foobar if = the users=20 of Foobar are few.)=20

Sometimes a program is free software in itself but depends on a = non-free=20 platform in order to run. For instance, many Java programs depend on the = parts=20 of Sun's Java implementation which are not yet free software, and won't = run on=20 the GNU Java Compiler (which does not yet have all the features) or = won't run=20 with the GNU Java libraries. We hope this particular problem will be = gone in a=20 few months, when Sun makes the standard Java libraries free software, = but of=20 course the general principle remains: you should not recommend programs = that=20 depend on non-free software to run.=20

Some free programs encourage the use of non-free software. A typical = example=20 is mplayer. It is free = software in=20 itself, and the free code can handle some kinds of files. However, = mplayer recommends use of non-free codecs = for other=20 kinds of files, and users that install mplayer are very likely to install those = codecs=20 along with it. To recommend mplayer is,=20 in effect, to recommend the non-free codecs. We must not do that, so we = cannot=20 recommend mplayer either.=20

In general, you should also not recommend programs that themselves = strongly=20 recommend the use of non-free software.=20

A GNU package should not refer the user to any non-free documentation = for=20 free software. Free documentation that can be included in free operating = systems=20 is essential for completing the GNU system, or any free operating = system, so it=20 is a major focus of the GNU Project; to recommend use of documentation = that we=20 are not allowed to use in GNU would weaken the impetus for the community = to=20 produce documentation that we can include. So GNU packages should never=20 recommend non-free documentation.=20

By contrast, it is ok to refer to journal articles and textbooks in = the=20 comments of a program for explanation of how it functions, even though = they be=20 non-free. This is because we don't include such things in the GNU system = even if=20 we are allowed to=97they are outside the scope of an operating system = project.=20

Referring to a web site that describes or recommends a non-free = program is in=20 effect promoting that software, so please do not make links (or mention = by name)=20 web sites that contain such material. This policy is relevant = particularly for=20 the web pages for a GNU package.=20

Following links from nearly any web site can lead to non-free = software; this=20 is an inescapable aspect of the nature of the web, and in itself is no = objection=20 to linking to a site. As long as the site does not itself recommend a = non-free=20 program, there is no need be concerned about the sites it links to for = other=20 reasons.=20

Thus, for example, you should not make a link to AT&T's web site, = because=20 that recommends AT&T's non-free software packages; you should not = make a=20 link to a site that links to AT&T's site saying it is a place to get = a=20 non-free program; but if a site you want to link to refers to AT&T's = web=20 site in some other context (such as long-distance telephone service), = that is=20 not a problem.=20

Next: Index, Previous: References, Up: Top=20

Appendix A Copying This Manual

Up: Copying This Manual

A.1 GNU Free Documentation License

Version 1.2, November 2002
     Copyright =A9 2000,2001,2002 Free Software =
Foundation, Inc.
     51 Franklin St, Fifth Floor, Boston, MA  02110-1301, USA
    =20
     Everyone is permitted to copy and distribute verbatim copies
     of this license document, but changing it is not allowed.
  1. PREAMBLE=20

    The purpose of this License is to make a manual, textbook, or other = functional and useful document free in the sense of = freedom: to=20 assure everyone the effective freedom to copy and redistribute it, = with or=20 without modifying it, either commercially or noncommercially. = Secondarily,=20 this License preserves for the author and publisher a way to get = credit for=20 their work, while not being considered responsible for modifications = made by=20 others.=20

    This License is a kind of =93copyleft=94, which means that = derivative works of=20 the document must themselves be free in the same sense. It complements = the GNU=20 General Public License, which is a copyleft license designed for free=20 software.=20

    We have designed this License in order to use it for manuals for = free=20 software, because free software needs free documentation: a free = program=20 should come with manuals providing the same freedoms that the software = does.=20 But this License is not limited to software manuals; it can be used = for any=20 textual work, regardless of subject matter or whether it is published = as a=20 printed book. We recommend this License principally for works whose = purpose is=20 instruction or reference.

