Information for maintainers of GNU software, last updated March 22, 2006.
Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
Permission is granted to make and distribute verbatim copies of this entire document without royalty provided the copyright notice and this permission notice are preserved.
This file contains guidelines and advice for someone who is the maintainer of a GNU program on behalf of the GNU Project. Everyone is entitled to change and redistribute GNU software; you need not pay attention to this file to get permission. But if you want to maintain a version for widespread distribution, we suggest you follow these guidelines; if you would like to be a GNU maintainer, then it is essential to follow these guidelines.
Please send corrections or suggestions for this document to maintainers@gnu.org. If you make a suggestion, please include a suggested new wording for it, to help us consider the suggestion efficiently. We prefer a context diff to the maintain.texi file, but if you don't have that file, you can make a context diff for some other version of this document, or propose it in any way that makes it clear.
This document uses the gender-neutral third-person pronouns “person”, “per”, “pers” and “perself” which were promoted, and perhaps invented, by Marge Piercy in Woman on the Edge of Time. They are used just like “she”, “her”, “hers” and “herself”, except that they apply equally to males and females. For example, “Person placed per new program under the GNU GPL, to let the public benefit from per work, and to enable per to feel person has done the right thing.”
The directory /gd/gnuorg is found on the GNU file server,
currently fencepost.gnu.org; if you are the maintainer of a GNU
package, you should have an account there. Contact
accounts@gnu.org if you don't have one. (You can also ask
for accounts for people who help you a large amount in working on the
package.) /gd/gnuorg/maintain.tar.gz is a tar file containing
all of these files in that directory which are mentioned in this file;
it is updated daily.
This release of the GNU Maintenance Instructions was last updated March 22, 2006.
With good fortune, you will continue maintaining your package for many decades. But sometimes for various reasons maintainers decide to step down.
If you're the official maintainer of a GNU package and you decide to step down, please inform the GNU Project (maintainers@gnu.org). We need to know that the package no longer has a maintainer, so we can look for and appoint a new maintainer.
If you have an idea for who should take over, please tell maintainers@gnu.org your suggestion. The appointment of a new maintainer needs the GNU Project's confirmation, but your judgment that a person is capable of doing the job will carry a lot of weight.
As your final act as maintainer, it would be helpful to set up the
package under savannah.gnu.org (see Old Versions). This will
make it much easier for the new maintainer to pick up where you left off
and will ensure that the CVS tree is not misplaced if it takes us a
while to find a new maintainer.
Unless your package is a fairly small, you probably won't do all the work on it yourself. Most maintainers recruit other developers to help.
Sometimes people will offer to help. Some of them will be capable, while others will not. It's up to you to determine who provides useful help, and encourage those people to participate more.
Some of the people who offer to help will support the GNU Project, while others may be interested for other reasons. Some will support the goals of the Free Software Movement, but some may not. They are all welcome to help with the work—we don't ask people's views or motivations before they contribute to GNU packages.
As a consequence, you cannot expect all contributors to support the GNU Project, or to have a concern for its policies and standards. So part of your job as maintainer is to exercise your authority on these points when they arise. No matter how much of the work other people do, you are in charge of what goes in the release. When a crucial point arises, you should calmly state your decision and stick to it.
Sometimes a package has several co-maintainers who share the role of maintainer. Unlike developers who help, co-maintainers have actually been appointed jointly as the maintainers of the package, and they carry out the maintainer's functions together. If you would like to propose some of your developers as co-maintainers, please contact maintainers@gnu.org.
This chapter describes procedures you should follow for legal reasons as you maintain the program, to avoid legal difficulties.
If you maintain an FSF-copyrighted package certain legal procedures are required when incorporating legally significant changes written by other people. This ensures that the FSF has the legal right to distribute the package, and the standing to defend its GPL-covered status in court if necessary.
Before incorporating significant changes, make sure that the person who wrote the changes has signed copyright papers and that the Free Software Foundation has received and signed them. We may also need a disclaimer from the person's employer.
To check whether papers have been received, look in /gd/gnuorg/copyright.list. If you can't look there directly, fsf-records@gnu.org can check for you. Our clerk can also check for papers that are waiting to be entered and inform you when expected papers arrive.
The directory /gd/gnuorg is found on the GNU machines,
currently fencepost.gnu.org; if you are the maintainer of a GNU
package, you should have an account on them. Contact
accounts@gnu.org if you don't have one. (You can also ask
for accounts for people who help you a large amount in working on the
package.)
In order for the contributor to know person should sign papers, you need to ask for the necessary papers. If you don't know per well, and you don't know that person is used to our ways of handling copyright papers, then it might be a good idea to raise the subject with a message like this:
Would you be willing to assign the copyright to the Free Software Foundation, so that we could install it in program?
or
Would you be willing to sign a copyright disclaimer to put this change in the public domain, so that we can install it in program?
If the contributor wants more information, you can send per /gd/gnuorg/conditions.text, which explains per options (assign vs. disclaim) and their consequences.
Once the conversation is under way and the contributor is ready for
more details, you should send one of the templates that are found in
the directory /gd/gnuorg/Copyright/; they are also available
from the doc/Copyright/ directory of the gnulib project
at http://savannah.gnu.org/projects/gnulib. This section
explains which templates you should use in which circumstances.
Please don't use any of the templates except for those listed
here, and please don't change the wording.
