CVS—Concurrent Versions System v1.11.22

This info manual describes how to use and administer CVS version 1.11.22.

CVS and the Real World.
-----------------------

References.
-----------

1. Overview

This chapter is for people who have never used CVS, and perhaps have never used version control software before.

If you are already familiar with CVS and are just trying to learn a particular feature or remember a certain command, you can probably skip everything here.

1.1 What is CVS?

CVS is a version control system. Using it, you can record the history of your source files.

For example, bugs sometimes creep in when software is modified, and you might not detect the bug until a long time after you make the modification. With CVS, you can easily retrieve old versions to see exactly which change caused the bug. This can sometimes be a big help.

You could of course save every version of every file you have ever created. This would however waste an enormous amount of disk space. CVS stores all the versions of a file in a single file in a clever way that only stores the differences between versions.

CVS also helps you if you are part of a group of people working on the same project. It is all too easy to overwrite each others' changes unless you are extremely careful. Some editors, like GNU Emacs, try to make sure that two people never modify the same file at the same time. Unfortunately, if someone is using another editor, that safeguard will not work. CVS solves this problem by insulating the different developers from each other. Every developer works in his own directory, and CVS merges the work when each developer is done.

CVS started out as a bunch of shell scripts written by Dick Grune, posted to the newsgroup comp.sources.unix in the volume 6 release of July, 1986. While no actual code from these shell scripts is present in the current version of CVS much of the CVS conflict resolution algorithms come from them.

In April, 1989, Brian Berliner designed and coded CVS. Jeff Polk later helped Brian with the design of the CVS module and vendor branch support.

You can get CVS in a variety of ways, including free download from the Internet. For more information on downloading CVS and other CVS topics, see:

http://cvs.nongnu.org/

There is a mailing list, known as info-cvs@nongnu.org, devoted to CVS. To subscribe or unsubscribe write to info-cvs-request@nongnu.org. If you prefer a Usenet group, there is a one-way mirror (posts to the email list are usually sent to the news group, but not vice versa) of info-cvs@nongnu.org at news:gnu.cvs.help. The right Usenet group for posts is news:comp.software.config-mgmt which is for CVS discussions (along with other configuration management systems). In the future, it might be possible to create a comp.software.config-mgmt.cvs, but probably only if there is sufficient CVS traffic on news:comp.software.config-mgmt.

You can also subscribe to the bug-cvs@nongnu.org mailing list, described in more detail in Dealing with bugs in CVS or this manual. To subscribe send mail to bug-cvs-request@nongnu.org. There is a two-way Usenet mirror (posts to the Usenet group are usually sent to the email list and vice versa) of bug-cvs@nongnu.org named news:gnu.cvs.bug.

1.2 What is CVS not?

CVS can do a lot of things for you, but it does not try to be everything for everyone.

CVS is not a build system.

Though the structure of your repository and modules file interact with your build system (e.g. `Makefile's), they are essentially independent.

CVS does not dictate how you build anything. It merely stores files for retrieval in a tree structure you devise.

CVS does not dictate how to use disk space in the checked out working directories. If you write your `Makefile's or scripts in every directory so they have to know the relative positions of everything else, you wind up requiring the entire repository to be checked out.

If you modularize your work, and construct a build system that will share files (via links, mounts, VPATH in `Makefile's, etc.), you can arrange your disk usage however you like.

But you have to remember that any such system is a lot of work to construct and maintain. CVS does not address the issues involved.

Of course, you should place the tools created to support such a build system (scripts, `Makefile's, etc) under CVS.

Figuring out what files need to be rebuilt when something changes is, again, something to be handled outside the scope of CVS. One traditional approach is to use make for building, and use some automated tool for generating the dependencies which make uses.

See How your build system interacts with CVS, for more information on doing builds in conjunction with CVS.

CVS is not a substitute for management.

Your managers and project leaders are expected to talk to you frequently enough to make certain you are aware of schedules, merge points, branch names and release dates. If they don't, CVS can't help.

CVS is an instrument for making sources dance to your tune. But you are the piper and the composer. No instrument plays itself or writes its own music.

CVS is not a substitute for developer communication.

When faced with conflicts within a single file, most developers manage to resolve them without too much effort. But a more general definition of “conflict” includes problems too difficult to solve without communication between developers.

CVS cannot determine when simultaneous changes within a single file, or across a whole collection of files, will logically conflict with one another. Its concept of a conflict is purely textual, arising when two changes to the same base file are near enough to spook the merge (i.e. diff3) command.

CVS does not claim to help at all in figuring out non-textual or distributed conflicts in program logic.

For example: Say you change the arguments to function X defined in file `A'. At the same time, someone edits file `B', adding new calls to function X using the old arguments. You are outside the realm of CVS's competence.

Acquire the habit of reading specs and talking to your peers.

CVS does not have change control

Change control refers to a number of things. First of all it can mean bug-tracking, that is being able to keep a database of reported bugs and the status of each one (is it fixed? in what release? has the bug submitter agreed that it is fixed?). For interfacing CVS to an external bug-tracking system, see the `rcsinfo' and `verifymsg' files (see section Reference manual for Administrative files).

Another aspect of change control is keeping track of the fact that changes to several files were in fact changed together as one logical change. If you check in several files in a single cvs commit operation, CVS then forgets that those files were checked in together, and the fact that they have the same log message is the only thing tying them together. Keeping a GNU style `ChangeLog' can help somewhat.

Another aspect of change control, in some systems, is the ability to keep track of the status of each change. Some changes have been written by a developer, others have been reviewed by a second developer, and so on. Generally, the way to do this with CVS is to generate a diff (using cvs diff or diff) and email it to someone who can then apply it using the patch utility. This is very flexible, but depends on mechanisms outside CVS to make sure nothing falls through the cracks.

CVS is not an automated testing program

It should be possible to enforce mandatory use of a test suite using the commitinfo file. I haven't heard a lot about projects trying to do that or whether there are subtle gotchas, however.

CVS does not have a built-in process model

Some systems provide ways to ensure that changes or releases go through various steps, with various approvals as needed. Generally, one can accomplish this with CVS but it might be a little more work. In some cases you'll want to use the `commitinfo', `loginfo', `rcsinfo', or `verifymsg' files, to require that certain steps be performed before cvs will allow a checkin. Also consider whether features such as branches and tags can be used to perform tasks such as doing work in a development tree and then merging certain changes over to a stable tree only once they have been proven.

1.3 A sample session

As a way of introducing CVS, we'll go through a typical work-session using CVS. The first thing to understand is that CVS stores all files in a centralized repository (see section The Repository); this section assumes that a repository is set up.

Suppose you are working on a simple compiler. The source consists of a handful of C files and a `Makefile'. The compiler is called `tc' (Trivial Compiler), and the repository is set up so that there is a module called `tc'.

1.3.1 Getting the source

The first thing you must do is to get your own working copy of the source for `tc'. For this, you use the checkout command:

$ cvs checkout tc

This will create a new directory called `tc' and populate it with the source files.

$ cd tc
$ ls
CVS         Makefile    backend.c   driver.c    frontend.c  parser.c

The `CVS' directory is used internally by CVS. Normally, you should not modify or remove any of the files in it.

You start your favorite editor, hack away at `backend.c', and a couple of hours later you have added an optimization pass to the compiler. A note to RCS and SCCS users: There is no need to lock the files that you want to edit. See section Multiple developers, for an explanation.

1.3.2 Committing your changes

When you have checked that the compiler is still compilable you decide to make a new version of `backend.c'. This will store your new `backend.c' in the repository and make it available to anyone else who is using that same repository.

$ cvs commit backend.c

CVS starts an editor, to allow you to enter a log message. You type in “Added an optimization pass.”, save the temporary file, and exit the editor.

The environment variable $CVSEDITOR determines which editor is started. If $CVSEDITOR is not set, then if the environment variable $EDITOR is set, it will be used. If both $CVSEDITOR and $EDITOR are not set then there is a default which will vary with your operating system, for example vi for unix or notepad for Windows NT/95.

In addition, CVS checks the $VISUAL environment variable. Opinions vary on whether this behavior is desirable and whether future releases of CVS should check $VISUAL or ignore it. You will be OK either way if you make sure that $VISUAL is either unset or set to the same thing as $EDITOR.

When CVS starts the editor, it includes a list of files which are modified. For the CVS client, this list is based on comparing the modification time of the file against the modification time that the file had when it was last gotten or updated. Therefore, if a file's modification time has changed but its contents have not, it will show up as modified. The simplest way to handle this is simply not to worry about it—if you proceed with the commit CVS will detect that the contents are not modified and treat it as an unmodified file. The next update will clue CVS in to the fact that the file is unmodified, and it will reset its stored timestamp so that the file will not show up in future editor sessions.

If you want to avoid starting an editor you can specify the log message on the command line using the `-m' flag instead, like this:

$ cvs commit -m "Added an optimization pass" backend.c

1.3.3 Cleaning up

Before you turn to other tasks you decide to remove your working copy of tc. One acceptable way to do that is of course

$ cd ..
$ rm -r tc

but a better way is to use the release command (see section release—Indicate that a Module is no longer in use):

$ cd ..
$ cvs release -d tc
M driver.c
? tc
You have [1] altered files in this repository.
Are you sure you want to release (and delete) directory `tc': n
** `release' aborted by user choice.

The release command checks that all your modifications have been committed. If history logging is enabled it also makes a note in the history file. See section The history file.

When you use the `-d' flag with release, it also removes your working copy.

In the example above, the release command wrote a couple of lines of output. `? tc' means that the file `tc' is unknown to CVS. That is nothing to worry about: `tc' is the executable compiler, and it should not be stored in the repository. See section Ignoring files via cvsignore, for information about how to make that warning go away. See section release output, for a complete explanation of all possible output from release.

`M driver.c' is more serious. It means that the file `driver.c' has been modified since it was checked out.

The release command always finishes by telling you how many modified files you have in your working copy of the sources, and then asks you for confirmation before deleting any files or making any note in the history file.

You decide to play it safe and answer n RET when release asks for confirmation.

1.3.4 Viewing differences

You do not remember modifying `driver.c', so you want to see what has happened to that file.

$ cd tc
$ cvs diff driver.c

This command runs diff to compare the version of `driver.c' that you checked out with your working copy. When you see the output you remember that you added a command line option that enabled the optimization pass. You check it in, and release the module.

$ cvs commit -m "Added an optimization pass" driver.c
Checking in driver.c;
/usr/local/cvsroot/tc/driver.c,v  <--  driver.c
new revision: 1.2; previous revision: 1.1
done
$ cd ..
$ cvs release -d tc
? tc
You have [0] altered files in this repository.
Are you sure you want to release (and delete) directory `tc': y

2. The Repository

The CVS repository stores a complete copy of all the files and directories which are under version control.

Normally, you never access any of the files in the repository directly. Instead, you use CVS commands to get your own copy of the files into a working directory, and then work on that copy. When you've finished a set of changes, you check (or commit) them back into the repository. The repository then contains the changes which you have made, as well as recording exactly what you changed, when you changed it, and other such information. Note that the repository is not a subdirectory of the working directory, or vice versa; they should be in separate locations.

CVS can access a repository by a variety of means. It might be on the local computer, or it might be on a computer across the room or across the world. To distinguish various ways to access a repository, the repository name can start with an access method. For example, the access method :local: means to access a repository directory, so the repository :local:/usr/local/cvsroot means that the repository is in `/usr/local/cvsroot' on the computer running CVS. For information on other access methods, see Remote repositories.

If the access method is omitted, then if the repository starts with `/', then :local: is assumed. If it does not start with `/' then either :ext: or :server: is assumed. For example, if you have a local repository in `/usr/local/cvsroot', you can use /usr/local/cvsroot instead of :local:/usr/local/cvsroot. But if (under Windows NT, for example) your local repository is `c:\src\cvsroot', then you must specify the access method, as in :local:c:/src/cvsroot.

The repository is split in two parts. `$CVSROOT/CVSROOT' contains administrative files for CVS. The other directories contain the actual user-defined modules.

2.1 Telling CVS where your repository is

There are several ways to tell CVS where to find the repository. You can name the repository on the command line explicitly, with the -d (for "directory") option:

cvs -d /usr/local/cvsroot checkout yoyodyne/tc

Or you can set the $CVSROOT environment variable to an absolute path to the root of the repository, `/usr/local/cvsroot' in this example. To set $CVSROOT, csh and tcsh users should have this line in their `.cshrc' or `.tcshrc' files:

setenv CVSROOT /usr/local/cvsroot

sh and bash users should instead have these lines in their `.profile' or `.bashrc':

CVSROOT=/usr/local/cvsroot
export CVSROOT

A repository specified with -d will override the $CVSROOT environment variable. Once you've checked a working copy out from the repository, it will remember where its repository is (the information is recorded in the `CVS/Root' file in the working copy).

The -d option and the `CVS/Root' file both override the $CVSROOT environment variable. If -d option differs from `CVS/Root', the former is used. Of course, for proper operation they should be two ways of referring to the same repository.

2.2 How data is stored in the repository

For most purposes it isn't important how CVS stores information in the repository. In fact, the format has changed in the past, and is likely to change in the future. Since in almost all cases one accesses the repository via CVS commands, such changes need not be disruptive.

However, in some cases it may be necessary to understand how CVS stores data in the repository, for example you might need to track down CVS locks (see section Several developers simultaneously attempting to run CVS) or you might need to deal with the file permissions appropriate for the repository.

2.2.1 Where files are stored within the repository

The overall structure of the repository is a directory tree corresponding to the directories in the working directory. For example, supposing the repository is in

/usr/local/cvsroot

here is a possible directory tree (showing only the directories):

/usr
 |
 +--local
 |   |
 |   +--cvsroot
 |   |    |
 |   |    +--CVSROOT
          |      (administrative files)
          |
          +--gnu
          |   |
          |   +--diff
          |   |   (source code to GNU diff)
          |   |
          |   +--rcs
          |   |   (source code to RCS)
          |   |
          |   +--cvs
          |       (source code to CVS)
          |
          +--yoyodyne
              |
              +--tc
              |    |
              |    +--man
              |    |
              |    +--testing
              |
              +--(other Yoyodyne software)

With the directories are history files for each file under version control. The name of the history file is the name of the corresponding file with `,v' appended to the end. Here is what the repository for the `yoyodyne/tc' directory might look like:

  $CVSROOT
    |
    +--yoyodyne
    |   |
    |   +--tc
    |   |   |
            +--Makefile,v
            +--backend.c,v
            +--driver.c,v
            +--frontend.c,v
            +--parser.c,v
            +--man
            |    |
            |    +--tc.1,v
            |
            +--testing
                 |
                 +--testpgm.t,v
                 +--test2.t,v

The history files contain, among other things, enough information to recreate any revision of the file, a log of all commit messages and the user-name of the person who committed the revision. The history files are known as RCS files, because the first program to store files in that format was a version control system known as RCS. For a full description of the file format, see the man page rcsfile(5), distributed with RCS, or the file `doc/RCSFILES' in the CVS source distribution. This file format has become very common—many systems other than CVS or RCS can at least import history files in this format.

The RCS files used in CVS differ in a few ways from the standard format. The biggest difference is magic branches; for more information see Magic branch numbers. Also in CVS the valid tag names are a subset of what RCS accepts; for CVS's rules see Tags–Symbolic revisions.

2.2.2 File permissions

All `,v' files are created read-only, and you should not change the permission of those files. The directories inside the repository should be writable by the persons that have permission to modify the files in each directory. This normally means that you must create a UNIX group (see group(5)) consisting of the persons that are to edit the files in a project, and set up the repository so that it is that group that owns the directory. (On some systems, you also need to set the set-group-ID-on-execution bit on the repository directories (see chmod(1)) so that newly-created files and directories get the group-ID of the parent directory rather than that of the current process.)

This means that you can only control access to files on a per-directory basis.

Note that users must also have write access to check out files, because CVS needs to create lock files (see section Several developers simultaneously attempting to run CVS). You can use LockDir in CVSROOT/config to put the lock files somewhere other than in the repository if you want to allow read-only access to some directories (see section The CVSROOT/config configuration file).

Also note that users must have write access to the `CVSROOT/val-tags' file. CVS uses it to keep track of what tags are valid tag names (it is sometimes updated when tags are used, as well as when they are created).

Each RCS file will be owned by the user who last checked it in. This has little significance; what really matters is who owns the directories.

CVS tries to set up reasonable file permissions for new directories that are added inside the tree, but you must fix the permissions manually when a new directory should have different permissions than its parent directory. If you set the CVSUMASK environment variable that will control the file permissions which CVS uses in creating directories and/or files in the repository. CVSUMASK does not affect the file permissions in the working directory; such files have the permissions which are typical for newly created files, except that sometimes CVS creates them read-only (see the sections on watches, Telling CVS to watch certain files; -r, Global options; or CVSREAD, All environment variables which affect CVS).

Note that using the client/server CVS (see section Remote repositories), there is no good way to set CVSUMASK; the setting on the client machine has no effect. If you are connecting with rsh, you can set CVSUMASK in `.bashrc' or `.cshrc', as described in the documentation for your operating system. This behavior might change in future versions of CVS; do not rely on the setting of CVSUMASK on the client having no effect.

Using pserver, you will generally need stricter permissions on the CVSROOT directory and directories above it in the tree; see Security considerations with password authentication.