  2. APPLICABILITY AND DEFINITIONS=20

    This License applies to any manual or other work, in any medium, = that=20 contains a notice placed by the copyright holder saying it can be = distributed=20 under the terms of this License. Such a notice grants a world-wide,=20 royalty-free license, unlimited in duration, to use that work under = the=20 conditions stated herein. The =93Document=94, below, refers to any = such manual or=20 work. Any member of the public is a licensee, and is addressed as = =93you=94. You=20 accept the license if you copy, modify or distribute the work in a way = requiring permission under copyright law.=20

    A =93Modified Version=94 of the Document means any work containing = the Document=20 or a portion of it, either copied verbatim, or with modifications = and/or=20 translated into another language.=20

    A =93Secondary Section=94 is a named appendix or a front-matter = section of the=20 Document that deals exclusively with the relationship of the = publishers or=20 authors of the Document to the Document's overall subject (or to = related=20 matters) and contains nothing that could fall directly within that = overall=20 subject. (Thus, if the Document is in part a textbook of mathematics, = a=20 Secondary Section may not explain any mathematics.) The relationship = could be=20 a matter of historical connection with the subject or with related = matters, or=20 of legal, commercial, philosophical, ethical or political position = regarding=20 them.=20

    The =93Invariant Sections=94 are certain Secondary Sections whose = titles are=20 designated, as being those of Invariant Sections, in the notice that = says that=20 the Document is released under this License. If a section does not fit = the=20 above definition of Secondary then it is not allowed to be designated = as=20 Invariant. The Document may contain zero Invariant Sections. If the = Document=20 does not identify any Invariant Sections then there are none.=20

    The =93Cover Texts=94 are certain short passages of text that are = listed, as=20 Front-Cover Texts or Back-Cover Texts, in the notice that says that = the=20 Document is released under this License. A Front-Cover Text may be at = most 5=20 words, and a Back-Cover Text may be at most 25 words.=20

    A =93Transparent=94 copy of the Document means a machine-readable = copy,=20 represented in a format whose specification is available to the = general=20 public, that is suitable for revising the document straightforwardly = with=20 generic text editors or (for images composed of pixels) generic paint = programs=20 or (for drawings) some widely available drawing editor, and that is = suitable=20 for input to text formatters or for automatic translation to a variety = of=20 formats suitable for input to text formatters. A copy made in an = otherwise=20 Transparent file format whose markup, or absence of markup, has been = arranged=20 to thwart or discourage subsequent modification by readers is not = Transparent.=20 An image format is not Transparent if used for any substantial amount = of text.=20 A copy that is not =93Transparent=94 is called =93Opaque=94.=20

    Examples of suitable formats for Transparent copies include plain = ascii without markup, Texinfo input format, LaTeX = input=20 format, SGML or XML using a = publicly=20 available DTD, and standard-conforming simple=20 HTML, PostScript or PDF designed = for=20 human modification. Examples of transparent image formats include=20 PNG, XCF and = JPG.=20 Opaque formats include proprietary formats that can be read and edited = only by=20 proprietary word processors, SGML or = XML=20 for which the DTD and/or processing tools are not = generally=20 available, and the machine-generated HTML, = PostScript or=20 PDF produced by some word processors for output = purposes=20 only.=20

    The =93Title Page=94 means, for a printed book, the title page = itself, plus=20 such following pages as are needed to hold, legibly, the material this = License=20 requires to appear in the title page. For works in formats which do = not have=20 any title page as such, =93Title Page=94 means the text near the most = prominent=20 appearance of the work's title, preceding the beginning of the body of = the=20 text.=20

    A section =93Entitled XYZ=94 means a named subunit of the Document = whose title=20 either is precisely XYZ or contains XYZ in parentheses following text = that=20 translates XYZ in another language. (Here XYZ stands for a specific = section=20 name mentioned below, such as =93Acknowledgements=94, = =93Dedications=94,=20 =93Endorsements=94, or =93History=94.) To =93Preserve the Title=94 of = such a section when=20 you modify the Document means that it remains a section =93Entitled = XYZ=94=20 according to this definition.=20

    The Document may include Warranty Disclaimers next to the notice = which=20 states that this License applies to the Document. These Warranty = Disclaimers=20 are considered to be included by reference in this License, but only = as=20 regards disclaiming warranties: any other implication that these = Warranty=20 Disclaimers may have is void and has no effect on the meaning of this = License.=20