Once the conversation is under way, you can send the contributor the precise wording and instructions by email. Before you do this, make sure to get the current version of the template you will use! We change these templates occasionally—don't keep using an old version.
For large changes, ask the contributor for an assignment. Send per a
copy of the file request-assign.changes. (Like all the
‘request-’ files, it is in /gd/gnuorg/Copyright and in
gnulib.)
For medium to small changes, request a disclaimer by sending per the file request-disclaim.changes.
If the contributor is likely to keep making changes, person might want to sign an assignment for all per future changes to the program. So it is useful to offer per that alternative. If person wants to do it that way, send per the request-assign.future.
When you send a request- file, you don't need to fill in anything before sending it. Just send the file verbatim to the contributor. The file gives per instructions for how to ask the FSF to mail per the papers to sign. The request- file also raises the issue of getting a copyright disclaimer from the contributor's employer.
When the contributor emails the form to the FSF, the FSF sends per papers to sign. If person signs them right away, the whole process takes about two weeks–mostly waiting for letters to go back and forth.
For less common cases, we have template files you should send to the contributor. Be sure to fill in the name of the person and the name of the program in these templates, where it says ‘NAME OF PERSON’ and ‘NAME OF PROGRAM’, before sending; otherwise person might sign without noticing them, and the papers would be useless. Note that in some templates there is more than one place to put the name of the program or the name of the person; be sure to change all of them. All the templates raise the issue of an employer's disclaimer as well.
You do not need to ask for separate papers for a manual that is distributed only in the software package it describes. But if we sometimes distribute the manual separately (for instance, if we publish it as a book), then we need separate legal papers for changes in the manual. For smaller changes, use disclaim.changes.manual; for larger ones, use assign.changes.manual. To cover both past and future changes to a manual, you can use assign.future.manual. For a translation of a manual, use assign.translation.manual.
If a contributor is reluctant to sign an assignment for a large change, and is willing to sign a disclaimer instead, that is acceptable, so you should offer this alternative if it helps you reach agreement. We prefer an assignment for a larger change, so that we can enforce the GNU GPL for the new text, but a disclaimer is enough to let us use the text.
If you maintain a collection of programs, occasionally someone will contribute an entire separate program or manual that should be added to the collection. Then you can use the files request-assign.program, disclaim.program, assign.manual, and disclaim.manual. We very much prefer an assignment for a new separate program or manual, unless it is quite small, but a disclaimer is acceptable if the contributor insists on handling the matter that way.
If a contributor wants the FSF to publish only a pseudonym, that is ok. The contributor should say this, and state the desired pseudonym, when answering the request- form. The actual legal papers will use the real name, but the FSF will publish only the pseudonym. When using one of the other forms, fill in the real name but ask the contributor to discuss the use of a pseudonym with assign@gnu.org before sending back the signed form.
Although there are other templates besides the ones listed here, they are for special circumstances; please do not use them without getting advice from assign@gnu.org.
If you are not sure what to do, then please ask assign@gnu.org for advice; if the contributor asks you questions about the meaning and consequences of the legal papers, and you don't know the answers, you can forward them to assign@gnu.org and we will answer.
Please do not try changing the wording of a template yourself. If you think a change is needed, please talk with assign@gnu.org, and we will work with a lawyer to decide what to do.
If a person contributes more than around 15 lines of code and/or text that is legally significant for copyright purposes, which means we need copyright papers for it as described above.
A change of just a few lines (less than 15 or so) is not legally significant for copyright. A regular series of repeated changes, such as renaming a symbol, is not legally significant even if the symbol has to be renamed in many places. Keep in mind, however, that a series of minor changes by the same person can add up to a significant contribution. What counts is the total contribution of the person; it is irrelevant which parts of it were contributed when.
Copyright does not cover ideas. If someone contributes ideas but no text, these ideas may be morally significant as contributions, and worth giving credit for, but they are not significant for copyright purposes. Likewise, bug reports do not count for copyright purposes.
When giving credit to people whose contributions are not legally significant for copyright purposes, be careful to make that fact clear. The credit should clearly say they did not contribute significant code or text.
When people's contributions are not legally significant because they did not write code, do this by stating clearly what their contribution was. For instance, you could write this:
/*
* Ideas by:
* Richard Mlynarik <mly@adoc.xerox.com> (1997)
* Masatake Yamato <masata-y@is.aist-nara.ac.jp> (1999)
*/
Ideas by: makes it clear that Mlynarik and Yamato here
contributed only ideas, not code. Without the Ideas by: note,
several years from now we would find it hard to be sure whether they
had contributed code, and we might have to track them down and ask
them.
When you record a small patch in a change log file, first search for previous changes by the same person, and see if his past contributions, plus the new one, add up to something legally significant. If so, you should get copyright papers for all his changes before you install the new change.
If that is not so, you can install the small patch. Write ‘(tiny change)’ after the patch author's name, like this:
2002-11-04 Robert Fenk <Robert.Fenk@gmx.de> (tiny change)
Keep correct records of which portions were written by whom. This is very important. These records should say which files parts of files, were written by each person, and which files or portions were revised by each person. This should include installation scripts as well as manuals and documentation files—everything.
These records don't need to be as detailed as a change log. They don't need to distinguish work done at different times, only different people. They don't need describe changes in more detail than which files or parts of a file were changed. And they don't need to say anything about the function or purpose of a file or change–the Register of Copyrights doesn't care what the text does, just who wrote or contributed to which parts.