Some operating systems have features which allow a particular program to run with the ability to perform operations which the caller of the program could not. For example, the set user ID (setuid) or set group ID (setgid) features of unix or the installed image feature of VMS. CVS was not written to use such features and therefore attempting to install CVS in this fashion will provide protection against only accidental lapses; anyone who is trying to circumvent the measure will be able to do so, and depending on how you have set it up may gain access to more than just CVS. You may wish to instead consider pserver. It shares some of the same attributes, in terms of possibly providing a false sense of security or opening security holes wider than the ones you are trying to fix, so read the documentation on pserver security carefully if you are considering this option (Security considerations with password authentication).

2.2.3 File Permission issues specific to Windows

Some file permission issues are specific to Windows operating systems (Windows 95, Windows NT, and presumably future operating systems in this family. Some of the following might apply to OS/2 but I'm not sure).

If you are using local CVS and the repository is on a networked file system which is served by the Samba SMB server, some people have reported problems with permissions. Enabling WRITE=YES in the samba configuration is said to fix/workaround it. Disclaimer: I haven't investigated enough to know the implications of enabling that option, nor do I know whether there is something which CVS could be doing differently in order to avoid the problem. If you find something out, please let us know as described in Dealing with bugs in CVS or this manual.

2.2.4 The attic

You will notice that sometimes CVS stores an RCS file in the Attic. For example, if the CVSROOT is `/usr/local/cvsroot' and we are talking about the file `backend.c' in the directory `yoyodyne/tc', then the file normally would be in

/usr/local/cvsroot/yoyodyne/tc/backend.c,v

but if it goes in the attic, it would be in

/usr/local/cvsroot/yoyodyne/tc/Attic/backend.c,v

instead. It should not matter from a user point of view whether a file is in the attic; CVS keeps track of this and looks in the attic when it needs to. But in case you want to know, the rule is that the RCS file is stored in the attic if and only if the head revision on the trunk has state dead. A dead state means that file has been removed, or never added, for that revision. For example, if you add a file on a branch, it will have a trunk revision in dead state, and a branch revision in a non-dead state.

2.2.5 The CVS directory in the repository

The `CVS' directory in each repository directory contains information such as file attributes (in a file called `CVS/fileattr'. In the future additional files may be added to this directory, so implementations should silently ignore additional files.

This behavior is implemented only by CVS 1.7 and later; for details see Using watches with old versions of CVS.

The format of the `fileattr' file is a series of entries of the following form (where `{' and `}' means the text between the braces can be repeated zero or more times):

ent-type filename <tab> attrname = attrval {; attrname = attrval} <linefeed>

ent-type is `F' for a file, in which case the entry specifies the attributes for that file.

ent-type is `D', and filename empty, to specify default attributes to be used for newly added files.

Other ent-type are reserved for future expansion. CVS 1.9 and older will delete them any time it writes file attributes. CVS 1.10 and later will preserve them.

Note that the order of the lines is not significant; a program writing the fileattr file may rearrange them at its convenience.

There is currently no way of quoting tabs or line feeds in the filename, `=' in attrname, `;' in attrval, etc. Note: some implementations also don't handle a NUL character in any of the fields, but implementations are encouraged to allow it.

By convention, attrname starting with `_' is for an attribute given special meaning by CVS; other attrnames are for user-defined attributes (or will be, once implementations start supporting user-defined attributes).

Built-in attributes:

_watched

Present means the file is watched and should be checked out read-only.

_watchers

Users with watches for this file. Value is watcher > type { , watcher > type } where watcher is a username, and type is zero or more of edit,unedit,commit separated by `+' (that is, nothing if none; there is no "none" or "all" keyword).

_editors

Users editing this file. Value is editor > val { , editor > val } where editor is a username, and val is time+hostname+pathname, where time is when the cvs edit command (or equivalent) happened, and hostname and pathname are for the working directory.

Example:

Ffile1 _watched=;_watchers=joe>edit,mary>commit
Ffile2 _watched=;_editors=sue>8 Jan 1975+workstn1+/home/sue/cvs
D _watched=

means that the file `file1' should be checked out read-only. Furthermore, joe is watching for edits and mary is watching for commits. The file `file2' should be checked out read-only; sue started editing it on 8 Jan 1975 in the directory `/home/sue/cvs' on the machine workstn1. Future files which are added should be checked out read-only. To represent this example here, we have shown a space after `D', `Ffile1', and `Ffile2', but in fact there must be a single tab character there and no spaces.

2.2.6 CVS locks in the repository

For an introduction to CVS locks focusing on user-visible behavior, see Several developers simultaneously attempting to run CVS. The following section is aimed at people who are writing tools which want to access a CVS repository without interfering with other tools accessing the same repository. If you find yourself confused by concepts described here, like read lock, write lock, and deadlock, you might consult the literature on operating systems or databases.

Any file in the repository with a name starting with `#cvs.rfl.' is a read lock. Any file in the repository with a name starting with `#cvs.wfl' is a write lock. Old versions of CVS (before CVS 1.5) also created files with names starting with `#cvs.tfl', but they are not discussed here. The directory `#cvs.lock' serves as a master lock. That is, one must obtain this lock first before creating any of the other locks.

To obtain a read lock, first create the `#cvs.lock' directory. This operation must be atomic (which should be true for creating a directory under most operating systems). If it fails because the directory already existed, wait for a while and try again. After obtaining the `#cvs.lock' lock, create a file whose name is `#cvs.rfl.' followed by information of your choice (for example, hostname and process identification number). Then remove the `#cvs.lock' directory to release the master lock. Then proceed with reading the repository. When you are done, remove the `#cvs.rfl' file to release the read lock.

To obtain a write lock, first create the `#cvs.lock' directory, as with read locks. Then check that there are no files whose names start with `#cvs.rfl.'. If there are, remove `#cvs.lock', wait for a while, and try again. If there are no readers, then create a file whose name is `#cvs.wfl' followed by information of your choice (for example, hostname and process identification number). Hang on to the `#cvs.lock' lock. Proceed with writing the repository. When you are done, first remove the `#cvs.wfl' file and then the `#cvs.lock' directory. Note that unlike the `#cvs.rfl' file, the `#cvs.wfl' file is just informational; it has no effect on the locking operation beyond what is provided by holding on to the `#cvs.lock' lock itself.

Note that each lock (write lock or read lock) only locks a single directory in the repository, including `Attic' and `CVS' but not including subdirectories which represent other directories under version control. To lock an entire tree, you need to lock each directory (note that if you fail to obtain any lock you need, you must release the whole tree before waiting and trying again, to avoid deadlocks).

Note also that CVS expects write locks to control access to individual `foo,v' files. RCS has a scheme where the `,foo,' file serves as a lock, but CVS does not implement it and so taking out a CVS write lock is recommended. See the comments at rcs_internal_lockfile in the CVS source code for further discussion/rationale.

2.2.7 How files are stored in the CVSROOT directory

The `$CVSROOT/CVSROOT' directory contains the various administrative files. In some ways this directory is just like any other directory in the repository; it contains RCS files whose names end in `,v', and many of the CVS commands operate on it the same way. However, there are a few differences.

For each administrative file, in addition to the RCS file, there is also a checked out copy of the file. For example, there is an RCS file `loginfo,v' and a file `loginfo' which contains the latest revision contained in `loginfo,v'. When you check in an administrative file, CVS should print

cvs commit: Rebuilding administrative file database

and update the checked out copy in `$CVSROOT/CVSROOT'. If it does not, there is something wrong (see section Dealing with bugs in CVS or this manual). To add your own files to the files to be updated in this fashion, you can add them to the `checkoutlist' administrative file (see section The checkoutlist file).

By default, the `modules' file behaves as described above. If the modules file is very large, storing it as a flat text file may make looking up modules slow (I'm not sure whether this is as much of a concern now as when CVS first evolved this feature; I haven't seen benchmarks). Therefore, by making appropriate edits to the CVS source code one can store the modules file in a database which implements the ndbm interface, such as Berkeley db or GDBM. If this option is in use, then the modules database will be stored in the files `modules.db', `modules.pag', and/or `modules.dir'.

For information on the meaning of the various administrative files, see Reference manual for Administrative files.

2.3 How data is stored in the working directory

While we are discussing CVS internals which may become visible from time to time, we might as well talk about what CVS puts in the `CVS' directories in the working directories. As with the repository, CVS handles this information and one can usually access it via CVS commands. But in some cases it may be useful to look at it, and other programs, such as the jCVS graphical user interface or the VC package for emacs, may need to look at it. Such programs should follow the recommendations in this section if they hope to be able to work with other programs which use those files, including future versions of the programs just mentioned and the command-line CVS client.

The `CVS' directory contains several files. Programs which are reading this directory should silently ignore files which are in the directory but which are not documented here, to allow for future expansion.

The files are stored according to the text file convention for the system in question. This means that working directories are not portable between systems with differing conventions for storing text files. This is intentional, on the theory that the files being managed by CVS probably will not be portable between such systems either.

`Root'

This file contains the current CVS root, as described in Telling CVS where your repository is.

`Repository'

This file contains the directory within the repository which the current directory corresponds with. It can be either an absolute pathname or a relative pathname; CVS has had the ability to read either format since at least version 1.3 or so. The relative pathname is relative to the root, and is the more sensible approach, but the absolute pathname is quite common and implementations should accept either. For example, after the command

cvs -d :local:/usr/local/cvsroot checkout yoyodyne/tc

`Root' will contain

:local:/usr/local/cvsroot

and `Repository' will contain either

/usr/local/cvsroot/yoyodyne/tc

or

yoyodyne/tc

If the particular working directory does not correspond to a directory in the repository, then `Repository' should contain `CVSROOT/Emptydir'.

`Entries'

This file lists the files and directories in the working directory. The first character of each line indicates what sort of line it is. If the character is unrecognized, programs reading the file should silently skip that line, to allow for future expansion.

If the first character is `/', then the format is:

/name/revision/timestamp[+conflict]/options/tagdate

where `[' and `]' are not part of the entry, but instead indicate that the `+' and conflict marker are optional. name is the name of the file within the directory. revision is the revision that the file in the working derives from, or `0' for an added file, or `-' followed by a revision for a removed file. timestamp is the timestamp of the file at the time that CVS created it; if the timestamp differs with the actual modification time of the file it means the file has been modified. It is stored in the format used by the ISO C asctime() function (for example, `Sun Apr 7 01:29:26 1996'). One may write a string which is not in that format, for example, `Result of merge', to indicate that the file should always be considered to be modified. This is not a special case; to see whether a file is modified a program should take the timestamp of the file and simply do a string compare with timestamp. If there was a conflict, conflict can be set to the modification time of the file after the file has been written with conflict markers (see section Conflicts example). Thus if conflict is subsequently the same as the actual modification time of the file it means that the user has obviously not resolved the conflict. options contains sticky options (for example `-kb' for a binary file). tagdate contains `T' followed by a tag name, or `D' for a date, followed by a sticky tag or date. Note that if timestamp contains a pair of timestamps separated by a space, rather than a single timestamp, you are dealing with a version of CVS earlier than CVS 1.5 (not documented here).

The timezone on the timestamp in CVS/Entries (local or universal) should be the same as the operating system stores for the timestamp of the file itself. For example, on Unix the file's timestamp is in universal time (UT), so the timestamp in CVS/Entries should be too. On VMS, the file's timestamp is in local time, so CVS on VMS should use local time. This rule is so that files do not appear to be modified merely because the timezone changed (for example, to or from summer time).

If the first character of a line in `Entries' is `D', then it indicates a subdirectory. `D' on a line all by itself indicates that the program which wrote the `Entries' file does record subdirectories (therefore, if there is such a line and no other lines beginning with `D', one knows there are no subdirectories). Otherwise, the line looks like:

D/name/filler1/filler2/filler3/filler4

where name is the name of the subdirectory, and all the filler fields should be silently ignored, for future expansion. Programs which modify Entries files should preserve these fields.

The lines in the `Entries' file can be in any order.

`Entries.Log'

This file does not record any information beyond that in `Entries', but it does provide a way to update the information without having to rewrite the entire `Entries' file, including the ability to preserve the information even if the program writing `Entries' and `Entries.Log' abruptly aborts. Programs which are reading the `Entries' file should also check for `Entries.Log'. If the latter exists, they should read `Entries' and then apply the changes mentioned in `Entries.Log'. After applying the changes, the recommended practice is to rewrite `Entries' and then delete `Entries.Log'. The format of a line in `Entries.Log' is a single character command followed by a space followed by a line in the format specified for a line in `Entries'. The single character command is `A' to indicate that the entry is being added, `R' to indicate that the entry is being removed, or any other character to indicate that the entire line in `Entries.Log' should be silently ignored (for future expansion). If the second character of the line in `Entries.Log' is not a space, then it was written by an older version of CVS (not documented here).

Programs which are writing rather than reading can safely ignore `Entries.Log' if they so choose.

`Entries.Backup'

This is a temporary file. Recommended usage is to write a new entries file to `Entries.Backup', and then to rename it (atomically, where possible) to `Entries'.

`Entries.Static'

The only relevant thing about this file is whether it exists or not. If it exists, then it means that only part of a directory was gotten and CVS will not create additional files in that directory. To clear it, use the update command with the `-d' option, which will get the additional files and remove `Entries.Static'.

`Tag'

This file contains per-directory sticky tags or dates. The first character is `T' for a branch tag, `N' for a non-branch tag, or `D' for a date, or another character to mean the file should be silently ignored, for future expansion. This character is followed by the tag or date. Note that per-directory sticky tags or dates are used for things like applying to files which are newly added; they might not be the same as the sticky tags or dates on individual files. For general information on sticky tags and dates, see Sticky tags.

`Notify'

This file stores notifications (for example, for edit or unedit) which have not yet been sent to the server. Its format is not yet documented here.

`Notify.tmp'

This file is to `Notify' as `Entries.Backup' is to `Entries'. That is, to write `Notify', first write the new contents to `Notify.tmp' and then (atomically where possible), rename it to `Notify'.

`Base'

If watches are in use, then an edit command stores the original copy of the file in the `Base' directory. This allows the unedit command to operate even if it is unable to communicate with the server.

`Baserev'

The file lists the revision for each of the files in the `Base' directory. The format is:

Bname/rev/expansion

where expansion should be ignored, to allow for future expansion.

`Baserev.tmp'

This file is to `Baserev' as `Entries.Backup' is to `Entries'. That is, to write `Baserev', first write the new contents to `Baserev.tmp' and then (atomically where possible), rename it to `Baserev'.

`Template'

This file contains the template specified by the `rcsinfo' file (see section Rcsinfo). It is only used by the client; the non-client/server CVS consults `rcsinfo' directly.

2.4 The administrative files

The directory `$CVSROOT/CVSROOT' contains some administrative files. See section Reference manual for Administrative files, for a complete description. You can use CVS without any of these files, but some commands work better when at least the `modules' file is properly set up.

The most important of these files is the `modules' file. It defines all modules in the repository. This is a sample `modules' file.

CVSROOT         CVSROOT
modules         CVSROOT modules
cvs             gnu/cvs
rcs             gnu/rcs
diff            gnu/diff
tc              yoyodyne/tc

The `modules' file is line oriented. In its simplest form each line contains the name of the module, whitespace, and the directory where the module resides. The directory is a path relative to $CVSROOT. The last four lines in the example above are examples of such lines.

The line that defines the module called `modules' uses features that are not explained here. See section The modules file, for a full explanation of all the available features.

2.4.1 Editing administrative files

You edit the administrative files in the same way that you would edit any other module. Use `cvs checkout CVSROOT' to get a working copy, edit it, and commit your changes in the normal way.

It is possible to commit an erroneous administrative file. You can often fix the error and check in a new revision, but sometimes a particularly bad error in the administrative file makes it impossible to commit new revisions. If and when this happens, you can correct the problem by temporarily copying a corrected administrative file directly into the $CVSROOT/CVSROOT repository directory, then committing the same correction via a checkout of the `CVSROOT' module. It is important that the correction also be made via the checked out copy, or the next checkout and commit to the <code>CVSROOT</code> module will overwrite the correction that was copied directly into the repository, possibly breaking things in such a way as to prevent commits again.

2.5 Multiple repositories

In some situations it is a good idea to have more than one repository, for instance if you have two development groups that work on separate projects without sharing any code. All you have to do to have several repositories is to specify the appropriate repository, using the CVSROOT environment variable, the `-d' option to CVS, or (once you have checked out a working directory) by simply allowing CVS to use the repository that was used to check out the working directory (see section Telling CVS where your repository is).

The big advantage of having multiple repositories is that they can reside on different servers. With CVS version 1.10, a single command cannot recurse into directories from different repositories. With development versions of CVS, you can check out code from multiple servers into your working directory. CVS will recurse and handle all the details of making connections to as many server machines as necessary to perform the requested command. Here is an example of how to set up a working directory:

cvs -d server1:/cvs co dir1
cd dir1
cvs -d server2:/root co sdir
cvs update

The cvs co commands set up the working directory, and then the cvs update command will contact server2, to update the dir1/sdir subdirectory, and server1, to update everything else.

2.6 Creating a repository

This section describes how to set up a CVS repository for any sort of access method. After completing the setup described in this section, you should be able to access your CVS repository immediately via the local access method and several remote access methods. For more information on setting up remote access to the repository you create in this section, please read the section on See section Remote repositories.

To set up a CVS repository, first choose the machine and disk on which you want to store the revision history of the source files. CPU and memory requirements are modest, so most machines should be adequate. For details see Server requirements.