  3. VERBATIM COPYING=20

    You may copy and distribute the Document in any medium, either = commercially=20 or noncommercially, provided that this License, the copyright notices, = and the=20 license notice saying this License applies to the Document are = reproduced in=20 all copies, and that you add no other conditions whatsoever to those = of this=20 License. You may not use technical measures to obstruct or control the = reading=20 or further copying of the copies you make or distribute. However, you = may=20 accept compensation in exchange for copies. If you distribute a large = enough=20 number of copies you must also follow the conditions in section 3.=20

    You may also lend copies, under the same conditions stated above, = and you=20 may publicly display copies.

  4. COPYING IN QUANTITY=20

    If you publish printed copies (or copies in media that commonly = have=20 printed covers) of the Document, numbering more than 100, and the = Document's=20 license notice requires Cover Texts, you must enclose the copies in = covers=20 that carry, clearly and legibly, all these Cover Texts: Front-Cover = Texts on=20 the front cover, and Back-Cover Texts on the back cover. Both covers = must also=20 clearly and legibly identify you as the publisher of these copies. The = front=20 cover must present the full title with all words of the title equally=20 prominent and visible. You may add other material on the covers in = addition.=20 Copying with changes limited to the covers, as long as they preserve = the title=20 of the Document and satisfy these conditions, can be treated as = verbatim=20 copying in other respects.=20

    If the required texts for either cover are too voluminous to fit = legibly,=20 you should put the first ones listed (as many as fit reasonably) on = the actual=20 cover, and continue the rest onto adjacent pages.=20

    If you publish or distribute Opaque copies of the Document = numbering more=20 than 100, you must either include a machine-readable Transparent copy = along=20 with each Opaque copy, or state in or with each Opaque copy a = computer-network=20 location from which the general network-using public has access to = download=20 using public-standard network protocols a complete Transparent copy of = the=20 Document, free of added material. If you use the latter option, you = must take=20 reasonably prudent steps, when you begin distribution of Opaque copies = in=20 quantity, to ensure that this Transparent copy will remain thus = accessible at=20 the stated location until at least one year after the last time you = distribute=20 an Opaque copy (directly or through your agents or retailers) of that = edition=20 to the public.=20

    It is requested, but not required, that you contact the authors of = the=20 Document well before redistributing any large number of copies, to = give them a=20 chance to provide you with an updated version of the Document.

  5. MODIFICATIONS=20

    You may copy and distribute a Modified Version of the Document = under the=20 conditions of sections 2 and 3 above, provided that you release the = Modified=20 Version under precisely this License, with the Modified Version = filling the=20 role of the Document, thus licensing distribution and modification of = the=20 Modified Version to whoever possesses a copy of it. In addition, you = must do=20 these things in the Modified Version:=20

    1. Use in the Title Page (and on the covers, if any) a title = distinct from=20 that of the Document, and from those of previous versions (which = should, if=20 there were any, be listed in the History section of the Document). = You may=20 use the same title as a previous version if the original publisher = of that=20 version gives permission.=20
    2. List on the Title Page, as authors, one or more persons or = entities=20 responsible for authorship of the modifications in the Modified = Version,=20 together with at least five of the principal authors of the Document = (all of=20 its principal authors, if it has fewer than five), unless they = release you=20 from this requirement.=20
    3. State on the Title page the name of the publisher of the = Modified=20 Version, as the publisher.=20
    4. Preserve all the copyright notices of the Document.=20
    5. Add an appropriate copyright notice for your modifications = adjacent to=20 the other copyright notices.=20
    6. Include, immediately after the copyright notices, a license = notice=20 giving the public permission to use the Modified Version under the = terms of=20 this License, in the form shown in the Addendum below.=20
    7. Preserve in that license notice the full lists of Invariant = Sections and=20 required Cover Texts given in the Document's license notice.=20
    8. Include an unaltered copy of this License.=20
    9. Preserve the section Entitled =93History=94, Preserve its Title, = and add to=20 it an item stating at least the title, year, new authors, and = publisher of=20 the Modified Version as given on the Title Page. If there is no = section=20 Entitled =93History=94 in the Document, create one stating the = title, year,=20 authors, and publisher of the Document as given on its Title Page, = then add=20 an item describing the Modified Version as stated in the previous = sentence.=20
    10. Preserve the network location, if any, given in the Document for = public=20 access to a Transparent copy of the Document, and likewise the = network=20 locations given in the Document for previous versions it was based = on. These=20 may be placed in the =93History=94 section. You may omit a network = location for=20 a work that was published at least four years before the Document = itself, or=20 if the original publisher of the version it refers to gives = permission.=20
    11. For any section Entitled =93Acknowledgements=94 or = =93Dedications=94, Preserve=20 the Title of the section, and preserve in the section all the = substance and=20 tone of each of the contributor acknowledgements and/or dedications = given=20 therein.=20
    12. Preserve all the Invariant Sections of the Document, unaltered = in their=20 text and in their titles. Section numbers or the equivalent are not=20 considered part of the section titles.=20
    13. Delete any section Entitled =93Endorsements=94. Such a section = may not be=20 included in the Modified Version.=20
    14. Do not retitle any existing section to be Entitled = =93Endorsements=94 or to=20 conflict in title with any Invariant Section.=20
    15. Preserve any Warranty Disclaimers.