The list should also mention if certain files distributed in the same package are really a separate program.
Only the contributions that are legally significant for copyright purposes (see Legally Significant) need to be listed. Small contributions, ideas, etc., can be omitted.
For example, this would describe an early version of GAS:
Dean Elsner first version of all files except gdb-lines.c and m68k.c.
Jay Fenlason entire files gdb-lines.c and m68k.c, most of app.c,
plus extensive changes in messages.c, input-file.c, write.c
and revisions elsewhere.
Note: GAS is distributed with the files obstack.c and obstack.h, but
they are considered a separate package, not part of GAS proper.
Please keep these records in a file named AUTHORS in the source directory for the program itself.
You should maintain a proper copyright notice and a license notice in each nontrivial file in the package. (Any file more than ten lines long is nontrivial for this purpose.) This includes header files and interface definitions for building or running the program, documentation files, and any supporting files. If a file has been explicitly placed in the public domain, then instead of a copyright notice, it should have a notice saying explicitly that it is in the public domain.
Even image files and sound files should contain copyright notices and license notices, if they can. Some formats do not have room for textual annotations; for these files, state the copyright and copying permissions in a README file in the same directory.
Change log files should have a copyright notice and license notice at the end, since new material is added at the beginning but the end remains the end.
When a file is automatically generated from some other file in the distribution, it is useful for the automatic procedure to copy the copyright notice and permission notice of the file it is generated from, if possible. Alternatively, put a notice at the beginning saying which file it is generated from.
A copyright notice looks like this:
Copyright (C) year1, year2, year3 copyright-holder
The copyright-holder may be the Free Software Foundation, Inc., or someone else; you should know who is the copyright holder for your package.
Replace the ‘(C)’ with a C-in-a-circle symbol if it is available. For example, use ‘@copyright{}’ in a Texinfo file. However, stick with parenthesized ‘C’ unless you know that C-in-a-circle will work. For example, a program's standard --version message should use parenthesized ‘C’ by default, though message translations may use C-in-a-circle in locales where that symbol is known to work.
To update the list of year numbers, add each year in which you have made nontrivial changes to the package. (Here we assume you're using a publicly accessible revision control server, so that every revision installed is also immediately and automatically published.) When you add the new year, it is not required to keep track which files have seen significant changes in the new year and which have not. It is recommended and simpler to add the new year to all files in the package, and be done with it for the rest of the year.
For files which are regularly copied from another project (such as ‘gnulib’), the copyright notice should left as it is in the original.
Don't delete old year numbers, though; they can indicate when older versions might theoretically go into the public domain. If you copy a file into the package from some other program, keep the copyright years that come with the file.
Do not abbreviate the year list using a range; for instance, do not write ‘1996--1998’; instead, write ‘1996, 1997, 1998’.
The copyright statement may be split across multiple lines, both in source files and in any generated output. This often happens for files with a long history, having many different years of publication.
For an FSF-copyrighted package, if you have followed the procedures to obtain legal papers, each file should have just one copyright holder: the Free Software Foundation, Inc. You should edit the file's copyright notice to list that name and only that name.
But if contributors are not all assigning their copyrights to a single copyright holder, it can easily happen that one file has several copyright holders. Each contributor of nontrivial text is a copyright holder.
In that case, you should always include a copyright notice in the name of main copyright holder of the file. You can also include copyright notices for other copyright holders as well, and this is a good idea for those who have contributed a large amount and for those who specifically ask for notices in their names. (Sometimes the license on code that you copy in may require preserving certain copyright notices.) But you don't have to include a notice for everyone who contributed to the file (which would be rather inconvenient).
Sometimes a program has an overall copyright notice that refers to the whole program. It might be in the README file, or it might be displayed when the program starts up. This copyright notice should mention the year of completion of the most recent major version; it can mention years of completion of previous major versions, but that is optional.
Every nontrivial file needs a license notice as well as the copyright notice. (Without a license notice giving permission to copy and change the file would make the file non-free.)
The package itself should contain a full copy of GPL (conventionally in a file named COPYING) and the GNU Free Documentation License (included within your documentation). If the package contains any files distributed under the Lesser GPL, it should contain a full copy of that as well (conventionally in a file named COPYING.LIB).
You can get the official versions of these files from three places. You can use whichever is the most convenient for you.
fencepost.gnu.org. (You can ask accounts@gnu.org
for an account there if you don't have one).
gnulib project on savannah.gnu.org, which you
can access via anonymous CVS. See
http://savannah.gnu.org/projects/gnulib.
The official Texinfo sources for the licenses are also available in those same places, so you can include them in your documentation. A GFDL-covered manual must include the GFDL in this way. See GNU Sample Texts, for a full example in a Texinfo manual.
Typically the license notice for program files (including build scripts, configure files and makefiles) should cite the GPL, like this:
This file is part of GNU programGNU program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2, or (at your option) any later version.
GNU program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with program; see the file COPYING. If not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
But in a small program which is just a few files, you can use this instead:
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
Documentation files should have license notices also. Manuals should use the GNU Free Documentation License. Here is an example of the license notice to use after the copyright notice. Please adjust the list of invariant sections as appropriate for your manual. (If there are none, then say “with no invariant sections”.) See GNU Sample Texts, for a full example in a Texinfo manual.