To estimate disk space requirements, if you are importing RCS files from another system, the size of those files is the approximate initial size of your repository, or if you are starting without any version history, a rule of thumb is to allow for the server approximately three times the size of the code to be under CVS for the repository (you will eventually outgrow this, but not for a while). On the machines on which the developers will be working, you'll want disk space for approximately one working directory for each developer (either the entire tree or a portion of it, depending on what each developer uses).

The repository should be accessible (directly or via a networked file system) from all machines which want to use CVS in server or local mode; the client machines need not have any access to it other than via the CVS protocol. It is not possible to use CVS to read from a repository which one only has read access to; CVS needs to be able to create lock files (see section Several developers simultaneously attempting to run CVS).

To create a repository, run the cvs init command. It will set up an empty repository in the CVS root specified in the usual way (see section The Repository). For example,

cvs -d /usr/local/cvsroot init

cvs init is careful to never overwrite any existing files in the repository, so no harm is done if you run cvs init on an already set-up repository.

cvs init will enable history logging; if you don't want that, remove the history file after running cvs init. See section The history file.

2.7 Backing up a repository

There is nothing particularly magical about the files in the repository; for the most part it is possible to back them up just like any other files. However, there are a few issues to consider.

The first is that to be paranoid, one should either not use CVS during the backup, or have the backup program lock CVS while doing the backup. To not use CVS, you might forbid logins to machines which can access the repository, turn off your CVS server, or similar mechanisms. The details would depend on your operating system and how you have CVS set up. To lock CVS, you would create `#cvs.rfl' locks in each repository directory. See Several developers simultaneously attempting to run CVS, for more on CVS locks. Having said all this, if you just back up without any of these precautions, the results are unlikely to be particularly dire. Restoring from backup, the repository might be in an inconsistent state, but this would not be particularly hard to fix manually.

When you restore a repository from backup, assuming that changes in the repository were made after the time of the backup, working directories which were not affected by the failure may refer to revisions which no longer exist in the repository. Trying to run CVS in such directories will typically produce an error message. One way to get those changes back into the repository is as follows:

• Get a new working directory.
• Copy the files from the working directory from before the failure over to the new working directory (do not copy the contents of the `CVS' directories, of course).
• Working in the new working directory, use commands such as cvs update and cvs diff to figure out what has changed, and then when you are ready, commit the changes into the repository.

2.8 Moving a repository

Just as backing up the files in the repository is pretty much like backing up any other files, if you need to move a repository from one place to another it is also pretty much like just moving any other collection of files.

The main thing to consider is that working directories point to the repository. The simplest way to deal with a moved repository is to just get a fresh working directory after the move. Of course, you'll want to make sure that the old working directory had been checked in before the move, or you figured out some other way to make sure that you don't lose any changes. If you really do want to reuse the existing working directory, it should be possible with manual surgery on the `CVS/Repository' files. You can see How data is stored in the working directory, for information on the `CVS/Repository' and `CVS/Root' files, but unless you are sure you want to bother, it probably isn't worth it.

2.9 Remote repositories

Your working copy of the sources can be on a different machine than the repository. Using CVS in this manner is known as client/server operation. You run CVS on a machine which can mount your working directory, known as the client, and tell it to communicate to a machine which can mount the repository, known as the server. Generally, using a remote repository is just like using a local one, except that the format of the repository name is:

[:method:][[user][:password]@]hostname[:[port]]/path/to/repository

Specifying a password in the repository name is not recommended during checkout, since this will cause CVS to store a cleartext copy of the password in each created directory. cvs login first instead (see section Using the client with password authentication).

The details of exactly what needs to be set up depend on how you are connecting to the server.

If method is not specified, and the repository name contains `:', then the default is ext or server, depending on your platform; both are described in Connecting with rsh.

2.9.1 Server requirements

The quick answer to what sort of machine is suitable as a server is that requirements are modest—a server with 32M of memory or even less can handle a fairly large source tree with a fair amount of activity.

The real answer, of course, is more complicated. Estimating the known areas of large memory consumption should be sufficient to estimate memory requirements. There are two such areas documented here; other memory consumption should be small by comparison (if you find that is not the case, let us know, as described in Dealing with bugs in CVS or this manual, so we can update this documentation).

The first area of big memory consumption is large checkouts, when using the CVS server. The server consists of two processes for each client that it is serving. Memory consumption on the child process should remain fairly small. Memory consumption on the parent process, particularly if the network connection to the client is slow, can be expected to grow to slightly more than the size of the sources in a single directory, or two megabytes, whichever is larger.

Multiplying the size of each CVS server by the number of servers which you expect to have active at one time should give an idea of memory requirements for the server. For the most part, the memory consumed by the parent process probably can be swap space rather than physical memory.

The second area of large memory consumption is diff, when checking in large files. This is required even for binary files. The rule of thumb is to allow about ten times the size of the largest file you will want to check in, although five times may be adequate. For example, if you want to check in a file which is 10 megabytes, you should have 100 megabytes of memory on the machine doing the checkin (the server machine for client/server, or the machine running CVS for non-client/server). This can be swap space rather than physical memory. Because the memory is only required briefly, there is no particular need to allow memory for more than one such checkin at a time.

Resource consumption for the client is even more modest—any machine with enough capacity to run the operating system in question should have little trouble.

For information on disk space requirements, see Creating a repository.

2.9.2 Connecting with rsh

CVS uses the `rsh' protocol to perform these operations, so the remote user host needs to have a `.rhosts' file which grants access to the local user.

For example, suppose you are the user `mozart' on the local machine `toe.example.com', and the server machine is `faun.example.org'. On faun, put the following line into the file `.rhosts' in `bach''s home directory:

toe.example.com  mozart

Then test that `rsh' is working with

rsh -l bach faun.example.org 'echo $PATH'

Next you have to make sure that rsh will be able to find the server. Make sure that the path which rsh printed in the above example includes the directory containing a program named cvs which is the server. You need to set the path in `.bashrc', `.cshrc', etc., not `.login' or `.profile'. Alternately, you can set the environment variable CVS_SERVER on the client machine to the filename of the server you want to use, for example `/usr/local/bin/cvs-1.6'.

There is no need to edit `inetd.conf' or start a CVS server daemon.

There are two access methods that you use in CVSROOT for rsh. :server: specifies an internal rsh client, which is supported only by some CVS ports. :ext: specifies an external rsh program. By default this is rsh but you may set the CVS_RSH environment variable to invoke another program which can access the remote server (for example, remsh on HP-UX 9 because rsh is something different). It must be a program which can transmit data to and from the server without modifying it; for example the Windows NT rsh is not suitable since it by default translates between CRLF and LF. The OS/2 CVS port has a hack to pass `-b' to rsh to get around this, but since this could potentially cause problems for programs other than the standard rsh, it may change in the future. If you set CVS_RSH to SSH or some other rsh replacement, the instructions in the rest of this section concerning `.rhosts' and so on are likely to be inapplicable; consult the documentation for your rsh replacement.

Continuing our example, supposing you want to access the module `foo' in the repository `/usr/local/cvsroot/', on machine `faun.example.org', you are ready to go:

cvs -d :ext:bach@faun.example.org:/usr/local/cvsroot checkout foo

(The `bach@' can be omitted if the username is the same on both the local and remote hosts.)

2.9.3 Direct connection with password authentication

The CVS client can also connect to the server using a password protocol. This is particularly useful if using rsh is not feasible (for example, the server is behind a firewall), and Kerberos also is not available.

To use this method, it is necessary to make some adjustments on both the server and client sides.

2.9.3.1 Setting up the server for password authentication

First of all, you probably want to tighten the permissions on the `$CVSROOT' and `$CVSROOT/CVSROOT' directories. See Security considerations with password authentication, for more details.

On the server side, the file `/etc/inetd.conf' needs to be edited so inetd knows to run the command cvs pserver when it receives a connection on the right port. By default, the port number is 2401; it would be different if your client were compiled with CVS_AUTH_PORT defined to something else, though. This can also be specified in the CVSROOT variable (see section Remote repositories) or overridden with the CVS_CLIENT_PORT environment variable (see section All environment variables which affect CVS).

If your inetd allows raw port numbers in `/etc/inetd.conf', then the following (all on a single line in `inetd.conf') should be sufficient:

2401  stream  tcp  nowait  root  /usr/local/bin/cvs
cvs -f --allow-root=/usr/cvsroot pserver

(You could also use the `-T' option to specify a temporary directory.)

The `--allow-root' option specifies the allowable CVSROOT directory. Clients which attempt to use a different CVSROOT directory will not be allowed to connect. If there is more than one CVSROOT directory which you want to allow, repeat the option. (Unfortunately, many versions of inetd have very small limits on the number of arguments and/or the total length of the command. The usual solution to this problem is to have inetd run a shell script which then invokes CVS with the necessary arguments.)

If your inetd wants a symbolic service name instead of a raw port number, then put this in `/etc/services':

cvspserver      2401/tcp

and put cvspserver instead of 2401 in `inetd.conf'.

If your system uses xinetd instead of inetd, the procedure is slightly different. Create a file called `/etc/xinetd.d/cvspserver' containing the following:

service cvspserver
{
   port        = 2401
   socket_type = stream
   protocol    = tcp
   wait        = no
   user        = root
   passenv     = PATH
   server      = /usr/local/bin/cvs
   server_args = -f --allow-root=/usr/cvsroot pserver
}

(If cvspserver is defined in `/etc/services', you can omit the port line.)

Once the above is taken care of, restart your inetd, or do whatever is necessary to force it to reread its initialization files.

If you are having trouble setting this up, see Trouble making a connection to a CVS server.

Because the client stores and transmits passwords in cleartext (almost—see Security considerations with password authentication, for details), a separate CVS password file is generally used, so people don't compromise their regular passwords when they access the repository. This file is `$CVSROOT/CVSROOT/passwd' (see section The administrative files). It uses a colon-separated format, similar to `/etc/passwd' on Unix systems, except that it has fewer fields: CVS username, optional password, and an optional system username for CVS to run as if authentication succeeds. Here is an example `passwd' file with five entries:

anonymous:
bach:ULtgRLXo7NRxs
spwang:1sOp854gDF3DY
melissa:tGX1fS8sun6rY:pubcvs
qproj:XR4EZcEs0szik:pubcvs

(The passwords are encrypted according to the standard Unix crypt() function, so it is possible to paste in passwords directly from regular Unix `/etc/passwd' files.)

The first line in the example will grant access to any CVS client attempting to authenticate as user anonymous, no matter what password they use, including an empty password. (This is typical for sites granting anonymous read-only access; for information on how to do the "read-only" part, see Read-only repository access.)

The second and third lines will grant access to bach and spwang if they supply their respective plaintext passwords.

The fourth line will grant access to melissa, if she supplies the correct password, but her CVS operations will actually run on the server side under the system user pubcvs. Thus, there need not be any system user named melissa, but there must be one named pubcvs.

The fifth line shows that system user identities can be shared: any client who successfully authenticates as qproj will actually run as pubcvs, just as melissa does. That way you could create a single, shared system user for each project in your repository, and give each developer their own line in the `$CVSROOT/CVSROOT/passwd' file. The CVS username on each line would be different, but the system username would be the same. The reason to have different CVS usernames is that CVS will log their actions under those names: when melissa commits a change to a project, the checkin is recorded in the project's history under the name melissa, not pubcvs. And the reason to have them share a system username is so that you can arrange permissions in the relevant area of the repository such that only that account has write-permission there.

If the system-user field is present, all password-authenticated CVS commands run as that user; if no system user is specified, CVS simply takes the CVS username as the system username and runs commands as that user. In either case, if there is no such user on the system, then the CVS operation will fail (regardless of whether the client supplied a valid password).

The password and system-user fields can both be omitted (and if the system-user field is omitted, then also omit the colon that would have separated it from the encrypted password). For example, this would be a valid `$CVSROOT/CVSROOT/passwd' file:

anonymous::pubcvs
fish:rKa5jzULzmhOo:kfogel
sussman:1sOp854gDF3DY

When the password field is omitted or empty, then the client's authentication attempt will succeed with any password, including the empty string. However, the colon after the CVS username is always necessary, even if the password is empty.

CVS can also fall back to use system authentication. When authenticating a password, the server first checks for the user in the `$CVSROOT/CVSROOT/passwd' file. If it finds the user, it will use that entry for authentication as described above. But if it does not find the user, or if the CVS `passwd' file does not exist, then the server can try to authenticate the username and password using the operating system's user-lookup routines (this "fallback" behavior can be disabled by setting SystemAuth=no in the CVS `config' file, see section The CVSROOT/config configuration file). Be aware, however, that falling back to system authentication might be a security risk: CVS operations would then be authenticated with that user's regular login password, and the password flies across the network in plaintext. See Security considerations with password authentication for more on this.

Right now, the only way to put a password in the CVS `passwd' file is to paste it there from somewhere else. Someday, there may be a cvs passwd command.

Unlike many of the files in `$CVSROOT/CVSROOT', it is normal to edit the `passwd' file in-place, rather than via CVS. This is because of the possible security risks of having the `passwd' file checked out to people's working copies. If you do want to include the `passwd' file in checkouts of `$CVSROOT/CVSROOT', see The checkoutlist file.

2.9.3.2 Using the client with password authentication

To run a CVS command on a remote repository via the password-authenticating server, one specifies the pserver protocol, optional username, repository host, an optional port number, and path to the repository. For example:

cvs -d :pserver:faun.example.org:/usr/local/cvsroot checkout someproj

or

CVSROOT=:pserver:bach@faun.example.org:2401/usr/local/cvsroot
cvs checkout someproj

However, unless you're connecting to a public-access repository (i.e., one where that username doesn't require a password), you'll need to supply a password or log in first. Logging in verifies your password with the repository and stores it in a file. It's done with the login command, which will prompt you interactively for the password if you didn't supply one as part of $CVSROOT:

cvs -d :pserver:bach@faun.example.org:/usr/local/cvsroot login
CVS password:

or

cvs -d :pserver:bach:p4ss30rd@faun.example.org:/usr/local/cvsroot login

After you enter the password, CVS verifies it with the server. If the verification succeeds, then that combination of username, host, repository, and password is permanently recorded, so future transactions with that repository won't require you to run cvs login. (If verification fails, CVS will exit complaining that the password was incorrect, and nothing will be recorded.)

The records are stored, by default, in the file `$HOME/.cvspass'. That file's format is human-readable, and to a degree human-editable, but note that the passwords are not stored in cleartext—they are trivially encoded to protect them from "innocent" compromise (i.e., inadvertent viewing by a system administrator or other non-malicious person).

You can change the default location of this file by setting the CVS_PASSFILE environment variable. If you use this variable, make sure you set it before cvs login is run. If you were to set it after running cvs login, then later CVS commands would be unable to look up the password for transmission to the server.

Once you have logged in, all CVS commands using that remote repository and username will authenticate with the stored password. So, for example

cvs -d :pserver:bach@faun.example.org:/usr/local/cvsroot checkout foo

should just work (unless the password changes on the server side, in which case you'll have to re-run cvs login).

Note that if the `:pserver:' were not present in the repository specification, CVS would assume it should use rsh to connect with the server instead (see section Connecting with rsh).

Of course, once you have a working copy checked out and are running CVS commands from within it, there is no longer any need to specify the repository explicitly, because CVS can deduce the repository from the working copy's `CVS' subdirectory.

The password for a given remote repository can be removed from the CVS_PASSFILE by using the cvs logout command.

2.9.3.3 Security considerations with password authentication

The passwords are stored on the client side in a trivial encoding of the cleartext, and transmitted in the same encoding. The encoding is done only to prevent inadvertent password compromises (i.e., a system administrator accidentally looking at the file), and will not prevent even a naive attacker from gaining the password.

The separate CVS password file (see section Setting up the server for password authentication) allows people to use a different password for repository access than for login access. On the other hand, once a user has non-read-only access to the repository, she can execute programs on the server system through a variety of means. Thus, repository access implies fairly broad system access as well. It might be possible to modify CVS to prevent that, but no one has done so as of this writing.

Note that because the `$CVSROOT/CVSROOT' directory contains `passwd' and other files which are used to check security, you must control the permissions on this directory as tightly as the permissions on `/etc'. The same applies to the `$CVSROOT' directory itself and any directory above it in the tree. Anyone who has write access to such a directory will have the ability to become any user on the system. Note that these permissions are typically tighter than you would use if you are not using pserver.

In summary, anyone who gets the password gets repository access (which may imply some measure of general system access as well). The password is available to anyone who can sniff network packets or read a protected (i.e., user read-only) file. If you want real security, get Kerberos.

2.9.4 Direct connection with GSSAPI

GSSAPI is a generic interface to network security systems such as Kerberos 5. If you have a working GSSAPI library, you can have CVS connect via a direct TCP connection, authenticating with GSSAPI.

To do this, CVS needs to be compiled with GSSAPI support; when configuring CVS it tries to detect whether GSSAPI libraries using Kerberos version 5 are present. You can also use the `--with-gssapi' flag to configure.

The connection is authenticated using GSSAPI, but the message stream is not authenticated by default. You must use the -a global option to request stream authentication.

The data transmitted is not encrypted by default. Encryption support must be compiled into both the client and the server; use the `--enable-encrypt' configure option to turn it on. You must then use the -x global option to request encryption.

GSSAPI connections are handled on the server side by the same server which handles the password authentication server; see Setting up the server for password authentication. If you are using a GSSAPI mechanism such as Kerberos which provides for strong authentication, you will probably want to disable the ability to authenticate via cleartext passwords. To do so, create an empty `CVSROOT/passwd' password file, and set SystemAuth=no in the config file (see section The CVSROOT/config configuration file).