    If the Modified Version includes new front-matter sections or = appendices=20 that qualify as Secondary Sections and contain no material copied from = the=20 Document, you may at your option designate some or all of these = sections as=20 invariant. To do this, add their titles to the list of Invariant = Sections in=20 the Modified Version's license notice. These titles must be distinct = from any=20 other section titles.=20

    You may add a section Entitled =93Endorsements=94, provided it = contains nothing=20 but endorsements of your Modified Version by various parties=97for = example,=20 statements of peer review or that the text has been approved by an=20 organization as the authoritative definition of a standard.=20

    You may add a passage of up to five words as a Front-Cover Text, = and a=20 passage of up to 25 words as a Back-Cover Text, to the end of the list = of=20 Cover Texts in the Modified Version. Only one passage of Front-Cover = Text and=20 one of Back-Cover Text may be added by (or through arrangements made = by) any=20 one entity. If the Document already includes a cover text for the same = cover,=20 previously added by you or by arrangement made by the same entity you = are=20 acting on behalf of, you may not add another; but you may replace the = old one,=20 on explicit permission from the previous publisher that added the old = one.=20

    The author(s) and publisher(s) of the Document do not by this = License give=20 permission to use their names for publicity for or to assert or imply=20 endorsement of any Modified Version.

  6. COMBINING DOCUMENTS=20

    You may combine the Document with other documents released under = this=20 License, under the terms defined in section 4 above for modified = versions,=20 provided that you include in the combination all of the Invariant = Sections of=20 all of the original documents, unmodified, and list them all as = Invariant=20 Sections of your combined work in its license notice, and that you = preserve=20 all their Warranty Disclaimers.=20

    The combined work need only contain one copy of this License, and = multiple=20 identical Invariant Sections may be replaced with a single copy. If = there are=20 multiple Invariant Sections with the same name but different contents, = make=20 the title of each such section unique by adding at the end of it, in=20 parentheses, the name of the original author or publisher of that = section if=20 known, or else a unique number. Make the same adjustment to the = section titles=20 in the list of Invariant Sections in the license notice of the = combined work.=20

    In the combination, you must combine any sections Entitled = =93History=94 in the=20 various original documents, forming one section Entitled = =93History=94; likewise=20 combine any sections Entitled =93Acknowledgements=94, and any sections = Entitled=20 =93Dedications=94. You must delete all sections Entitled = =93Endorsements.=94

  7. COLLECTIONS OF DOCUMENTS=20

    You may make a collection consisting of the Document and other = documents=20 released under this License, and replace the individual copies of this = License=20 in the various documents with a single copy that is included in the=20 collection, provided that you follow the rules of this License for = verbatim=20 copying of each of the documents in all other respects.=20

    You may extract a single document from such a collection, and = distribute it=20 individually under this License, provided you insert a copy of this = License=20 into the extracted document, and follow this License in all other = respects=20 regarding verbatim copying of that document.