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 the
Invariant Sections being "GNU General Public License", with the
Front-Cover Texts being ``A GNU Manual,'' and with the Back-Cover Texts
as in (a) below. A copy of the license is included in the section
entitled "GNU Free Documentation License".
(a) The FSF's Back-Cover Text is: ``You are free to copy and modify
this GNU Manual. Buying copies from GNU Press supports the FSF in
developing GNU and promoting software freedom.''
If the FSF does not publish this manual on paper, then omit the last sentence in (a) that talks about copies from GNU Press. If the FSF is not the copyright holder, then replace ‘FSF’ with the appropriate name.
See http://www.gnu.org/licenses/fdl-howto.html for more advice about how to use the GNU FDL.
If the manual is over 400 pages, or if the FSF thinks it might be a good choice for publishing on paper, then please include our standard invariant section which explains the importance of free documentation. Write to assign@gnu.org to get a copy of this section.
Note that when you distribute several manuals together in one software package, their on-line forms can share a single copy of the GFDL (see section 6). However, the printed (‘.dvi’) forms should each contain a copy of the GFDL, unless they are set up to be printed and published only together. Therefore, it is usually simplest to include the GFDL in each manual.
Small supporting files, short manuals (under 300 lines long) and rough documentation (README files, INSTALL files, etc) can use a simple all-permissive license like this one:
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.
If your package distributes Autoconf macros that are intended to be used (hence distributed) by third-party packages under possibly incompatible licenses, you may also use the above all-permissive license for these macros.
If you would like help with license issues or with using the GFDL, please contact licensing@gnu.org.
When maintaining an FSF-copyrighted GNU package, you may occasionally want to use a general-purpose free software module which offers a useful functionality, as a “library” facility (though the module is not always packaged technically as a library).
In a case like this, it would be unreasonable to ask the author of that module to assign the copyright to the FSF. After all, person did not write it specifically as a contribution to your package, so it would be impertinent to ask per, out of the blue, “Please give the FSF your copyright.”
So the thing to do in this case is to make your program use the module, but not consider it a part of your program. There are two reasonable methods of doing this:
For this method, it is not necessary to treat the module as a library and make a ‘.a’ file from it. You can link with the ‘.o’ files directly in the usual manner.
Both of these methods create an irregularity, and our lawyers have told us to minimize the amount of such irregularity. So consider using these methods only for general-purpose modules that were written for other programs and released separately for general use. For anything that was written as a contribution to your package, please get papers signed.
Don't feel obligated to include every change that someone asks you to include. You must judge which changes are improvements—partly based on what you think the users will like, and partly based on your own judgment of what is better. If you think a change is not good, you should reject it.
If someone sends you changes which are useful, but written in an ugly way or hard to understand and maintain in the future, don't hesitate to ask per to clean up their changes before you merge them. Since the amount of work we can do is limited, the more we convince others to help us work efficiently, the faster GNU will advance.
If the contributor will not or can not make the changes clean enough, then it is legitimate to say “I can't install this in its present form; I can only do so if you clean it up.” Invite per to distribute per changes another way, or to find other people to make them clean enough for you to install and maintain.
The only reason to do these cleanups yourself is if (1) it is easy, less work than telling the author what to clean up, or (2) the change is an important one, important enough to be worth the work of cleaning it up.
The GNU Coding Standards are a good thing to send people when you ask them to clean up changes (see Contents). The Emacs Lisp manual contains an appendix that gives coding standards for Emacs Lisp programs; it is good to urge authors to read it (see Tips and Standards).
Most GNU packages run on a wide range of platforms. These platforms are not equally important.
The most important platforms for a GNU package to support are GNU and GNU/Linux. Developing the GNU operating system is the whole point of the GNU Project; a GNU package exists to make the whole GNU system more powerful. So please keep that goal in mind and let it shape your work. For instance, every new feature you add should work on GNU, and GNU/Linux if possible too. If a new feature only runs on GNU and GNU/Linux, it could still be acceptable. However, a feature that runs only on other systems and not on GNU or GNU/Linux makes no sense in a GNU package.
You will naturally want to keep the program running on all the platforms it supports. But you personally will not have access to most of these platforms–so how should you do it?
Don't worry about trying to get access to all of these platforms. Even if you did have access to all the platforms, it would be inefficient for you to test the program on each platform yourself. Instead, you should test the program on a few platforms, including GNU or GNU/Linux, and let the users test it on the other platforms. You can do this through a pretest phase before the real release; when there is no reason to expect problems, in a package that is mostly portable, you can just make a release and let the users tell you if anything unportable was introduced.
It is important to test the program personally on GNU or GNU/Linux, because these are the most important platforms for a GNU package. If you don't have access to one of these platforms, please ask maintainers@gnu.org to help you out.
Supporting other platforms is optional—we do it when that seems like a good idea, but we don't consider it obligatory. If the users don't take care of a certain platform, you may have to desupport it unless and until users come forward to help. Conversely, if a user offers changes to support an additional platform, you will probably want to install them, but you don't have to. If you feel the changes are complex and ugly, if you think that they will increase the burden of future maintenance, you can and should reject them. This includes both free platforms such as NetBSD or FreeBSD and non-free platforms such as Windows.
Once a program is in use, you will get bug reports for it. Most GNU programs have their own special lists for sending bug reports. The advertised bug-reporting email address should always be ‘bug-program@gnu.org’, to help show users that the program is a GNU package, but it is ok to set up that list to forward to another site for further forwarding. The package distribution should state the name of the bug-reporting list in a prominent place, and ask users to help us by reporting bugs there.