The GSSAPI server uses a principal name of cvs/hostname, where hostname is the canonical name of the server host. You will have to set this up as required by your GSSAPI mechanism.

To connect using GSSAPI, use the `:gserver:' method. For example,

cvs -d :gserver:faun.example.org:/usr/local/cvsroot checkout foo

2.9.5 Direct connection with Kerberos

The easiest way to use Kerberos is to use the Kerberos rsh, as described in Connecting with rsh. The main disadvantage of using rsh is that all the data needs to pass through additional programs, so it may be slower. So if you have Kerberos installed you can connect via a direct TCP connection, authenticating with Kerberos.

This section concerns the Kerberos network security system, version 4. Kerberos version 5 is supported via the GSSAPI generic network security interface, as described in the previous section.

To do this, CVS needs to be compiled with Kerberos support; when configuring CVS it tries to detect whether Kerberos is present or you can use the `--with-krb4' flag to configure.

The data transmitted is not encrypted by default. Encryption support must be compiled into both the client and server; use the `--enable-encryption' configure option to turn it on. You must then use the -x global option to request encryption.

You need to edit `inetd.conf' on the server machine to run cvs kserver. The client uses port 1999 by default; if you want to use another port specify it in the CVSROOT (see section Remote repositories) or the CVS_CLIENT_PORT environment variable (see section All environment variables which affect CVS) on the client.

When you want to use CVS, get a ticket in the usual way (generally kinit); it must be a ticket which allows you to log into the server machine. Then you are ready to go:

cvs -d :kserver:faun.example.org:/usr/local/cvsroot checkout foo

Previous versions of CVS would fall back to a connection via rsh; this version will not do so.

2.9.6 Connecting with fork

This access method allows you to connect to a repository on your local disk via the remote protocol. In other words it does pretty much the same thing as :local:, but various quirks, bugs and the like are those of the remote CVS rather than the local CVS.

For day-to-day operations you might prefer either :local: or :fork:, depending on your preferences. Of course :fork: comes in particularly handy in testing or debugging cvs and the remote protocol. Specifically, we avoid all of the network-related setup/configuration, timeouts, and authentication inherent in the other remote access methods but still create a connection which uses the remote protocol.

To connect using the fork method, use `:fork:' and the pathname to your local repository. For example:

cvs -d :fork:/usr/local/cvsroot checkout foo

As with :ext:, the server is called `cvs' by default, or the value of the CVS_SERVER environment variable.

2.10 Read-only repository access

It is possible to grant read-only repository access to people using the password-authenticated server (see section Direct connection with password authentication). (The other access methods do not have explicit support for read-only users because those methods all assume login access to the repository machine anyway, and therefore the user can do whatever local file permissions allow her to do.)

A user who has read-only access can do only those CVS operations which do not modify the repository, except for certain “administrative” files (such as lock files and the history file). It may be desirable to use this feature in conjunction with user-aliasing (see section Setting up the server for password authentication).

Unlike with previous versions of CVS, read-only users should be able merely to read the repository, and not to execute programs on the server or otherwise gain unexpected levels of access. Or to be more accurate, the known holes have been plugged. Because this feature is new and has not received a comprehensive security audit, you should use whatever level of caution seems warranted given your attitude concerning security.

There are two ways to specify read-only access for a user: by inclusion, and by exclusion.

"Inclusion" means listing that user specifically in the `$CVSROOT/CVSROOT/readers' file, which is simply a newline-separated list of users. Here is a sample `readers' file:

melissa
splotnik
jrandom

(Don't forget the newline after the last user.)

"Exclusion" means explicitly listing everyone who has write access—if the file

$CVSROOT/CVSROOT/writers

exists, then only those users listed in it have write access, and everyone else has read-only access (of course, even the read-only users still need to be listed in the CVS `passwd' file). The `writers' file has the same format as the `readers' file.

Note: if your CVS `passwd' file maps cvs users onto system users (see section Setting up the server for password authentication), make sure you deny or grant read-only access using the cvs usernames, not the system usernames. That is, the `readers' and `writers' files contain cvs usernames, which may or may not be the same as system usernames.

Here is a complete description of the server's behavior in deciding whether to grant read-only or read-write access:

If `readers' exists, and this user is listed in it, then she gets read-only access. Or if `writers' exists, and this user is NOT listed in it, then she also gets read-only access (this is true even if `readers' exists but she is not listed there). Otherwise, she gets full read-write access.

Of course there is a conflict if the user is listed in both files. This is resolved in the more conservative way, it being better to protect the repository too much than too little: such a user gets read-only access.

2.11 Temporary directories for the server

While running, the CVS server creates temporary directories. They are named

cvs-servpid

where pid is the process identification number of the server. They are located in the directory specified by the `-T' global option (see section Global options), the TMPDIR environment variable (see section All environment variables which affect CVS), or, failing that, `/tmp'.

In most cases the server will remove the temporary directory when it is done, whether it finishes normally or abnormally. However, there are a few cases in which the server does not or cannot remove the temporary directory, for example:

• If the server aborts due to an internal server error, it may preserve the directory to aid in debugging
• If the server is killed in a way that it has no way of cleaning up (most notably, `kill -KILL' on unix).
• If the system shuts down without an orderly shutdown, which tells the server to clean up.

In cases such as this, you will need to manually remove the `cvs-servpid' directories. As long as there is no server running with process identification number pid, it is safe to do so.

3. Starting a project with CVS

Because renaming files and moving them between directories is somewhat inconvenient, the first thing you do when you start a new project should be to think through your file organization. It is not impossible to rename or move files, but it does increase the potential for confusion and CVS does have some quirks particularly in the area of renaming directories. See section Moving and renaming files.

What to do next depends on the situation at hand.

3.1 Setting up the files

The first step is to create the files inside the repository. This can be done in a couple of different ways.

3.1.1 Creating a directory tree from a number of files

When you begin using CVS, you will probably already have several projects that can be put under CVS control. In these cases the easiest way is to use the import command. An example is probably the easiest way to explain how to use it. If the files you want to install in CVS reside in `wdir', and you want them to appear in the repository as `$CVSROOT/yoyodyne/rdir', you can do this:

$ cd wdir
$ cvs import -m "Imported sources" yoyodyne/rdir yoyo start

Unless you supply a log message with the `-m' flag, CVS starts an editor and prompts for a message. The string `yoyo' is a vendor tag, and `start' is a release tag. They may fill no purpose in this context, but since CVS requires them they must be present. See section Tracking third-party sources, for more information about them.

You can now verify that it worked, and remove your original source directory.

$ cd ..
$ cvs checkout yoyodyne/rdir       # Explanation below
$ diff -r wdir yoyodyne/rdir
$ rm -r wdir

Erasing the original sources is a good idea, to make sure that you do not accidentally edit them in wdir, bypassing CVS. Of course, it would be wise to make sure that you have a backup of the sources before you remove them.

The checkout command can either take a module name as argument (as it has done in all previous examples) or a path name relative to $CVSROOT, as it did in the example above.

It is a good idea to check that the permissions CVS sets on the directories inside $CVSROOT are reasonable, and that they belong to the proper groups. See section File permissions.

If some of the files you want to import are binary, you may want to use the wrappers features to specify which files are binary and which are not. See section The cvswrappers file.

3.1.2 Creating Files From Other Version Control Systems

If you have a project which you are maintaining with another version control system, such as RCS, you may wish to put the files from that project into CVS, and preserve the revision history of the files.

From RCS

If you have been using RCS, find the RCS files—usually a file named `foo.c' will have its RCS file in `RCS/foo.c,v' (but it could be other places; consult the RCS documentation for details). Then create the appropriate directories in CVS if they do not already exist. Then copy the files into the appropriate directories in the CVS repository (the name in the repository must be the name of the source file with `,v' added; the files go directly in the appropriate directory of the repository, not in an `RCS' subdirectory). This is one of the few times when it is a good idea to access the CVS repository directly, rather than using CVS commands. Then you are ready to check out a new working directory.

The RCS file should not be locked when you move it into CVS; if it is, CVS will have trouble letting you operate on it.

From another version control system

Many version control systems have the ability to export RCS files in the standard format. If yours does, export the RCS files and then follow the above instructions.

Failing that, probably your best bet is to write a script that will check out the files one revision at a time using the command line interface to the other system, and then check the revisions into CVS. The `sccs2rcs' script mentioned below may be a useful example to follow.

From SCCS

There is a script in the `contrib' directory of the CVS source distribution called `sccs2rcs' which converts SCCS files to RCS files. Note: you must run it on a machine which has both SCCS and RCS installed, and like everything else in contrib it is unsupported (your mileage may vary).

From PVCS

There is a script in the `contrib' directory of the CVS source distribution called `pvcs_to_rcs' which converts PVCS archives to RCS files. You must run it on a machine which has both PVCS and RCS installed, and like everything else in contrib it is unsupported (your mileage may vary). See the comments in the script for details.

3.1.3 Creating a directory tree from scratch

For a new project, the easiest thing to do is probably to create an empty directory structure, like this:

$ mkdir tc
$ mkdir tc/man
$ mkdir tc/testing

After that, you use the import command to create the corresponding (empty) directory structure inside the repository:

$ cd tc
$ cvs import -m "Created directory structure" yoyodyne/dir yoyo start

This will add yoyodyne/dir as a directory under $CVSROOT.

Use checkout to get the new project. Then, use add to add files (and new directories) as needed.

$ cd ..
$ cvs co yoyodyne/dir

Check that the permissions CVS sets on the directories inside $CVSROOT are reasonable.

3.2 Defining the module

The next step is to define the module in the `modules' file. This is not strictly necessary, but modules can be convenient in grouping together related files and directories.

In simple cases these steps are sufficient to define a module.

1. Get a working copy of the modules file.

   $ cvs checkout CVSROOT/modules $ cd CVSROOT

2. Edit the file and insert a line that defines the module. See section The administrative files, for an introduction. See section The modules file, for a full description of the modules file. You can use the following line to define the module `tc':

   tc yoyodyne/tc

3. Commit your changes to the modules file.

   $ cvs commit -m "Added the tc module." modules

4. Release the modules module.

   $ cd .. $ cvs release -d CVSROOT

4. Revisions

For many uses of CVS, one doesn't need to worry too much about revision numbers; CVS assigns numbers such as 1.1, 1.2, and so on, and that is all one needs to know. However, some people prefer to have more knowledge and control concerning how CVS assigns revision numbers.

If one wants to keep track of a set of revisions involving more than one file, such as which revisions went into a particular release, one uses a tag, which is a symbolic revision which can be assigned to a numeric revision in each file.

4.1 Revision numbers

Each version of a file has a unique revision number. Revision numbers look like `1.1', `1.2', `1.3.2.2' or even `1.3.2.2.4.5'. A revision number always has an even number of period-separated decimal integers. By default revision 1.1 is the first revision of a file. Each successive revision is given a new number by increasing the rightmost number by one. The following figure displays a few revisions, with newer revisions to the right.

       +-----+    +-----+    +-----+    +-----+    +-----+
       ! 1.1 !----! 1.2 !----! 1.3 !----! 1.4 !----! 1.5 !
       +-----+    +-----+    +-----+    +-----+    +-----+

It is also possible to end up with numbers containing more than one period, for example `1.3.2.2'. Such revisions represent revisions on branches (see section Branching and merging); such revision numbers are explained in detail in Branches and revisions.

4.2 Versions, revisions and releases

A file can have several versions, as described above. Likewise, a software product can have several versions. A software product is often given a version number such as `4.1.1'.

Versions in the first sense are called revisions in this document, and versions in the second sense are called releases. To avoid confusion, the word version is almost never used in this document.

4.3 Assigning revisions

By default, CVS will assign numeric revisions by leaving the first number the same and incrementing the second number. For example, 1.1, 1.2, 1.3, etc.

When adding a new file, the second number will always be one and the first number will equal the highest first number of any file in that directory. For example, the current directory contains files whose highest numbered revisions are 1.7, 3.1, and 4.12, then an added file will be given the numeric revision 4.1. (When using client/server CVS, only files that are actually sent to the server are considered.)

Normally there is no reason to care about the revision numbers—it is easier to treat them as internal numbers that CVS maintains, and tags provide a better way to distinguish between things like release 1 versus release 2 of your product (see section Tags–Symbolic revisions). However, if you want to set the numeric revisions, the `-r' option to cvs commit can do that. The `-r' option implies the `-f' option, in the sense that it causes the files to be committed even if they are not modified.

For example, to bring all your files up to revision 3.0 (including those that haven't changed), you might invoke:

$ cvs commit -r 3.0

Note that the number you specify with `-r' must be larger than any existing revision number. That is, if revision 3.0 exists, you cannot `cvs commit -r 1.3'. If you want to maintain several releases in parallel, you need to use a branch (see section Branching and merging).

4.4 Tags–Symbolic revisions

The revision numbers live a life of their own. They need not have anything at all to do with the release numbers of your software product. Depending on how you use CVS the revision numbers might change several times between two releases. As an example, some of the source files that make up RCS 5.6 have the following revision numbers:

ci.c            5.21
co.c            5.9
ident.c         5.3
rcs.c           5.12
rcsbase.h       5.11
rcsdiff.c       5.10
rcsedit.c       5.11
rcsfcmp.c       5.9
rcsgen.c        5.10
rcslex.c        5.11
rcsmap.c        5.2
rcsutil.c       5.10

You can use the tag command to give a symbolic name to a certain revision of a file. You can use the `-v' flag to the status command to see all tags that a file has, and which revision numbers they represent. Tag names must start with an uppercase or lowercase letter and can contain uppercase and lowercase letters, digits, `-', and `_'. The two tag names BASE and HEAD are reserved for use by CVS. It is expected that future names which are special to CVS will be specially named, for example by starting with `.', rather than being named analogously to BASE and HEAD, to avoid conflicts with actual tag names.

You'll want to choose some convention for naming tags, based on information such as the name of the program and the version number of the release. For example, one might take the name of the program, immediately followed by the version number with `.' changed to `-', so that CVS 1.9 would be tagged with the name cvs1-9. If you choose a consistent convention, then you won't constantly be guessing whether a tag is cvs-1-9 or cvs1_9 or what. You might even want to consider enforcing your convention in the `taginfo' file (see section Taginfo).

The following example shows how you can add a tag to a file. The commands must be issued inside your working directory. That is, you should issue the command in the directory where `backend.c' resides.

$ cvs tag rel-0-4 backend.c
T backend.c
$ cvs status -v backend.c
===================================================================
File: backend.c         Status: Up-to-date

    Version:            1.4     Tue Dec  1 14:39:01 1992
    RCS Version:        1.4     /u/cvsroot/yoyodyne/tc/backend.c,v
    Sticky Tag:         (none)
    Sticky Date:        (none)
    Sticky Options:     (none)

    Existing Tags:
        rel-0-4                     (revision: 1.4)

For a complete summary of the syntax of cvs tag, including the various options, see Quick reference to CVS commands.

There is seldom reason to tag a file in isolation. A more common use is to tag all the files that constitute a module with the same tag at strategic points in the development life-cycle, such as when a release is made.

$ cvs tag rel-1-0 .
cvs tag: Tagging .
T Makefile
T backend.c
T driver.c
T frontend.c
T parser.c

(When you give CVS a directory as argument, it generally applies the operation to all the files in that directory, and (recursively), to any subdirectories that it may contain. See section Recursive behavior.)

The checkout command has a flag, `-r', that lets you check out a certain revision of a module. This flag makes it easy to retrieve the sources that make up release 1.0 of the module `tc' at any time in the future:

$ cvs checkout -r rel-1-0 tc

This is useful, for instance, if someone claims that there is a bug in that release, but you cannot find the bug in the current working copy.

You can also check out a module as it was at any given date. See section checkout options. When specifying `-r' to any of these commands, you will need beware of sticky tags; see Sticky tags.

When you tag more than one file with the same tag you can think about the tag as "a curve drawn through a matrix of filename vs. revision number." Say we have 5 files with the following revisions:

        file1   file2   file3   file4   file5

        1.1     1.1     1.1     1.1  /--1.1*      <-*-  TAG
        1.2*-   1.2     1.2    -1.2*-
        1.3  \- 1.3*-   1.3   / 1.3
        1.4          \  1.4  /  1.4
                      \-1.5*-   1.5
                        1.6

At some time in the past, the * versions were tagged. You can think of the tag as a handle attached to the curve drawn through the tagged revisions. When you pull on the handle, you get all the tagged revisions. Another way to look at it is that you "sight" through a set of revisions that is "flat" along the tagged revisions, like this:

        file1   file2   file3   file4   file5

                        1.1
                        1.2
                1.1     1.3                       _
        1.1     1.2     1.4     1.1              /
        1.2*----1.3*----1.5*----1.2*----1.1     (--- <--- Look here
        1.3             1.6     1.3              \_
        1.4                     1.4
                                1.5

4.5 Specifying what to tag from the working directory

The example in the previous section demonstrates one of the most common ways to choose which revisions to tag. Namely, running the cvs tag command without arguments causes CVS to select the revisions which are checked out in the current working directory. For example, if the copy of `backend.c' in working directory was checked out from revision 1.4, then CVS will tag revision 1.4. Note that the tag is applied immediately to revision 1.4 in the repository; tagging is not like modifying a file, or other operations in which one first modifies the working directory and then runs cvs commit to transfer that modification to the repository.