  8. AGGREGATION WITH INDEPENDENT WORKS=20

    A compilation of the Document or its derivatives with other = separate and=20 independent documents or works, in or on a volume of a storage or = distribution=20 medium, is called an =93aggregate=94 if the copyright resulting from = the=20 compilation is not used to limit the legal rights of the compilation's = users=20 beyond what the individual works permit. When the Document is included = in an=20 aggregate, this License does not apply to the other works in the = aggregate=20 which are not themselves derivative works of the Document.=20

    If the Cover Text requirement of section 3 is applicable to these = copies of=20 the Document, then if the Document is less than one half of the entire = aggregate, the Document's Cover Texts may be placed on covers that = bracket the=20 Document within the aggregate, or the electronic equivalent of covers = if the=20 Document is in electronic form. Otherwise they must appear on printed = covers=20 that bracket the whole aggregate.

  9. TRANSLATION=20

    Translation is considered a kind of modification, so you may = distribute=20 translations of the Document under the terms of section 4. Replacing = Invariant=20 Sections with translations requires special permission from their = copyright=20 holders, but you may include translations of some or all Invariant = Sections in=20 addition to the original versions of these Invariant Sections. You may = include=20 a translation of this License, and all the license notices in the = Document,=20 and any Warranty Disclaimers, provided that you also include the = original=20 English version of this License and the original versions of those = notices and=20 disclaimers. In case of a disagreement between the translation and the = original version of this License or a notice or disclaimer, the = original=20 version will prevail.=20

    If a section in the Document is Entitled =93Acknowledgements=94, = =93Dedications=94,=20 or =93History=94, the requirement (section 4) to Preserve its Title = (section 1)=20 will typically require changing the actual title.

  10. TERMINATION=20

    You may not copy, modify, sublicense, or distribute the Document = except as=20 expressly provided for under this License. Any other attempt to copy, = modify,=20 sublicense or distribute the Document is void, and will automatically=20 terminate your rights under this License. However, parties who have = received=20 copies, or rights, from you under this License will not have their = licenses=20 terminated so long as such parties remain in full compliance.

  11. FUTURE REVISIONS OF THIS LICENSE=20

    The Free Software Foundation may publish new, revised versions of = the GNU=20 Free Documentation License from time to time. Such new versions will = be=20 similar in spirit to the present version, but may differ in detail to = address=20 new problems or concerns. See http://www.gnu.org/copyleft/.=20

    Each version of the License is given a distinguishing version = number. If=20 the Document specifies that a particular numbered version of this = License =93or=20 any later version=94 applies to it, you have the option of following = the terms=20 and conditions either of that specified version or of any later = version that=20 has been published (not as a draft) by the Free Software Foundation. = If the=20 Document does not specify a version number of this License, you may = choose any=20 version ever published (not as a draft) by the Free Software = Foundation.=20

A.1.1 ADDENDUM: How to use this License for = your=20 documents

To use this License in a document you have written, include a copy of = the=20 License in the document and put the following copyright and license = notices just=20 after the title page: 

       Copyright (C)  =
year  your name.
       Permission is granted to copy, distribute and/or modify this =
document
       under the terms of the GNU Free Documentation License, Version =
1.2
       or any later version published by the Free Software Foundation;
       with no Invariant Sections, no Front-Cover Texts, and no =
Back-Cover
       Texts.  A copy of the license is included in the section entitled =
``GNU
       Free Documentation License''.

If you have Invariant Sections, Front-Cover Texts and Back-Cover = Texts,=20 replace the =93with...Texts.=94 line with this: 

         with the Invariant Sections being =
list their titles, with
         the Front-Cover Texts being list, and with the =
Back-Cover Texts
         being list.

If you have Invariant Sections without Cover Texts, or some other = combination=20 of the three, merge those two alternatives to suit the situation.=20

If your document contains nontrivial examples of program code, we = recommend=20 releasing these examples in parallel under your choice of free software = license,=20 such as the GNU General Public License, to permit their use in free = software.

Previous: Copying This Manual, Up: Top=20

Index

Footnotes

[1] texi2dvi uses TeX to do the real = work of=20 formatting. TeX is not distributed with Texinfo.