We also have a catch-all list, bug-gnu-utils@gnu.org, which is used for all GNU programs that don't have their own specific lists. But nowadays we want to give each program its own bug-reporting list and move away from using bug-gnu-utils.
If you are the maintainer of a GNU package, you should have an account on the GNU servers; contact accounts@gnu.org if you don't have one. (You can also ask for accounts for people who help you a large amount in working on the package.) With this account, you can edit /com/mailer/aliases to create a new unmanaged list or add yourself to an existing unmanaged list. A comment near the beginning of that file explains how to create a Mailman-managed mailing list.
But if you don't want to learn how to do those things, you can alternatively ask alias-file@gnu.org to add you to the bug-reporting list for your program. To set up a new list, contact new-mailing-list@gnu.org. You can subscribe to a list managed by Mailman by sending mail to the corresponding ‘-request’ address.
You should moderate postings from non-subscribed addresses on your
mailing lists, to prevent propagation of unwanted messages (“spam”)
to subscribers and to the list archives. For lists controlled by
Mailman, you can do this by setting Privacy Options - Sender
Filter - generic_nonmember_action to Hold, and then
periodically (daily is best) reviewing the held messages, accepting
the real ones and discarding the junk.
When you receive bug reports, keep in mind that bug reports are crucial for your work. If you don't know about problems, you cannot fix them. So always thank each person who sends a bug report.
You don't have an obligation to give more response than that, though. The main purpose of bug reports is to help you contribute to the community by improving the next version of the program. Many of the people who report bugs don't realize this—they think that the point is for you to help them individually. Some will ask you to focus on that instead of on making the program better. If you comply with their wishes, you will have been distracted from the job of maintaining the program.
For example, people sometimes report a bug in a vague (and therefore useless) way, and when you ask for more information, they say, “I just wanted to see if you already knew the solution” (in which case the bug report would do nothing to help improve the program). When this happens, you should explain to them the real purpose of bug reports. (A canned explanation will make this more efficient.)
When people ask you to put your time into helping them use the program, it may seem “helpful” to do what they ask. But it is much less helpful than improving the program, which is the maintainer's real job.
By all means help individual users when you feel like it, if you feel you have the time available. But be careful to limit the amount of time you spend doing this—don't let it eat away the time you need to maintain the program! Know how to say no; when you are pressed for time, just “thanks for the bug report—I will fix it” is enough response.
Some GNU packages, such as Emacs and GCC, come with advice about how to make bug reports useful. If you want to copy and adapt that, it could be a very useful thing to do.
It is very important to keep backup files of all source files of GNU. You can do this using RCS, CVS or PRCS if you like. The easiest way to use RCS or CVS is via the Version Control library in Emacs; Concepts of Version Control.
The history of previous revisions and log entries is very important for future maintainers of the package, so even if you do not make it publicly accessible, be careful not to put anything in the repository or change log that you would not want to hand over to another maintainer some day.
The GNU Project provides a CVS server that GNU software packages can
use: subversions.gnu.org. (The name refers to the multiple
versions and their subversions that are stored in a CVS repository.)
You don't have to use this repository, but if you plan to allow public
read-only access to your development sources, it is convenient for
people to be able to find various GNU packages in a central place. The
CVS Server is managed by cvs-hackers@gnu.org.
The GNU project also provides additional developer resources on
subversions.gnu.org through its savannah.gnu.org
interface. All GNU maintainers are encouraged to take advantage of
these facilities, as savannah can serve to foster a sense of
community among all GNU developers and help in keeping up with project
management.
It is important to follow the GNU conventions when making GNU software distributions.
The tar file for version m.n of program foo should be
named foo-m.n.tar. It should unpack into a
subdirectory named foo-m.n. Tar files should not
unpack into files in the current directory, because this is inconvenient
if the user happens to unpack into a directory with other files in it.
Here is how the Makefile for Bison creates the tar file. This method is good for other programs.
dist: bison.info
echo bison-`sed -e '/version_string/!d' \
-e 's/[^0-9.]*\([0-9.]*\).*/\1/' -e q version.c` > .fname
-rm -rf `cat .fname`
mkdir `cat .fname`
dst=`cat .fname`; for f in $(DISTFILES); do \
ln $(srcdir)/$$f $$dst/$$f || { echo copying $$f; \
cp -p $(srcdir)/$$f $$dst/$$f ; } \
done
tar --gzip -chf `cat .fname`.tar.gz `cat .fname`
-rm -rf `cat .fname` .fname
Source files that are symbolic links to other file systems cannot be
installed in the temporary directory using ln, so use cp
if ln fails.
Using Automake is a good way to take care of writing the dist
target.
If the program is large, it is useful to make a set of diffs for each release, against the previous important release.
At the front of the set of diffs, put a short explanation of which version this is for and which previous version it is relative to. Also explain what else people need to do to update the sources properly (for example, delete or rename certain files before installing the diffs).
The purpose of having diffs is that they are small. To keep them small, exclude files that the user can easily update. For example, exclude info files, DVI files, tags tables, output files of Bison or Flex. In Emacs diffs, we exclude compiled Lisp files, leaving it up to the installer to recompile the patched sources.
When you make the diffs, each version should be in a directory suitably named—for example, gcc-2.3.2 and gcc-2.3.3. This way, it will be very clear from the diffs themselves which version is which.