One potentially surprising aspect of the fact that cvs tag operates on the repository is that you are tagging the checked-in revisions, which may differ from locally modified files in your working directory. If you want to avoid doing this by mistake, specify the `-c' option to cvs tag. If there are any locally modified files, CVS will abort with an error before it tags any files:

$ cvs tag -c rel-0-4
cvs tag: backend.c is locally modified
cvs [tag aborted]: correct the above errors first!

4.6 Specifying what to tag by date or revision

The cvs rtag command tags the repository as of a certain date or time (or can be used to tag the latest revision). rtag works directly on the repository contents (it requires no prior checkout and does not look for a working directory).

The following options specify which date or revision to tag. See Common command options, for a complete description of them.

-D date

Tag the most recent revision no later than date.

-f

Only useful with the `-D date' or `-r tag' flags. If no matching revision is found, use the most recent revision (instead of ignoring the file).

-r tag

Only tag those files that contain existing tag tag.

The cvs tag command also allows one to specify files by revision or date, using the same `-r', `-D', and `-f' options. However, this feature is probably not what you want. The reason is that cvs tag chooses which files to tag based on the files that exist in the working directory, rather than the files which existed as of the given tag/date. Therefore, you are generally better off using cvs rtag. The exceptions might be cases like:

cvs tag -r 1.4 backend.c

4.7 Deleting, moving, and renaming tags

Normally one does not modify tags. They exist in order to record the history of the repository and so deleting them or changing their meaning would, generally, not be what you want.

However, there might be cases in which one uses a tag temporarily or accidentally puts one in the wrong place. Therefore, one might delete, move, or rename a tag.

WARNING: the commands in this section are dangerous; they permanently discard historical information and it can be difficult or impossible to recover from errors. If you are a CVS administrator, you may consider restricting these commands with the `taginfo' file (see section Taginfo).

To delete a tag, specify the `-d' option to either cvs tag or cvs rtag. For example:

cvs rtag -d rel-0-4 tc

deletes the non-branch tag rel-0-4 from the module tc. In the event that branch tags are encountered within the repository with the given name, a warning message will be issued and the branch tag will not be deleted. If you are absolutely certain you know what you are doing, the -B option may be specified to allow deletion of branch tags. In that case, any non-branch tags encountered will trigger warnings and will not be deleted.

WARNING: Moving branch tags is very dangerous! If you think you need the -B option, think again and ask your CVS administrator about it (if that isn't you). There is almost certainly another way to accomplish what you want to accomplish.

When we say move a tag, we mean to make the same name point to different revisions. For example, the stable tag may currently point to revision 1.4 of `backend.c' and perhaps we want to make it point to revision 1.6. To move a non-branch tag, specify the `-F' option to either cvs tag or cvs rtag. For example, the task just mentioned might be accomplished as:

cvs tag -r 1.6 -F stable backend.c

If any branch tags are encountered in the repository with the given name, a warning is issued and the branch tag is not disturbed. If you are absolutely certain you wish to move the branch tag, the -B option may be specified. In that case, non-branch tags encountered with the given name are ignored with a warning message.

WARNING: Moving branch tags is very dangerous! If you think you need the -B option, think again and ask your CVS administrator about it (if that isn't you). There is almost certainly another way to accomplish what you want to accomplish.

When we say rename a tag, we mean to make a different name point to the same revisions as the old tag. For example, one may have misspelled the tag name and want to correct it (hopefully before others are relying on the old spelling). To rename a tag, first create a new tag using the `-r' option to cvs rtag, and then delete the old name. (Caution: this method will not work with branch tags.) This leaves the new tag on exactly the same files as the old tag. For example:

cvs rtag -r old-name-0-4 rel-0-4 tc
cvs rtag -d old-name-0-4 tc

4.8 Tagging and adding and removing files

The subject of exactly how tagging interacts with adding and removing files is somewhat obscure; for the most part CVS will keep track of whether files exist or not without too much fussing. By default, tags are applied to only files which have a revision corresponding to what is being tagged. Files which did not exist yet, or which were already removed, simply omit the tag, and CVS knows to treat the absence of a tag as meaning that the file didn't exist as of that tag.

However, this can lose a small amount of information. For example, suppose a file was added and then removed. Then, if the tag is missing for that file, there is no way to know whether the tag refers to the time before the file was added, or the time after it was removed. If you specify the `-r' option to cvs rtag, then CVS tags the files which have been removed, and thereby avoids this problem. For example, one might specify -r HEAD to tag the head.

On the subject of adding and removing files, the cvs rtag command has a `-a' option which means to clear the tag from removed files that would not otherwise be tagged. For example, one might specify this option in conjunction with `-F' when moving a tag. If one moved a tag without `-a', then the tag in the removed files might still refer to the old revision, rather than reflecting the fact that the file had been removed. I don't think this is necessary if `-r' is specified, as noted above.

4.9 Sticky tags

Sometimes a working copy's revision has extra data associated with it, for example it might be on a branch (see section Branching and merging), or restricted to versions prior to a certain date by `checkout -D' or `update -D'. Because this data persists – that is, it applies to subsequent commands in the working copy – we refer to it as sticky.

Most of the time, stickiness is an obscure aspect of CVS that you don't need to think about. However, even if you don't want to use the feature, you may need to know something about sticky tags (for example, how to avoid them!).

You can use the status command to see if any sticky tags or dates are set:

$ cvs status driver.c
===================================================================
File: driver.c          Status: Up-to-date

    Version:            1.7.2.1 Sat Dec  5 19:35:03 1992
    RCS Version:        1.7.2.1 /u/cvsroot/yoyodyne/tc/driver.c,v
    Sticky Tag:         rel-1-0-patches (branch: 1.7.2)
    Sticky Date:        (none)
    Sticky Options:     (none)

The sticky tags will remain on your working files until you delete them with `cvs update -A'. The `-A' option merges local changes into the version of the file from the head of the trunk, removing any sticky tags, dates, or options (other than sticky `-k' options on locally modified files). See update—Bring work tree in sync with repository for more on the operation of cvs update.

The most common use of sticky tags is to identify which branch one is working on, as described in Accessing branches. However, non-branch sticky tags have uses as well. For example, suppose that you want to avoid updating your working directory, to isolate yourself from possibly destabilizing changes other people are making. You can, of course, just refrain from running cvs update. But if you want to avoid updating only a portion of a larger tree, then sticky tags can help. If you check out a certain revision (such as 1.4) it will become sticky. Subsequent cvs update commands will not retrieve the latest revision until you reset the tag with cvs update -A. Likewise, use of the `-D' option to update or checkout sets a sticky date, which, similarly, causes that date to be used for future retrievals.

People often want to retrieve an old version of a file without setting a sticky tag. This can be done with the `-p' option to checkout or update, which sends the contents of the file to standard output. For example:

$ cvs update -p -r 1.1 file1 >file1
===================================================================
Checking out file1
RCS:  /tmp/cvs-sanity/cvsroot/first-dir/Attic/file1,v
VERS: 1.1
***************
$

However, this isn't the easiest way, if you are asking how to undo a previous checkin (in this example, put `file1' back to the way it was as of revision 1.1). In that case you are better off using the `-j' option to update; for further discussion see Merging differences between any two revisions.

5. Branching and merging

CVS allows you to isolate changes onto a separate line of development, known as a branch. When you change files on a branch, those changes do not appear on the main trunk or other branches.

Later you can move changes from one branch to another branch (or the main trunk) by merging. Merging involves first running cvs update -j, to merge the changes into the working directory. You can then commit that revision, and thus effectively copy the changes onto another branch.

5.1 What branches are good for

Suppose that release 1.0 of tc has been made. You are continuing to develop tc, planning to create release 1.1 in a couple of months. After a while your customers start to complain about a fatal bug. You check out release 1.0 (see section Tags–Symbolic revisions) and find the bug (which turns out to have a trivial fix). However, the current revision of the sources are in a state of flux and are not expected to be stable for at least another month. There is no way to make a bug fix release based on the newest sources.

The thing to do in a situation like this is to create a branch on the revision trees for all the files that make up release 1.0 of tc. You can then make modifications to the branch without disturbing the main trunk. When the modifications are finished you can elect to either incorporate them on the main trunk, or leave them on the branch.

5.2 Creating a branch

You can create a branch with tag -b; for example, assuming you're in a working copy:

$ cvs tag -b rel-1-0-patches

This splits off a branch based on the current revisions in the working copy, assigning that branch the name `rel-1-0-patches'.

It is important to understand that branches get created in the repository, not in the working copy. Creating a branch based on current revisions, as the above example does, will not automatically switch the working copy to be on the new branch. For information on how to do that, see Accessing branches.

You can also create a branch without reference to any working copy, by using rtag:

$ cvs rtag -b -r rel-1-0 rel-1-0-patches tc

`-r rel-1-0' says that this branch should be rooted at the revision that corresponds to the tag `rel-1-0'. It need not be the most recent revision – it's often useful to split a branch off an old revision (for example, when fixing a bug in a past release otherwise known to be stable).

As with `tag', the `-b' flag tells rtag to create a branch (rather than just a symbolic revision name). Note that the numeric revision number that matches `rel-1-0' will probably be different from file to file.

So, the full effect of the command is to create a new branch – named `rel-1-0-patches' – in module `tc', rooted in the revision tree at the point tagged by `rel-1-0'.

5.3 Accessing branches

You can retrieve a branch in one of two ways: by checking it out fresh from the repository, or by switching an existing working copy over to the branch.

To check out a branch from the repository, invoke `checkout' with the `-r' flag, followed by the tag name of the branch (see section Creating a branch):

$ cvs checkout -r rel-1-0-patches tc

Or, if you already have a working copy, you can switch it to a given branch with `update -r':

$ cvs update -r rel-1-0-patches tc

or equivalently:

$ cd tc
$ cvs update -r rel-1-0-patches

It does not matter if the working copy was originally on the main trunk or on some other branch – the above command will switch it to the named branch. And similarly to a regular `update' command, `update -r' merges any changes you have made, notifying you of conflicts where they occur.

Once you have a working copy tied to a particular branch, it remains there until you tell it otherwise. This means that changes checked in from the working copy will add new revisions on that branch, while leaving the main trunk and other branches unaffected.

To find out what branch a working copy is on, you can use the `status' command. In its output, look for the field named `Sticky tag' (see section Sticky tags) – that's CVS's way of telling you the branch, if any, of the current working files:

$ cvs status -v driver.c backend.c
===================================================================
File: driver.c          Status: Up-to-date

    Version:            1.7     Sat Dec  5 18:25:54 1992
    RCS Version:        1.7     /u/cvsroot/yoyodyne/tc/driver.c,v
    Sticky Tag:         rel-1-0-patches (branch: 1.7.2)
    Sticky Date:        (none)
    Sticky Options:     (none)

    Existing Tags:
        rel-1-0-patches             (branch: 1.7.2)
        rel-1-0                     (revision: 1.7)

===================================================================
File: backend.c         Status: Up-to-date

    Version:            1.4     Tue Dec  1 14:39:01 1992
    RCS Version:        1.4     /u/cvsroot/yoyodyne/tc/backend.c,v
    Sticky Tag:         rel-1-0-patches (branch: 1.4.2)
    Sticky Date:        (none)
    Sticky Options:     (none)

    Existing Tags:
        rel-1-0-patches             (branch: 1.4.2)
        rel-1-0                     (revision: 1.4)
        rel-0-4                     (revision: 1.4)

Don't be confused by the fact that the branch numbers for each file are different (`1.7.2' and `1.4.2' respectively). The branch tag is the same, `rel-1-0-patches', and the files are indeed on the same branch. The numbers simply reflect the point in each file's revision history at which the branch was made. In the above example, one can deduce that `driver.c' had been through more changes than `backend.c' before this branch was created.

See Branches and revisions for details about how branch numbers are constructed.

5.4 Branches and revisions

Ordinarily, a file's revision history is a linear series of increments (see section Revision numbers):

       +-----+    +-----+    +-----+    +-----+    +-----+
       ! 1.1 !----! 1.2 !----! 1.3 !----! 1.4 !----! 1.5 !
       +-----+    +-----+    +-----+    +-----+    +-----+

However, CVS is not limited to linear development. The revision tree can be split into branches, where each branch is a self-maintained line of development. Changes made on one branch can easily be moved back to the main trunk.

Each branch has a branch number, consisting of an odd number of period-separated decimal integers. The branch number is created by appending an integer to the revision number where the corresponding branch forked off. Having branch numbers allows more than one branch to be forked off from a certain revision.

All revisions on a branch have revision numbers formed by appending an ordinal number to the branch number. The following figure illustrates branching with an example.

                                                      +-------------+
                           Branch 1.2.2.3.2 ->        ! 1.2.2.3.2.1 !
                                                    / +-------------+
                                                   /
                                                  /
                 +---------+    +---------+    +---------+
Branch 1.2.2 -> _! 1.2.2.1 !----! 1.2.2.2 !----! 1.2.2.3 !
               / +---------+    +---------+    +---------+
              /
             /
+-----+    +-----+    +-----+    +-----+    +-----+
! 1.1 !----! 1.2 !----! 1.3 !----! 1.4 !----! 1.5 !  <- The main trunk
+-----+    +-----+    +-----+    +-----+    +-----+
                !
                !
                !   +---------+    +---------+    +---------+
Branch 1.2.4 -> +---! 1.2.4.1 !----! 1.2.4.2 !----! 1.2.4.3 !
                    +---------+    +---------+    +---------+

The exact details of how the branch number is constructed is not something you normally need to be concerned about, but here is how it works: When CVS creates a branch number it picks the first unused even integer, starting with 2. So when you want to create a branch from revision 6.4 it will be numbered 6.4.2. All branch numbers ending in a zero (such as 6.4.0) are used internally by CVS (see section Magic branch numbers). The branch 1.1.1 has a special meaning. See section Tracking third-party sources.

5.5 Magic branch numbers

This section describes a CVS feature called magic branches. For most purposes, you need not worry about magic branches; CVS handles them for you. However, they are visible to you in certain circumstances, so it may be useful to have some idea of how it works.

Externally, branch numbers consist of an odd number of dot-separated decimal integers. See section Revision numbers. That is not the whole truth, however. For efficiency reasons CVS sometimes inserts an extra 0 in the second rightmost position (1.2.4 becomes 1.2.0.4, 8.9.10.11.12 becomes 8.9.10.11.0.12 and so on).

CVS does a pretty good job at hiding these so called magic branches, but in a few places the hiding is incomplete:

• The magic branch number appears in the output from cvs log.
• You cannot specify a symbolic branch name to cvs admin.

You can use the admin command to reassign a symbolic name to a branch the way RCS expects it to be. If R4patches is assigned to the branch 1.4.2 (magic branch number 1.4.0.2) in file `numbers.c' you can do this:

$ cvs admin -NR4patches:1.4.2 numbers.c

It only works if at least one revision is already committed on the branch. Be very careful so that you do not assign the tag to the wrong number. (There is no way to see how the tag was assigned yesterday).

5.6 Merging an entire branch

You can merge changes made on a branch into your working copy by giving the `-j branchname' flag to the update subcommand. With one `-j branchname' option it merges the changes made between the greatest common ancestor (GCA) of the branch and the destination revision (in the simple case below the GCA is the point where the branch forked) and the newest revision on that branch into your working copy.

The `-j' stands for “join”.

Consider this revision tree:

+-----+    +-----+    +-----+    +-----+
! 1.1 !----! 1.2 !----! 1.3 !----! 1.4 !      <- The main trunk
+-----+    +-----+    +-----+    +-----+
                !
                !
                !   +---------+    +---------+
Branch R1fix -> +---! 1.2.2.1 !----! 1.2.2.2 !
                    +---------+    +---------+

The branch 1.2.2 has been given the tag (symbolic name) `R1fix'. The following example assumes that the module `mod' contains only one file, `m.c'.

$ cvs checkout mod               # Retrieve the latest revision, 1.4

$ cvs update -j R1fix m.c        # Merge all changes made on the branch,
                                 # i.e. the changes between revision 1.2
                                 # and 1.2.2.2, into your working copy
                                 # of the file.

$ cvs commit -m "Included R1fix" # Create revision 1.5.

A conflict can result from a merge operation. If that happens, you should resolve it before committing the new revision. See section Conflicts example.

If your source files contain keywords (see section Keyword substitution), you might be getting more conflicts than strictly necessary. See Merging and keywords, for information on how to avoid this.

The checkout command also supports the `-j branchname' flag. The same effect as above could be achieved with this:

$ cvs checkout -j R1fix mod
$ cvs commit -m "Included R1fix"

It should be noted that update -j tagname will also work but may not produce the desired result. See section Merging can add or remove files, for more.

5.7 Merging from a branch several times

Continuing our example, the revision tree now looks like this:

+-----+    +-----+    +-----+    +-----+    +-----+
! 1.1 !----! 1.2 !----! 1.3 !----! 1.4 !----! 1.5 !   <- The main trunk
+-----+    +-----+    +-----+    +-----+    +-----+
                !                           *
                !                          *
                !   +---------+    +---------+
Branch R1fix -> +---! 1.2.2.1 !----! 1.2.2.2 !
                    +---------+    +---------+

where the starred line represents the merge from the `R1fix' branch to the main trunk, as just discussed.

Now suppose that development continues on the `R1fix' branch:

+-----+    +-----+    +-----+    +-----+    +-----+
! 1.1 !----! 1.2 !----! 1.3 !----! 1.4 !----! 1.5 !   <- The main trunk
+-----+    +-----+    +-----+    +-----+    +-----+
                !                           *
                !                          *
                !   +---------+    +---------+    +---------+
Branch R1fix -> +---! 1.2.2.1 !----! 1.2.2.2 !----! 1.2.2.3 !
                    +---------+    +---------+    +---------+

and then you want to merge those new changes onto the main trunk. If you just use the cvs update -j R1fix m.c command again, CVS will attempt to merge again the changes which you have already merged, which can have undesirable side effects.