If you use GNU diff to make the patch, use the options
‘-rc2P’. That will put any new files into the output as “entirely
different.” Also, the patch's context diff headers should have dates
and times in Universal Time using traditional Unix format, so that patch
recipients can use GNU patch's ‘-Z’ option. For example,
you could use the following Bourne shell command to create the patch:
LC_ALL=C TZ=UTC0 diff -rc2P gcc-2.3.2 gcc-2.3.3 | \
gzip -9 >gcc-2.3.2-2.3.3.patch.gz
If the distribution has subdirectories in it, then the diffs probably include some files in the subdirectories. To help users install such patches reliably, give them precise directions for how to run patch. For example, say this:
To apply these patches, cd to the main directory of the program
and then use `patch -p1'. `-p1' avoids guesswork in choosing
which subdirectory to find each file in.
It's wise to test your patch by applying it to a copy of the old version, and checking that the result exactly matches the new version.
ftp.gnu.org
GNU packages are distributed through directory /gnu on
ftp.gnu.org. Each package should have a subdirectory
named after the package, and all the distribution files for the package
should go in that subdirectory.
See Automated FTP Uploads, for procedural details of putting new
versions on ftp.gnu.org.
When you release a greatly changed new major version of a program, you might want to do so as a pretest. This means that you make a tar file, but send it only to a group of volunteers that you have recruited. (Use a suitable GNU mailing list/newsgroup to recruit them.)
We normally use the FTP server alpha.gnu.org for pretests and
prerelease versions. See Automated FTP Uploads, for procedural details
of putting new versions on alpha.gnu.org.
Once a program gets to be widely used and people expect it to work solidly, it is a good idea to do pretest releases before each “real” release.
There are two ways of handling version numbers for pretest versions. One method is to treat them as versions preceding the release you are going to make.
In this method, if you are about to release version 4.6 but you want to do a pretest first, call it 4.5.90. If you need a second pretest, call it 4.5.91, and so on. If you are really unlucky and ten pretests are not enough, after 4.5.99 you could advance to 4.5.990 and so on. (You could also use 4.5.100, but 990 has the advantage of sorting in the right order.)
The other method is to attach a date to the release number that is coming. For a pretest for version 4.6, made on Dec 10, 2002, this would be 4.6.20021210. A second pretest made the same day could be 4.6.20021210.1.
For development snapshots that are not formal pretests, using just the date without the version numbers is ok too.
One thing that you should never do is to release a pretest with the same version number as the planned real release. Many people will look only at the version number (in the tar file name, in the directory name that it unpacks into, or wherever they can find it) to determine whether a tar file is the latest version. People might look at the test release in this way and mistake it for the real release. Therefore, always change the number when you release changed code.
In order to upload new releases to ftp.gnu.org or
alpha.gnu.org, you first need to register the necessary
information. Then, you can perform uploads yourself, with no
intervention needed by the system administrators.
To register your information to perform automated uploads, send a message, preferably GPG-signed, to ftp-upload@gnu.org with the following:
The administrators will acknowledge your message when they have added the proper GPG keys as authorized to upload files for the corresponding packages.
Once you have registered your information, as described in the previous section, you will be able to do unattended ftp uploads using the following procedure.
For each upload destined for ftp.gnu.org or
alpha.gnu.org, three files (a triplet) need to be
uploaded via ftp to the host ftp-upload.gnu.org.
Upload the triplet via anonymous ftp to ftp-upload.gnu.org. If
the upload is destined for ftp.gnu.org, then place the triplet
in the /incoming/ftp directory. If the upload is destined for
alpha.gnu.org, then place the triplet in the
/incoming/alpha directory.
Uploads are processed every five minutes. Uploads that are in progress when the upload processing script is running are handled properly, so do not worry about the timing of your upload.
The directive file should contain one line, excluding the clearsigned data GPG that inserts, which specifies the final destination directory where items (1) and (2) to be placed.
For example, the foo.tar.gz.directive file might contain the single line:
directory: bar/v1
This directory line indicates that foo.tar.gz and
foo.tar.gz.sig are part of package bar. If you were to
upload the triplet to /incoming/ftp, and the system can
positively authenticate the signatures, then the files
foo.tar.gz and foo.tar.gz.sig will be placed in the
directory gnu/bar/v1 of the ftp.gnu.org site.
The directive file can be used to create currently non-existent
directory trees, as long as they are under the package directory for
your package (in the example above, that is bar).
Your designated upload email addresses (see Automated Upload Registration) are sent a message if there are any problems processing an upload for your package. You also receive a message when your +upload has been successfully processed.
If you have difficulties processing an upload, email ftp-upload@gnu.org.
When you have a new release, please make an announcement. You can maintain your own mailing list for announcements if you like, or you can use the moderated general GNU announcements list, info-gnu@gnu.org.
If you use your own list, you can decide as you see fit what events are worth announcing. If you use info-gnu@gnu.org, please do not announce pretest releases, only real releases. But real releases do include releases made just to fix bugs.
Please write pages about your package for installation on
www.gnu.org. The pages should follow our usual standards for web
pages (see http://www.gnu.org/server); we chose them in order to
support a wide variety of browsers, to focus on information rather than
flashy eye candy, and to keep the site simple and uniform.
The simplest way to maintain the web pages for your project is to
register the project on savannah.gnu.org. Then you can edit
the pages using CVS. You can keep the source files there too, but if
you want to use savannah.gnu.org only for the web pages, simply
register a “web-only” project.
If you don't want to use that method, please talk with webmasters@gnu.org about other possible methods. For instance, you can mail them pages to install, if necessary. But that is more work for them, so please use CVS if you can.
Some GNU packages have just simple web pages, but the more information
you provide, the better. So please write as much as you usefully can,
and put all of it on www.gnu.org. However, pages that access
databases (including mail logs and bug tracking) are an exception; set
them up on whatever site is convenient for you, and make the pages on
www.gnu.org link to that site.
Web pages for GNU packages should not include GIF images, since the GNU project avoids GIFs due to patent problems. See Ethical and Philosophical Consideration.
The web pages for the package should include its manuals, in HTML, DVI, Info, PostScript, PDF, plain ASCII, and Texinfo format (source). (All of these can be generated automatically from the Texinfo source using Makeinfo and other programs.) When there is only one manual, put it in a subdirectory called manual; the file manual/index.html should have a link to the manual in each of its forms.
If the package has more than one manual, put each one in a subdirectory of manual, set up index.html in each subdirectory to link to that manual in all its forms, and make manual/index.html link to each manual through its subdirectory.
See the section below for details on a script to make the job of creating all these different formats and index pages easier.
We would like to include links to all these manuals in the page
http://www.gnu.org/manual. Just send mail to
webmasters@gnu.org telling them the name of your package and
asking them to edit http://www.gnu.org/manual, and they will do
so based on the contents of your manual directory.
The script gendocs.sh eases the task of generating the Texinfo documentation output for your web pages section above. It has a companion template file, used as the basis for the html index pages. Both are available from the Texinfo CVS sources:
http://savannah.gnu.org/cgi-bin/viewcvs/texinfo/texinfo/util/gendocs.sh http://savannah.gnu.org/cgi-bin/viewcvs/texinfo/texinfo/util/gendocs_template
Invoke the script like this, in the directory containing the Texinfo source:
gendocs.sh yourmanual "GNU yourmanual manual"
where yourmanual is the short name for your package. The script processes the file yourmanual.texinfo (or .texi or .txi). For example:
cd .../emacs/man
# download gendocs.sh and gendocs_template
gendocs.sh emacs "GNU Emacs manual"
gendocs.sh creates a subdirectory manual/ containing the manual generated in all the standard output formats: Info, HTML, DVI, and so on, as well as the Texinfo source. You then need to move all those files, retaining the subdirectories, into the web pages for your package.
You can specify the option -o outdir to override the name manual. Any previous contents of outdir will be deleted.
The second argument, with the description, is included as part of the
HTML <title> of the overall manual/index.html file. It
should include the name of the package being documented, as shown.
manual/index.html is created by substitution from the file
gendocs_template. (Feel free to modify the generic template
for your own purposes.)
If you have several manuals, you'll need to run this script several times with different arguments, specifying a different output directory with -o each time, and moving all the output to your web page. Then write (by hand) an overall index.html with links to them all. For example:
cd .../texinfo/doc
gendocs.sh -o texinfo texinfo "GNU Texinfo manual"
gendocs.sh -o info info "GNU Info manual"
gendocs.sh -o info-stnd info-stnd "GNU info-stnd manual"
You can set the environment variables MAKEINFO, TEXI2DVI, and DVIPS to control the programs that get executed, and GENDOCS_TEMPLATE_DIR to control where the gendocs_template file is found.
Please email bug reports, enhancement requests, or other correspondence to bug-texinfo@gnu.org.
Since www.gnu.org works through CVS, CVS keywords in your
manual, such as $Log$, need special treatment (even if you
don't happen to maintain your manual in CVS).
If these keywords end up in the generated output as literal strings, they will be expanded. The most robust way to handle this is to turn off keyword expansion for such generated files. For existing files, this is done with:
cvs admin -ko file1 file2 ...
For new files:
cvs add -ko file1 file2 ...
See Keyword substitution.
In Texinfo source, the recommended way to literally specify a “dollar” keyword is:
@w{$}Log$
The @w prevents keyword expansion in the Texinfo source
itself. Also, makeinfo notices the @w and generates
output avoiding the literal keyword string.
The GNU project takes a strong stand for software freedom. Many times, this means you'll need to avoid certain technologies when such technologies conflict with our ethics of software freedom.
Software patents threaten the advancement of free software and freedom to program. For our safety (which includes yours), we try to avoid using algorithms and techniques that we know are patented in the US or elsewhere, unless the patent looks so absurd that we doubt it will be enforced, or we have a suitable patent license allowing release of free software.
Beyond that, sometimes the GNU project takes a strong stand against a particular patented technology in order to encourage everyone to reject it.
For example, the GIF file format is covered by the LZW software patent in the USA. A patent holder has threatened lawsuits against not only developers of software to produce GIFs, but even web sites that contain them.
For this reason, you should not include GIFs in the web pages for your package, nor in the distribution of the package itself. It is ok for a GNU package to support displaying GIFs which will come into play if a user asks it to operate on one. However, it is essential to provide equal or better support for the competing PNG and JPG formats—otherwise, the GNU package would be pressuring users to use GIF format, and that it must not do. More about our stand on GIF is available at http://www.gnu.org/philosophy/gif.html.