So instead you need to specify that you only want to merge the changes on the branch which have not yet been merged into the trunk. To do that you specify two `-j' options, and CVS merges the changes from the first revision to the second revision. For example, in this case the simplest way would be

cvs update -j 1.2.2.2 -j R1fix m.c    # Merge changes from 1.2.2.2 to the
                                      # head of the R1fix branch

The problem with this is that you need to specify the 1.2.2.2 revision manually. A slightly better approach might be to use the date the last merge was done:

cvs update -j R1fix:yesterday -j R1fix m.c

Better yet, tag the R1fix branch after every merge into the trunk, and then use that tag for subsequent merges:

cvs update -j merged_from_R1fix_to_trunk -j R1fix m.c

5.8 Merging differences between any two revisions

With two `-j revision' flags, the update (and checkout) command can merge the differences between any two revisions into your working file.

$ cvs update -j 1.5 -j 1.3 backend.c

will undo all changes made between revision 1.3 and 1.5. Note the order of the revisions!

If you try to use this option when operating on multiple files, remember that the numeric revisions will probably be very different between the various files. You almost always use symbolic tags rather than revision numbers when operating on multiple files.

Specifying two `-j' options can also undo file removals or additions. For example, suppose you have a file named `file1' which existed as revision 1.1, and you then removed it (thus adding a dead revision 1.2). Now suppose you want to add it again, with the same contents it had previously. Here is how to do it:

$ cvs update -j 1.2 -j 1.1 file1
U file1
$ cvs commit -m test
Checking in file1;
/tmp/cvs-sanity/cvsroot/first-dir/file1,v  <--  file1
new revision: 1.3; previous revision: 1.2
done
$

5.9 Merging can add or remove files

If the changes which you are merging involve removing or adding some files, update -j will reflect such additions or removals.

For example:

cvs update -A
touch a b c
cvs add a b c ; cvs ci -m "added" a b c
cvs tag -b branchtag
cvs update -r branchtag
touch d ; cvs add d
rm a ; cvs rm a
cvs ci -m "added d, removed a"
cvs update -A
cvs update -jbranchtag

After these commands are executed and a `cvs commit' is done, file `a' will be removed and file `d' added in the main branch.

Note that using a single static tag (`-j tagname') rather than a dynamic tag (`-j branchname') to merge changes from a branch will usually not remove files which were removed on the branch since CVS does not automatically add static tags to dead revisions. The exception to this rule occurs when a static tag has been attached to a dead revision manually. Use the branch tag to merge all changes from the branch or use two static tags as merge endpoints to be sure that all intended changes are propagated in the merge.

5.10 Merging and keywords

If you merge files containing keywords (see section Keyword substitution), you will normally get numerous conflicts during the merge, because the keywords are expanded differently in the revisions which you are merging.

Therefore, you will often want to specify the `-kk' (see section Substitution modes) switch to the merge command line. By substituting just the name of the keyword, not the expanded value of that keyword, this option ensures that the revisions which you are merging will be the same as each other, and avoid spurious conflicts.

For example, suppose you have a file like this:

       +---------+
      _! 1.1.2.1 !   <-  br1
     / +---------+
    /
   /
+-----+    +-----+
! 1.1 !----! 1.2 !
+-----+    +-----+

and your working directory is currently on the trunk (revision 1.2). Then you might get the following results from a merge:

$ cat file1
key $Revision: 1.2 $
. . .
$ cvs update -j br1
U file1
RCS file: /cvsroot/first-dir/file1,v
retrieving revision 1.1
retrieving revision 1.1.2.1
Merging differences between 1.1 and 1.1.2.1 into file1
rcsmerge: warning: conflicts during merge
$ cat file1
<<<<<<< file1
key $Revision: 1.2 $
=======
key $Revision: 1.1.2.1 $
>>>>>>> 1.1.2.1
. . .

What happened was that the merge tried to merge the differences between 1.1 and 1.1.2.1 into your working directory. So, since the keyword changed from Revision: 1.1 to Revision: 1.1.2.1, CVS tried to merge that change into your working directory, which conflicted with the fact that your working directory had contained Revision: 1.2.

Here is what happens if you had used `-kk':

$ cat file1
key $Revision: 1.2 $
. . .
$ cvs update -kk -j br1
U file1
RCS file: /cvsroot/first-dir/file1,v
retrieving revision 1.1
retrieving revision 1.1.2.1
Merging differences between 1.1 and 1.1.2.1 into file1
$ cat file1
key $Revision$
. . .

What is going on here is that revision 1.1 and 1.1.2.1 both expand as plain Revision, and therefore merging the changes between them into the working directory need not change anything. Therefore, there is no conflict.

There is, however, one major caveat with using `-kk' on merges. Namely, it overrides whatever keyword expansion mode CVS would normally have used. In particular, this is a problem if the mode had been `-kb' for a binary file. Therefore, if your repository contains binary files, you will need to deal with the conflicts rather than using `-kk'.

As a result of using `-kk' during the merge, each file examined by the update will have `-kk' set as sticky options. Running update -A will clear the sticky options on unmodified files, but it will not clear the sticky options on modified files. To get back to the default keyword substitution for modified files, you must commit the results of the merge and then run update -A.

6. Recursive behavior

Almost all of the subcommands of CVS work recursively when you specify a directory as an argument. For instance, consider this directory structure:

      $HOME
        |
        +--tc
        |   |
            +--CVS
            |      (internal CVS files)
            +--Makefile
            +--backend.c
            +--driver.c
            +--frontend.c
            +--parser.c
            +--man
            |    |
            |    +--CVS
            |    |  (internal CVS files)
            |    +--tc.1
            |
            +--testing
                 |
                 +--CVS
                 |  (internal CVS files)
                 +--testpgm.t
                 +--test2.t

If `tc' is the current working directory, the following is true:

• `cvs update testing' is equivalent to

  cvs update testing/testpgm.t testing/test2.t

• `cvs update testing man' updates all files in the subdirectories
• `cvs update .' or just `cvs update' updates all files in the tc directory

If no arguments are given to update it will update all files in the current working directory and all its subdirectories. In other words, `.' is a default argument to update. This is also true for most of the CVS subcommands, not only the update command.

The recursive behavior of the CVS subcommands can be turned off with the `-l' option. Conversely, the `-R' option can be used to force recursion if `-l' is specified in `~/.cvsrc' (see section Default options and the ~/.cvsrc file).

$ cvs update -l         # Don't update files in subdirectories

7. Adding, removing, and renaming files and directories

In the course of a project, one will often add new files. Likewise with removing or renaming, or with directories. The general concept to keep in mind in all these cases is that instead of making an irreversible change you want CVS to record the fact that a change has taken place, just as with modifying an existing file. The exact mechanisms to do this in CVS vary depending on the situation.

7.1 Adding files to a directory

To add a new file to a directory, follow these steps.

• You must have a working copy of the directory. See section Getting the source.
• Create the new file inside your working copy of the directory.
• Use `cvs add filename' to tell CVS that you want to version control the file. If the file contains binary data, specify `-kb' (see section Handling binary files).
• Use `cvs commit filename' to actually check in the file into the repository. Other developers cannot see the file until you perform this step.

You can also use the add command to add a new directory.

Unlike most other commands, the add command is not recursive. You have to expcicitly name files and directories that you wish to add to the repository. However, each directory will need to be added separately before you will be able to add new files to those directories.

$ mkdir -p foo/bar
$ cp ~/myfile foo/bar/myfile
$ cvs add foo foo/bar
$ cvs add foo/bar/myfile

Command: cvs add [-k kflag] [-m message] files …

Schedule files to be added to the repository. The files or directories specified with add must already exist in the current directory. To add a whole new directory hierarchy to the source repository (for example, files received from a third-party vendor), use the import command instead. See section import—Import sources into CVS, using vendor branches.

The added files are not placed in the source repository until you use commit to make the change permanent. Doing an add on a file that was removed with the remove command will undo the effect of the remove, unless a commit command intervened. See section Removing files, for an example.

The `-k' option specifies the default way that this file will be checked out; for more information see Substitution modes.

The `-m' option specifies a description for the file. This description appears in the history log (if it is enabled, see section The history file). It will also be saved in the version history inside the repository when the file is committed. The log command displays this description. The description can be changed using `admin -t'. See section admin—Administration. If you omit the `-m description' flag, an empty string will be used. You will not be prompted for a description.

For example, the following commands add the file `backend.c' to the repository:

$ cvs add backend.c
$ cvs commit -m "Early version. Not yet compilable." backend.c

When you add a file it is added only on the branch which you are working on (see section Branching and merging). You can later merge the additions to another branch if you want (see section Merging can add or remove files).

7.2 Removing files

Directories change. New files are added, and old files disappear. Still, you want to be able to retrieve an exact copy of old releases.

Here is what you can do to remove a file, but remain able to retrieve old revisions:

• Make sure that you have not made any uncommitted modifications to the file. See section Viewing differences, for one way to do that. You can also use the status or update command. If you remove the file without committing your changes, you will of course not be able to retrieve the file as it was immediately before you deleted it.
• Remove the file from your working copy of the directory. You can for instance use rm.
• Use `cvs remove filename' to tell CVS that you really want to delete the file.
• Use `cvs commit filename' to actually perform the removal of the file from the repository.

When you commit the removal of the file, CVS records the fact that the file no longer exists. It is possible for a file to exist on only some branches and not on others, or to re-add another file with the same name later. CVS will correctly create or not create the file, based on the `-r' and `-D' options specified to checkout or update.

Command: cvs remove [options] files …

Schedule file(s) to be removed from the repository (files which have not already been removed from the working directory are not processed). This command does not actually remove the file from the repository until you commit the removal. For a full list of options, see Quick reference to CVS commands.

Here is an example of removing several files:

$ cd test
$ rm *.c
$ cvs remove
cvs remove: Removing .
cvs remove: scheduling a.c for removal
cvs remove: scheduling b.c for removal
cvs remove: use 'cvs commit' to remove these files permanently
$ cvs ci -m "Removed unneeded files"
cvs commit: Examining .
cvs commit: Committing .

As a convenience you can remove the file and cvs remove it in one step, by specifying the `-f' option. For example, the above example could also be done like this:

$ cd test
$ cvs remove -f *.c
cvs remove: scheduling a.c for removal
cvs remove: scheduling b.c for removal
cvs remove: use 'cvs commit' to remove these files permanently
$ cvs ci -m "Removed unneeded files"
cvs commit: Examining .
cvs commit: Committing .

If you execute remove for a file, and then change your mind before you commit, you can undo the remove with an add command.

$ ls
CVS   ja.h  oj.c
$ rm oj.c
$ cvs remove oj.c
cvs remove: scheduling oj.c for removal
cvs remove: use 'cvs commit' to remove this file permanently
$ cvs add oj.c
U oj.c
cvs add: oj.c, version 1.1.1.1, resurrected

If you realize your mistake before you run the remove command you can use update to resurrect the file:

$ rm oj.c
$ cvs update oj.c
cvs update: warning: oj.c was lost
U oj.c

When you remove a file it is removed only on the branch which you are working on (see section Branching and merging). You can later merge the removals to another branch if you want (see section Merging can add or remove files).

7.3 Removing directories

In concept, removing directories is somewhat similar to removing files—you want the directory to not exist in your current working directories, but you also want to be able to retrieve old releases in which the directory existed.

The way that you remove a directory is to remove all the files in it. You don't remove the directory itself; there is no way to do that. Instead you specify the `-P' option to cvs update or cvs checkout, which will cause CVS to remove empty directories from working directories. (Note that cvs export always removes empty directories.) Probably the best way to do this is to always specify `-P'; if you want an empty directory then put a dummy file (for example `.keepme') in it to prevent `-P' from removing it.

Note that `-P' is implied by the `-r' or `-D' options of checkout. This way, CVS will be able to correctly create the directory or not depending on whether the particular version you are checking out contains any files in that directory.

7.4 Moving and renaming files

Moving files to a different directory or renaming them is not difficult, but some of the ways in which this works may be non-obvious. (Moving or renaming a directory is even harder. See section Moving and renaming directories.).

The examples below assume that the file old is renamed to new.

7.4.1 The Normal way to Rename

The normal way to move a file is to copy old to new, and then issue the normal CVS commands to remove old from the repository, and add new to it.

$ mv old new
$ cvs remove old
$ cvs add new
$ cvs commit -m "Renamed old to new" old new

This is the simplest way to move a file, it is not error-prone, and it preserves the history of what was done. Note that to access the history of the file you must specify the old or the new name, depending on what portion of the history you are accessing. For example, cvs log old will give the log up until the time of the rename.

When new is committed its revision numbers will start again, usually at 1.1, so if that bothers you, use the `-r rev' option to commit. For more information see Assigning revisions.

7.4.2 Moving the history file

This method is more dangerous, since it involves moving files inside the repository. Read this entire section before trying it out!

$ cd $CVSROOT/dir
$ mv old,v new,v

Advantages:

• The log of changes is maintained intact.
• The revision numbers are not affected.

Disadvantages:

• Old releases cannot easily be fetched from the repository. (The file will show up as new even in revisions from the time before it was renamed).
• There is no log information of when the file was renamed.
• Nasty things might happen if someone accesses the history file while you are moving it. Make sure no one else runs any of the CVS commands while you move it.

7.4.3 Copying the history file

This way also involves direct modifications to the repository. It is safe, but not without drawbacks.

# Copy the RCS file inside the repository
$ cd $CVSROOT/dir
$ cp old,v new,v
# Remove the old file
$ cd ~/dir
$ rm old
$ cvs remove old
$ cvs commit old
# Remove all tags from new
$ cvs update new
$ cvs log new             # Remember the non-branch tag names
$ cvs tag -d tag1 new
$ cvs tag -d tag2 new
…

By removing the tags you will be able to check out old revisions.

Advantages:

• Checking out old revisions works correctly, as long as you use `-rtag' and not `-Ddate' to retrieve the revisions.
• The log of changes is maintained intact.
• The revision numbers are not affected.

Disadvantages:

• You cannot easily see the history of the file across the rename.

7.5 Moving and renaming directories

The normal way to rename or move a directory is to rename or move each file within it as described in The Normal way to Rename. Then check out with the `-P' option, as described in Removing directories.

If you really want to hack the repository to rename or delete a directory in the repository, you can do it like this:

1. Inform everyone who has a checked out copy of the directory that the directory will be renamed. They should commit all their changes in all their copies of the project containing the directory to be removed, and remove all their working copies of said project, before you take the steps below.
2. Rename the directory inside the repository.

   $ cd $CVSROOT/parent-dir $ mv old-dir new-dir

3. Fix the CVS administrative files, if necessary (for instance if you renamed an entire module).
4. Tell everyone that they can check out again and continue working.

If someone had a working copy the CVS commands will cease to work for him, until he removes the directory that disappeared inside the repository.

It is almost always better to move the files in the directory instead of moving the directory. If you move the directory you are unlikely to be able to retrieve old releases correctly, since they probably depend on the name of the directories.

8. History browsing

Once you have used CVS to store a version control history—what files have changed when, how, and by whom, there are a variety of mechanisms for looking through the history.

8.1 Log messages

Whenever you commit a file you specify a log message.

To look through the log messages which have been specified for every revision which has been committed, use the cvs log command (see section log—Print out log information for files).

8.2 The history database

You can use the history file (see section The history file) to log various CVS actions. To retrieve the information from the history file, use the cvs history command (see section history—Show status of files and users).

Note: you can control what is logged to this file by using the `LogHistory' keyword in the `CVSROOT/config' file (see section The CVSROOT/config configuration file).

8.3 User-defined logging

You can customize CVS to log various kinds of actions, in whatever manner you choose. These mechanisms operate by executing a script at various times. The script might append a message to a file listing the information and the programmer who created it, or send mail to a group of developers, or, perhaps, post a message to a particular newsgroup. To log commits, use the `loginfo' file (see section Loginfo). To log tags, use the `taginfo' file (see section Taginfo). To log commits, checkouts, exports, and tags, respectively, you can also use the `-i', `-o', `-e', and `-t' options in the modules file. For a more flexible way of giving notifications to various users, which requires less in the way of keeping centralized scripts up to date, use the cvs watch add command (see section Telling CVS to notify you); this command is useful even if you are not using cvs watch on.

9. Handling binary files

The most common use for CVS is to store text files. With text files, CVS can merge revisions, display the differences between revisions in a human-visible fashion, and other such operations. However, if you are willing to give up a few of these abilities, CVS can store binary files. For example, one might store a web site in CVS including both text files and binary images.

9.1 The issues with binary files

While the need to manage binary files may seem obvious if the files that you customarily work with are binary, putting them into version control does present some additional issues.

One basic function of version control is to show the differences between two revisions. For example, if someone else checked in a new version of a file, you may wish to look at what they changed and determine whether their changes are good. For text files, CVS provides this functionality via the cvs diff command. For binary files, it may be possible to extract the two revisions and then compare them with a tool external to CVS (for example, word processing software often has such a feature). If there is no such tool, one must track changes via other mechanisms, such as urging people to write good log messages, and hoping that the changes they actually made were the changes that they intended to make.