Software patents are not the only matter for ethical concern. A GNU package should not recommend use of any non-free program, nor should it require a non-free program (such as a non-free compiler or IDE) to build. Thus, a GNU package cannot be written in a programming language that does not have a free software implementation. Now that GNU/Linux systems are widely available, all GNU packages should function completely with the GNU/Linux system and not require any non-free software to build or function.
A GNU package should not refer the user to any non-free documentation for free software. The need for free documentation to come with free software is now a major focus of the GNU project; to show that we are serious about the need for free documentation, we must not contradict our position by recommending use of documentation that isn't free.
Finally, new issues concerning the ethics of software freedom come up frequently. We ask that GNU maintainers, at least on matters that pertain specifically to their package, stand with the rest of the GNU project when such issues come up.
This chapter explains a couple of issues of terminology which are important for correcting two widespread and important misunderstandings about GNU.
The terms “free software” and “open source” are the slogans of two different movements which differ in their basic philosophy. The Free Software Movement is idealistic, and raises issues of freedom, ethics, principle and what makes for a good society. The Open Source Movement, founded in 1998, studiously avoids such questions. For more explanation, see http://www.gnu.org/philosophy/free-software-for-freedom.html.
The GNU Project is aligned with the Free Software Movement. This doesn't mean that all GNU contributors and maintainers have to agree; your views on these issues are up to you, and you're entitled to express them when speaking for yourself.
However, due to the much greater publicity that the Open Source Movement receives, the GNU Project needs to overcome a widespread mistaken impression that GNU is and always was an activity of the Open Source Movement. For this reason, please use the term “free software,” not “open source,” in GNU software releases, GNU documentation, and announcements and articles that you publish in your role as the maintainer of a GNU package. A reference to the URL given above, to explain the difference, is a useful thing to include as well.
The GNU Project was formed to develop a free Unix-like operating system, GNU. The existence of this system is our major accomplishment. However, the widely used version of the GNU system, in which Linux is used as the kernel, is often called simply “Linux”. As a result, most users don't know about the GNU Project's major accomplishment—or more precisely, they know about it, but don't realize it is the GNU Project's accomplishment and reason for existence. Even people who believe they know the real history often believe that the goal of GNU was to develop “tools” or “utilities.”
To correct this confusion, we have made a years-long effort to distinguish between Linux, the kernel that Linus Torvalds wrote, and GNU/Linux, the operating system that is the combination of GNU and Linux. The resulting increased awareness of what the GNU Project has already done helps every activity of the GNU Project recruit more support and contributors.
Please make this distinction consistently in GNU software releases, GNU documentation, and announcements and articles that you publish in your role as the maintainer of a GNU package. If you want to explain the terminology and its reasons, you can refer to the URL http://www.gnu.org/gnu/linux-and-gnu.html.
Do contrast the GNU system properly speaking to GNU/Linux, you can call it “GNU/Hurd” or “the GNU/Hurd system.” However, when that contrast is not specifically the focus, please call it just “GNU” or “the GNU system.”
When referring to the collection of servers that is the higher level of the GNU kernel, please call it “the Hurd” or “the GNU Hurd.” Note that this uses a space, not a slash.
We would like to recommend using subversions.gnu.org as the CVS
repository for your package, and using ftp.gnu.org as the
standard FTP site. It is ok to use other machines if you wish. If you
use a company's machine to hold the repository for your program, or as
its ftp site, please put this statement in a prominent place on the
site, so as to prevent people from getting the wrong idea about the
relationship between the package and the company:
The programs <list of them> hosted here are free software packages
of the GNU Project, not products of <company name>. We call them
"free software" because you are free to copy and redistribute them,
following the rules stated in the license of each package. For more
information, see http://www.gnu.org/philosophy/free-sw.html.
If you are looking for service or support for GNU software, see
http://www.gnu.org/help/gethelp.html for suggestions of where to ask.
If you would like to contribute to the development of one of these
packages, contact the package maintainer or the bug-reporting address
of the package (which should be listed in the package itself), or look
on www.gnu.org for more information on how to contribute.
The Free Software Directory aims to be a complete list of free software packages, within certain criteria. Every GNU package should be listed there, so please contact bug-directory@gnu.org to ask for information on how to write an entry for your package.
If you want help finding errors in documentation, or help improving the quality of writing, or if you are not a native speaker of English and want help producing good English documentation, you can use the GNU proofreaders mailing list: proofreaders@gnu.org.
But be careful when you use the list, because there are over 200 people on it. If you simply ask everyone on the list to read your work, there will probably be tremendous duplication of effort by the proofreaders, and you will probably get the same errors reported 100 times. This must be avoided.
Also, the people on the list do not want to get a large amount of mail from it. So do not ever ask people on the list to send mail to the list!
Here are a few methods that seem reasonable to use:
Be sure to specify the random choice procedure; otherwise people will probably use a mental procedure that is not really random, such as "pick a part near the middle", and you will not get even coverage.
You can either divide up the work physically, into 20 separate files, or describe a virtual division, such as by sections (if your work has approximately 20 sections). If you do the latter, be sure to be precise about it—for example, do you want the material before the first section heading to count as a section, or not?
alpha.gnu.org, ftp site for test releases: Test Releasesautomake: Distribution tar Filesdiff: Distribution Patchesftp.gnu.org, the GNU ftp site: Distribution on ftp.gnu.orggendocs.sh: Invoking gendocs.shpatch: Distribution Patches