Another ability of a version control system is the ability to merge two revisions. For CVS this happens in two contexts. The first is when users make changes in separate working directories (see section Multiple developers). The second is when one merges explicitly with the `update -j' command (see section Branching and merging).

In the case of text files, CVS can merge changes made independently, and signal a conflict if the changes conflict. With binary files, the best that CVS can do is present the two different copies of the file, and leave it to the user to resolve the conflict. The user may choose one copy or the other, or may run an external merge tool which knows about that particular file format, if one exists. Note that having the user merge relies primarily on the user to not accidentally omit some changes, and thus is potentially error prone.

If this process is thought to be undesirable, the best choice may be to avoid merging. To avoid the merges that result from separate working directories, see the discussion of reserved checkouts (file locking) in Multiple developers. To avoid the merges resulting from branches, restrict use of branches.

9.2 How to store binary files

There are two issues with using CVS to store binary files. The first is that CVS by default converts line endings between the canonical form in which they are stored in the repository (linefeed only), and the form appropriate to the operating system in use on the client (for example, carriage return followed by line feed for Windows NT).

The second is that a binary file might happen to contain data which looks like a keyword (see section Keyword substitution), so keyword expansion must be turned off.

The `-kb' option available with some CVS commands insures that neither line ending conversion nor keyword expansion will be done.

Here is an example of how you can create a new file using the `-kb' flag:

$ echo '$Id$' > kotest
$ cvs add -kb -m"A test file" kotest
$ cvs ci -m"First checkin; contains a keyword" kotest

If a file accidentally gets added without `-kb', one can use the cvs admin command to recover. For example:

$ echo '$Id$' > kotest
$ cvs add -m"A test file" kotest
$ cvs ci -m"First checkin; contains a keyword" kotest
$ cvs admin -kb kotest
$ cvs update -A kotest
# For non-unix systems:
# Copy in a good copy of the file from outside CVS
$ cvs commit -m "make it binary" kotest

When you check in the file `kotest' the file is not preserved as a binary file, because you did not check it in as a binary file. The cvs admin -kb command sets the default keyword substitution method for this file, but it does not alter the working copy of the file that you have. If you need to cope with line endings (that is, you are using CVS on a non-unix system), then you need to check in a new copy of the file, as shown by the cvs commit command above. On unix, the cvs update -A command suffices. (Note that you can use cvs log to determine the default keyword substitution method for a file and cvs status to determine the keyword substitution method for a working copy.)

However, in using cvs admin -k to change the keyword expansion, be aware that the keyword expansion mode is not version controlled. This means that, for example, that if you have a text file in old releases, and a binary file with the same name in new releases, CVS provides no way to check out the file in text or binary mode depending on what version you are checking out. There is no good workaround for this problem.

You can also set a default for whether cvs add and cvs import treat a file as binary based on its name; for example you could say that files who names end in `.exe' are binary. See section The cvswrappers file. There is currently no way to have CVS detect whether a file is binary based on its contents. The main difficulty with designing such a feature is that it is not clear how to distinguish between binary and non-binary files, and the rules to apply would vary considerably with the operating system.

10. Multiple developers

When more than one person works on a software project things often get complicated. Often, two people try to edit the same file simultaneously. One solution, known as file locking or reserved checkouts, is to allow only one person to edit each file at a time. This is the only solution with some version control systems, including RCS and SCCS. Currently the usual way to get reserved checkouts with CVS is the cvs admin -l command (see section admin options). This is not as nicely integrated into CVS as the watch features, described below, but it seems that most people with a need for reserved checkouts find it adequate. It also may be possible to use the watches features described below, together with suitable procedures (not enforced by software), to avoid having two people edit at the same time.

The default model with CVS is known as unreserved checkouts. In this model, developers can edit their own working copy of a file simultaneously. The first person that commits his changes has no automatic way of knowing that another has started to edit it. Others will get an error message when they try to commit the file. They must then use CVS commands to bring their working copy up to date with the repository revision. This process is almost automatic.

CVS also supports mechanisms which facilitate various kinds of communication, without actually enforcing rules like reserved checkouts do.

The rest of this chapter describes how these various models work, and some of the issues involved in choosing between them.

10.1 File status

Based on what operations you have performed on a checked out file, and what operations others have performed to that file in the repository, one can classify a file in a number of states. The states, as reported by the status command, are:

Up-to-date

The file is identical with the latest revision in the repository for the branch in use.

Locally Modified

You have edited the file, and not yet committed your changes.

Locally Added

You have added the file with add, and not yet committed your changes.

Locally Removed

You have removed the file with remove, and not yet committed your changes.

Needs Checkout

Someone else has committed a newer revision to the repository. The name is slightly misleading; you will ordinarily use update rather than checkout to get that newer revision.

Needs Patch

Like Needs Checkout, but the CVS server will send a patch rather than the entire file. Sending a patch or sending an entire file accomplishes the same thing.

Needs Merge

Someone else has committed a newer revision to the repository, and you have also made modifications to the file.

Unresolved Conflict

A file with the same name as this new file has been added to the repository from a second workspace. This file will need to be moved out of the way to allow an update to complete.

File had conflicts on merge

This is like Locally Modified, except that a previous update command gave a conflict. If you have not already done so, you need to resolve the conflict as described in Conflicts example.

Unknown

CVS doesn't know anything about this file. For example, you have created a new file and have not run add.

To help clarify the file status, status also reports the Working revision which is the revision that the file in the working directory derives from, and the Repository revision which is the latest revision in the repository for the branch in use.

The options to status are listed in Quick reference to CVS commands. For information on its Sticky tag and Sticky date output, see Sticky tags. For information on its Sticky options output, see the `-k' option in update options.

You can think of the status and update commands as somewhat complementary. You use update to bring your files up to date, and you can use status to give you some idea of what an update would do (of course, the state of the repository might change before you actually run update). In fact, if you want a command to display file status in a more brief format than is displayed by the status command, you can invoke

$ cvs -n -q update

The `-n' option means to not actually do the update, but merely to display statuses; the `-q' option avoids printing the name of each directory. For more information on the update command, and these options, see Quick reference to CVS commands.

10.2 Bringing a file up to date

When you want to update or merge a file, use the cvs update -d command. For files that are not up to date this is roughly equivalent to a checkout command: the newest revision of the file is extracted from the repository and put in your working directory. The -d option, not necessary with checkout, tells CVS that you wish it to create directories added by other developers.

Your modifications to a file are never lost when you use update. If no newer revision exists, running update has no effect. If you have edited the file, and a newer revision is available, CVS will merge all changes into your working copy.

For instance, imagine that you checked out revision 1.4 and started editing it. In the meantime someone else committed revision 1.5, and shortly after that revision 1.6. If you run update on the file now, CVS will incorporate all changes between revision 1.4 and 1.6 into your file.

If any of the changes between 1.4 and 1.6 were made too close to any of the changes you have made, an overlap occurs. In such cases a warning is printed, and the resulting file includes both versions of the lines that overlap, delimited by special markers. See section update—Bring work tree in sync with repository, for a complete description of the update command.

10.3 Conflicts example

Suppose revision 1.4 of `driver.c' contains this:

#include <stdio.h>

void main()
{
    parse();
    if (nerr == 0)
        gencode();
    else
        fprintf(stderr, "No code generated.\n");
    exit(nerr == 0 ? 0 : 1);
}

Revision 1.6 of `driver.c' contains this:

#include <stdio.h>

int main(int argc,
         char **argv)
{
    parse();
    if (argc != 1)
    {
        fprintf(stderr, "tc: No args expected.\n");
        exit(1);
    }
    if (nerr == 0)
        gencode();
    else
        fprintf(stderr, "No code generated.\n");
    exit(!!nerr);
}

Your working copy of `driver.c', based on revision 1.4, contains this before you run `cvs update':

#include <stdlib.h>
#include <stdio.h>

void main()
{
    init_scanner();
    parse();
    if (nerr == 0)
        gencode();
    else
        fprintf(stderr, "No code generated.\n");
    exit(nerr == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
}

You run `cvs update':

$ cvs update driver.c
RCS file: /usr/local/cvsroot/yoyodyne/tc/driver.c,v
retrieving revision 1.4
retrieving revision 1.6
Merging differences between 1.4 and 1.6 into driver.c
rcsmerge warning: overlaps during merge
cvs update: conflicts found in driver.c
C driver.c

CVS tells you that there were some conflicts. Your original working file is saved unmodified in `.#driver.c.1.4'. The new version of `driver.c' contains this:

#include <stdlib.h>
#include <stdio.h>

int main(int argc,
         char **argv)
{
    init_scanner();
    parse();
    if (argc != 1)
    {
        fprintf(stderr, "tc: No args expected.\n");
        exit(1);
    }
    if (nerr == 0)
        gencode();
    else
        fprintf(stderr, "No code generated.\n");
<<<<<<< driver.c
    exit(nerr == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
=======
    exit(!!nerr);
>>>>>>> 1.6
}

Note how all non-overlapping modifications are incorporated in your working copy, and that the overlapping section is clearly marked with `<<<<<<<', `=======' and `>>>>>>>'.

You resolve the conflict by editing the file, removing the markers and the erroneous line. Suppose you end up with this file:

#include <stdlib.h>
#include <stdio.h>

int main(int argc,
         char **argv)
{
    init_scanner();
    parse();
    if (argc != 1)
    {
        fprintf(stderr, "tc: No args expected.\n");
        exit(1);
    }
    if (nerr == 0)
        gencode();
    else
        fprintf(stderr, "No code generated.\n");
    exit(nerr == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
}

You can now go ahead and commit this as revision 1.7.

$ cvs commit -m "Initialize scanner. Use symbolic exit values." driver.c
Checking in driver.c;
/usr/local/cvsroot/yoyodyne/tc/driver.c,v  <--  driver.c
new revision: 1.7; previous revision: 1.6
done

For your protection, CVS will refuse to check in a file if a conflict occurred and you have not resolved the conflict. Currently to resolve a conflict, you must change the timestamp on the file. In previous versions of CVS, you also needed to insure that the file contains no conflict markers. Because your file may legitimately contain conflict markers (that is, occurrences of `>>>>>>> ' at the start of a line that don't mark a conflict), the current version of CVS will print a warning and proceed to check in the file.

If you use release 1.04 or later of pcl-cvs (a GNU Emacs front-end for CVS) you can use an Emacs package called emerge to help you resolve conflicts. See the documentation for pcl-cvs.

10.4 Informing others about commits

It is often useful to inform others when you commit a new revision of a file. The `-i' option of the `modules' file, or the `loginfo' file, can be used to automate this process. See section The modules file. See section Loginfo. You can use these features of CVS to, for instance, instruct CVS to mail a message to all developers, or post a message to a local newsgroup.

10.5 Several developers simultaneously attempting to run CVS

If several developers try to run CVS at the same time, one may get the following message:

[11:43:23] waiting for bach's lock in /usr/local/cvsroot/foo

CVS will try again every 30 seconds, and either continue with the operation or print the message again, if it still needs to wait. If a lock seems to stick around for an undue amount of time, find the person holding the lock and ask them about the cvs command they are running. If they aren't running a cvs command, look in the repository directory mentioned in the message and remove files which they own whose names start with `#cvs.rfl', `#cvs.wfl', or `#cvs.lock'.

Note that these locks are to protect CVS's internal data structures and have no relationship to the word lock in the sense used by RCS—which refers to reserved checkouts (see section Multiple developers).

Any number of people can be reading from a given repository at a time; only when someone is writing do the locks prevent other people from reading or writing.

One might hope for the following property:

If someone commits some changes in one cvs command, then an update by someone else will either get all the changes, or none of them.

but CVS does not have this property. For example, given the files

a/one.c
a/two.c
b/three.c
b/four.c

if someone runs

cvs ci a/two.c b/three.c

and someone else runs cvs update at the same time, the person running update might get only the change to `b/three.c' and not the change to `a/two.c'.

10.6 Mechanisms to track who is editing files

For many groups, use of CVS in its default mode is perfectly satisfactory. Users may sometimes go to check in a modification only to find that another modification has intervened, but they deal with it and proceed with their check in. Other groups prefer to be able to know who is editing what files, so that if two people try to edit the same file they can choose to talk about who is doing what when rather than be surprised at check in time. The features in this section allow such coordination, while retaining the ability of two developers to edit the same file at the same time.

For maximum benefit developers should use cvs edit (not chmod) to make files read-write to edit them, and cvs release (not rm) to discard a working directory which is no longer in use, but CVS is not able to enforce this behavior.

10.6.1 Telling CVS to watch certain files

To enable the watch features, you first specify that certain files are to be watched.

Command: cvs watch on [-lR] [files]…

Specify that developers should run cvs edit before editing files. CVS will create working copies of files read-only, to remind developers to run the cvs edit command before working on them.

If files includes the name of a directory, CVS arranges to watch all files added to the corresponding repository directory, and sets a default for files added in the future; this allows the user to set notification policies on a per-directory basis. The contents of the directory are processed recursively, unless the -l option is given. The -R option can be used to force recursion if the -l option is set in `~/.cvsrc' (see section Default options and the ~/.cvsrc file).

If files is omitted, it defaults to the current directory.

Command: cvs watch off [-lR] [files]…

Do not create files read-only on checkout; thus, developers will not be reminded to use cvs edit and cvs unedit.

The files and options are processed as for cvs watch on.

10.6.2 Telling CVS to notify you

You can tell CVS that you want to receive notifications about various actions taken on a file. You can do this without using cvs watch on for the file, but generally you will want to use cvs watch on, to remind developers to use the cvs edit command.

Command: cvs watch add [-lR] [-a action]… [files]…

Add the current user to the list of people to receive notification of work done on files.

The -a option specifies what kinds of events CVS should notify the user about. action is one of the following:

edit

Another user has applied the cvs edit command (described below) to a watched file.

commit

Another user has committed changes to one of the named files.

unedit

Another user has abandoned editing a file (other than by committing changes). They can do this in several ways, by:

• applying the cvs unedit command (described below) to the file
• applying the cvs release command (see section release—Indicate that a Module is no longer in use) to the file's parent directory (or recursively to a directory more than one level up)
• deleting the file and allowing cvs update to recreate it

all

All of the above.

none

None of the above. (This is useful with cvs edit, described below.)

The -a option may appear more than once, or not at all. If omitted, the action defaults to all.

The files and options are processed as for cvs watch on.

Command: cvs watch remove [-lR] [-a action]… [files]…

Remove a notification request established using cvs watch add; the arguments are the same. If the -a option is present, only watches for the specified actions are removed.

When the conditions exist for notification, CVS calls the `notify' administrative file. Edit `notify' as one edits the other administrative files (see section The administrative files). This file follows the usual conventions for administrative files (see section The common syntax), where each line is a regular expression followed by a command to execute. The command should contain a single occurrence of `%s' which will be replaced by the user to notify; the rest of the information regarding the notification will be supplied to the command on standard input. The standard thing to put in the notify file is the single line:

ALL mail %s -s "CVS notification"

This causes users to be notified by electronic mail.

Note that if you set this up in the straightforward way, users receive notifications on the server machine. One could of course write a `notify' script which directed notifications elsewhere, but to make this easy, CVS allows you to associate a notification address for each user. To do so create a file `users' in `CVSROOT' with a line for each user in the format user:value. Then instead of passing the name of the user to be notified to `notify', CVS will pass the value (normally an email address on some other machine).

CVS does not notify you for your own changes. Currently this check is done based on whether the user name of the person taking the action which triggers notification matches the user name of the person getting notification. In fact, in general, the watches features only track one edit by each user. It probably would be more useful if watches tracked each working directory separately, so this behavior might be worth changing.

10.6.3 How to edit a file which is being watched

Since a file which is being watched is checked out read-only, you cannot simply edit it. To make it read-write, and inform others that you are planning to edit it, use the cvs edit command. Some systems call this a checkout, but CVS uses that term for obtaining a copy of the sources (see section Getting the source), an operation which those systems call a get or a fetch.

Command: cvs edit [-lR] [-a action]… [files]…

Prepare to edit the working files files. CVS makes the files read-write, and notifies users who have requested edit notification for any of files.

The cvs edit command accepts the same options as the cvs watch add command, and establishes a temporary watch for the user on files; CVS will remove the watch when files are unedited or committed. If the user does not wish to receive notifications, she should specify -a none.

The files and the options are processed as for the cvs watch commands.

Normally when you are done with a set of changes, you use the cvs commit command, which checks in your changes and returns the watched files to their usual read-only state. But if you instead decide to abandon your changes, or not to make any changes, you can use the cvs unedit command.

Command: cvs unedit [-lR] [files]…

Abandon work on the working files files, and revert them to the repository versions on which they are based. CVS makes those files read-only for which users have requested notification using cvs watch on. CVS notifies users who have requested unedit notification for any of files.

The files and options are processed as for the cvs watch commands.

If watches are not in use, the unedit command probably does not work, and the way to revert to the repository version is with the command cvs update -C file (see section update—Bring work tree in sync with repository). The meaning is not precisely the same; the latter may also bring in some changes which have been made in the repository since the last time you updated.

When using client/server CVS, you can use the cvs edit and cvs unedit commands even if CVS is unable to successfully communicate with the server; the notifications will be sent upon the next successful CVS command.

10.6.4 Information about who is watching and editing

Command: cvs watchers [-lR] [files]…

List the users currently watching changes to files. The report includes the files being watched, and the mail address of each watcher.

The files and options are processed as for the cvs watch commands.

Command: cvs editors [-lR] [files]…

List the users currently working on files. The report includes the mail address of each user, the time when the user began working with the file, and the host and path of the working directory containing the file.

The files and options are processed as for the cvs watch commands.

10.6.5 Using watches with old versions of CVS

If you use the watch features on a repository, it creates `CVS' directories in the repository and stores the information about watches in that directory. If you attempt to use CVS 1.6 or earlier with the repository, you get an error message such as the following (all on one line):

cvs update: cannot open CVS/Entries for reading:
No such file or directory

and your operation will likely be aborted. To use the watch features, you must upgrade all copies of CVS which use that repository in local or server mode. If you cannot upgrade, use the watch off and watch remove commands to remove all watches, and that will restore the repository to a state which CVS 1.6 can cope with.

10.7 Choosing between reserved or unreserved checkouts

Reserved and unreserved checkouts each have pros and cons. Let it be said that a lot of this is a matter of opinion or what works given different groups' working styles, but here is a brief description of some of the issues. There are many ways to organize a team of developers. CVS does not try to enforce a certain organization. It is a tool that can be used in several ways.

Reserved checkouts can be very counter-productive. If two persons want to edit different parts of a file, there may be no reason to prevent either of them from doing so. Also, it is common for someone to take out a lock on a file, because they are planning to edit it, but then forget to release the lock.

People, especially people who are familiar with reserved checkouts, often wonder how often conflicts occur if unreserved checkouts are used, and how difficult they are to resolve. The experience with many groups is that they occur rarely and usually are relatively straightforward to resolve.

The rarity of serious conflicts may be surprising, until one realizes that they occur only when two developers disagree on the proper design for a given section of code; such a disagreement suggests that the team has not been communicating properly in the first place. In order to collaborate under any source management regimen, developers must agree on the general design of the system; given this agreement, overlapping changes are usually straightforward to merge.

In some cases unreserved checkouts are clearly inappropriate. If no merge tool exists for the kind of file you are managing (for example word processor files or files edited by Computer Aided Design programs), and it is not desirable to change to a program which uses a mergeable data format, then resolving conflicts is going to be unpleasant enough that you generally will be better off to simply avoid the conflicts instead, by using reserved checkouts.

The watches features described above in Mechanisms to track who is editing files can be considered to be an intermediate model between reserved checkouts and unreserved checkouts. When you go to edit a file, it is possible to find out who else is editing it. And rather than having the system simply forbid both people editing the file, it can tell you what the situation is and let you figure out whether it is a problem in that particular case or not. Therefore, for some groups it can be considered the best of both the reserved checkout and unreserved checkout worlds.

11. Revision management

If you have read this far, you probably have a pretty good grasp on what CVS can do for you. This chapter talks a little about things that you still have to decide.

If you are doing development on your own using CVS you could probably skip this chapter. The questions this chapter takes up become more important when more than one person is working in a repository.

11.1 When to commit?

Your group should decide which policy to use regarding commits. Several policies are possible, and as your experience with CVS grows you will probably find out what works for you.

If you commit files too quickly you might commit files that do not even compile. If your partner updates his working sources to include your buggy file, he will be unable to compile the code. On the other hand, other persons will not be able to benefit from the improvements you make to the code if you commit very seldom, and conflicts will probably be more common.

It is common to only commit files after making sure that they can be compiled. Some sites require that the files pass a test suite. Policies like this can be enforced using the commitinfo file (see section Commitinfo), but you should think twice before you enforce such a convention. By making the development environment too controlled it might become too regimented and thus counter-productive to the real goal, which is to get software written.

12. Keyword substitution

As long as you edit source files inside a working directory you can always find out the state of your files via `cvs status' and `cvs log'. But as soon as you export the files from your development environment it becomes harder to identify which revisions they are.

CVS can use a mechanism known as keyword substitution (or keyword expansion) to help identifying the files. Embedded strings of the form $keyword$ and $keyword:…$ in a file are replaced with strings of the form $keyword:value$ whenever you obtain a new revision of the file.

12.1 Keyword List

This is a list of the keywords:

$Author$

The login name of the user who checked in the revision.

$Date$

The date and time (UTC) the revision was checked in.

$Header$

A standard header containing the full pathname of the RCS file, the revision number, the date (UTC), the author, the state, and the locker (if locked). Files will normally never be locked when you use CVS.

$Id$

Same as $Header$, except that the RCS filename is without a path.

$Name$

Tag name used to check out this file. The keyword is expanded only if one checks out with an explicit tag name. For example, when running the command cvs co -r first, the keyword expands to `Name: first'.

$Locker$

The login name of the user who locked the revision (empty if not locked, which is the normal case unless cvs admin -l is in use).

$Log$

The log message supplied during commit, preceded by a header containing the RCS filename, the revision number, the author, and the date (UTC). Existing log messages are not replaced. Instead, the new log message is inserted after $Log:…$. Each new line is prefixed with the same string which precedes the $Log keyword. For example, if the file contains:

/* Here is what people have been up to: * * $Log: frob.c,v $ * Revision 1.1 1997/01/03 14:23:51 joe * Add the superfrobnicate option * */

then additional lines which are added when expanding the $Log keyword will be preceded by ` * '. Unlike previous versions of CVS and RCS, the comment leader from the RCS file is not used. The $Log keyword is useful for accumulating a complete change log in a source file, but for several reasons it can be problematic. See section Problems with the $Log$ keyword..

$RCSfile$

The name of the RCS file without a path.

$Revision$

The revision number assigned to the revision.

$Source$

The full pathname of the RCS file.

$State$

The state assigned to the revision. States can be assigned with cvs admin -s—see admin options.

12.2 Using keywords

To include a keyword string you simply include the relevant text string, such as $Id$, inside the file, and commit the file. CVS will automatically (Or, more accurately, as part of the update run that automatically happens after a commit.) expand the string as part of the commit operation.

It is common to embed the $Id$ string in the source files so that it gets passed through to generated files. For example, if you are managing computer program source code, you might include a variable which is initialized to contain that string. Or some C compilers may provide a #pragma ident directive. Or a document management system might provide a way to pass a string through to generated files.

The ident command (which is part of the RCS package) can be used to extract keywords and their values from a file. This can be handy for text files, but it is even more useful for extracting keywords from binary files.

$ ident samp.c
samp.c:
     $Id: samp.c,v 1.5 1993/10/19 14:57:32 ceder Exp $
$ gcc samp.c
$ ident a.out
a.out:
     $Id: samp.c,v 1.5 1993/10/19 14:57:32 ceder Exp $

SCCS is another popular revision control system. It has a command, what, which is very similar to ident and used for the same purpose. Many sites without RCS have SCCS. Since what looks for the character sequence @(#) it is easy to include keywords that are detected by either command. Simply prefix the keyword with the magic SCCS phrase, like this:

static char *id="@(#) $Id: ab.c,v 1.5 1993/10/19 14:57:32 ceder Exp $";

12.3 Avoiding substitution

Keyword substitution has its disadvantages. Sometimes you might want the literal text string `$Author$' to appear inside a file without CVS interpreting it as a keyword and expanding it into something like `$Author: ceder $'.

There is unfortunately no way to selectively turn off keyword substitution. You can use `-ko' (see section Substitution modes) to turn off keyword substitution entirely.

In many cases you can avoid using keywords in the source, even though they appear in the final product. For example, the source for this manual contains `$@asis{}Author$' whenever the text `$Author$' should appear. In nroff and troff you can embed the null-character \& inside the keyword for a similar effect.

12.4 Substitution modes

Each file has a stored default substitution mode, and each working directory copy of a file also has a substitution mode. The former is set by the `-k' option to cvs add and cvs admin; the latter is set by the `-k' or `-A' options to cvs checkout or cvs update. cvs diff and cvs rdiff also have `-k' options. For some examples, see Handling binary files, and Merging and keywords.

The modes available are:

`-kkv'

Generate keyword strings using the default form, e.g. $Revision: 5.7 $ for the Revision keyword.

`-kkvl'

Like `-kkv', except that a locker's name is always inserted if the given revision is currently locked. The locker's name is only relevant if cvs admin -l is in use.

`-kk'

Generate only keyword names in keyword strings; omit their values. For example, for the Revision keyword, generate the string $Revision$ instead of $Revision: 5.7 $. This option is useful to ignore differences due to keyword substitution when comparing different revisions of a file (see section Merging and keywords).

`-ko'

Generate the old keyword string, present in the working file just before it was checked in. For example, for the Revision keyword, generate the string $Revision: 1.1 $ instead of $Revision: 5.7 $ if that is how the string appeared when the file was checked in.

`-kb'

Like `-ko', but also inhibit conversion of line endings between the canonical form in which they are stored in the repository (linefeed only), and the form appropriate to the operating system in use on the client. For systems, like unix, which use linefeed only to terminate lines, this is the same as `-ko'. For more information on binary files, see Handling binary files.

`-kv'

Generate only keyword values for keyword strings. For example, for the Revision keyword, generate the string 5.7 instead of $Revision: 5.7 $. This can help generate files in programming languages where it is hard to strip keyword delimiters like $Revision: $ from a string. However, further keyword substitution cannot be performed once the keyword names are removed, so this option should be used with care.

One often would like to use `-kv' with cvs export—see section export—Export sources from CVS, similar to checkout. But be aware that doesn't handle an export containing binary files correctly.

12.5 Problems with the $Log$ keyword.

The $Log$ keyword is somewhat controversial. As long as you are working on your development system the information is easily accessible even if you do not use the $Log$ keyword—just do a cvs log. Once you export the file the history information might be useless anyhow.

A more serious concern is that CVS is not good at handling $Log$ entries when a branch is merged onto the main trunk. Conflicts often result from the merging operation.

People also tend to "fix" the log entries in the file (correcting spelling mistakes and maybe even factual errors). If that is done the information from cvs log will not be consistent with the information inside the file. This may or may not be a problem in real life.

It has been suggested that the $Log$ keyword should be inserted last in the file, and not in the files header, if it is to be used at all. That way the long list of change messages will not interfere with everyday source file browsing.

13. Tracking third-party sources

If you modify a program to better fit your site, you probably want to include your modifications when the next release of the program arrives. CVS can help you with this task.

In the terminology used in CVS, the supplier of the program is called a vendor. The unmodified distribution from the vendor is checked in on its own branch, the vendor branch. CVS reserves branch 1.1.1 for this use.

When you modify the source and commit it, your revision will end up on the main trunk. When a new release is made by the vendor, you commit it on the vendor branch and copy the modifications onto the main trunk.

Use the import command to create and update the vendor branch. When you import a new file, the vendor branch is made the `head' revision, so anyone that checks out a copy of the file gets that revision. When a local modification is committed it is placed on the main trunk, and made the `head' revision.

13.1 Importing for the first time

Use the import command to check in the sources for the first time. When you use the import command to track third-party sources, the vendor tag and release tags are useful. The vendor tag is a symbolic name for the branch (which is always 1.1.1, unless you use the `-b branch' flag—see Multiple vendor branches.). The release tags are symbolic names for a particular release, such as `FSF_0_04'.

Note that import does not change the directory in which you invoke it. In particular, it does not set up that directory as a CVS working directory; if you want to work with the sources import them first and then check them out into a different directory (see section Getting the source).

Suppose you have the sources to a program called wdiff in a directory `wdiff-0.04', and are going to make private modifications that you want to be able to use even when new releases are made in the future. You start by importing the source to your repository:

$ cd wdiff-0.04
$ cvs import -m "Import of FSF v. 0.04" fsf/wdiff FSF_DIST WDIFF_0_04

The vendor tag is named `FSF_DIST' in the above example, and the only release tag assigned is `WDIFF_0_04'.

13.2 Updating with the import command

When a new release of the source arrives, you import it into the repository with the same import command that you used to set up the repository in the first place. The only difference is that you specify a different release tag this time:

$ tar xfz wdiff-0.05.tar.gz
$ cd wdiff-0.05
$ cvs import -m "Import of FSF v. 0.05" fsf/wdiff FSF_DIST WDIFF_0_05

WARNING: If you use a release tag that already exists in one of the repository archives, files removed by an import may not be detected.

For files that have not been modified locally, the newly created revision becomes the head revision. If you have made local changes, import will warn you that you must merge the changes into the main trunk, and tell you to use `checkout -j' to do so:

$ cvs checkout -jFSF_DIST:yesterday -jFSF_DIST wdiff

The above command will check out the latest revision of `wdiff', merging the changes made on the vendor branch `FSF_DIST' since yesterday into the working copy. If any conflicts arise during the merge they should be resolved in the normal way (see section Conflicts example). Then, the modified files may be committed.

However, it is much better to use the two release tags rather than using a date on the branch as suggested above:

$ cvs checkout -jWDIFF_0_04 -jWDIFF_0_05 wdiff

The reason this is better is that using a date, as suggested above, assumes that you do not import more than one release of a product per day. More importantly, using the release tags allows CVS to detect files that were removed between the two vendor releases and mark them for removal. Since import has no way to detect removed files, you should do a merge like this even if import doesn't tell you to.

13.3 Reverting to the latest vendor release

You can also revert local changes completely and return to the latest vendor release by changing the `head' revision back to the vendor branch on all files. For example, if you have a checked-out copy of the sources in `~/work.d/wdiff', and you want to revert to the vendor's version for all the files in that directory, you would type:

$ cd ~/work.d/wdiff
$ cvs admin -bFSF_DIST .

You must specify the `-bFSF_DIST' without any space after the `-b'. See section admin options.

13.4 How to handle binary files with cvs import

Use the `-k' wrapper option to tell import which files are binary. See section The cvswrappers file.

13.5 How to handle keyword substitution with cvs import

The sources which you are importing may contain keywords (see section Keyword substitution). For example, the vendor may use CVS or some other system which uses similar keyword expansion syntax. If you just import the files in the default fashion, then the keyword expansions supplied by the vendor will be replaced by keyword expansions supplied by your own copy of CVS. It may be more convenient to maintain the expansions supplied by the vendor, so that this information can supply information about the sources that you imported from the vendor.

To maintain the keyword expansions supplied by the vendor, supply the `-ko' option to cvs import the first time you import the file. This will turn off keyword expansion for that file entirely, so if you want to be more selective you'll have to think about what you want and use the `-k' option to cvs update or cvs admin as appropriate.

13.6 Multiple vendor branches

All the examples so far assume that there is only one vendor from which you are getting sources. In some situations you might get sources from a variety of places. For example, suppose that you are dealing with a project where many different people and teams are modifying the software. There are a variety of ways to handle this, but in some cases you have a bunch of source trees lying around and what you want to do more than anything else is just to all put them in CVS so that you at least have them in one place.

For handling situations in which there may be more than one vendor, you may specify the `-b' option to cvs import. It takes as an argument the vendor branch to import to. The default is `-b 1.1.1'.

For example, suppose that there are two teams, the red team and the blue team, that are sending you sources. You want to import the red team's efforts to branch 1.1.1 and use the vendor tag RED. You want to import the blue team's efforts to branch 1.1.3 and use the vendor tag BLUE. So the commands you might use are:

$ cvs import dir RED RED_1-0
$ cvs import -b 1.1.3 dir BLUE BLUE_1-5

Note that if your vendor tag does not match your `-b' option, CVS will not detect this case! For example,

$ cvs import -b 1.1.3 dir RED RED_1-0

Be careful; this kind of mismatch is sure to sow confusion or worse. I can't think of a useful purpose for the ability to specify a mismatch here, but if you discover such a use, don't. CVS is likely to make this an error in some future release.

14. How your build system interacts with CVS

As mentioned in the introduction, CVS does not contain software for building your software from source code. This section describes how various aspects of your build system might interact with CVS.

One common question, especially from people who are accustomed to RCS, is how to make their build get an up to date copy of the sources. The answer to this with CVS is two-fold. First of all, since CVS itself can recurse through directories, there is no need to modify your `Makefile' (or whatever configuration file your build tool uses) to make sure each file is up to date. Instead, just use two commands, first cvs -q update and then make or whatever the command is to invoke your build tool. Secondly, you do not necessarily want to get a copy of a change someone else made until you have finished your own work. One suggested approach is to first update your sources, then implement, build and test the change you were thinking of, and then commit your sources (updating first if necessary). By periodically (in between changes, using the approach just described) updating your entire tree, you ensure that your sources are sufficiently up to date.

One common need is to record which versions of which source files went into a particular build. This kind of functionality is sometimes called bill of materials or something similar. The best way to do this with CVS is to use the tag command to record which versions went into a given build (see section Tags–Symbolic revisions).

Using CVS in the most straightforward manner possible, each developer will have a copy of the entire source tree which is used in a particular build. If the source tree is small, or if developers are geographically dispersed, this is the preferred solution. In fact one approach for larger projects is to break a project down into smaller separately-compiled subsystems, and arrange a way of releasing them internally so that each developer need check out only those subsystems which they are actively working on.

Another approach is to set up a structure which allows developers to have their own copies of some files, and for other files to access source files from a central location. Many people have come up with some such a system using features such as the symbolic link feature found in many operating systems, or the VPATH feature found in many versions of make. One build tool which is designed to help with this kind of thing is Odin (see ftp://ftp.cs.colorado.edu/pub/distribs/odin).

15. Special Files

In normal circumstances, CVS works only with regular files. Every file in a project is assumed to be persistent; it must be possible to open, read and close them; and so on. CVS also ignores file permissions and ownerships, leaving such issues to be resolved by the developer at installation time. In other words, it is not possible to "check in" a device into a repository; if the device file cannot be opened, CVS will refuse to handle it. Files also lose their ownerships and permissions during repository transactions.

A. Guide to CVS commands

This appendix describes the overall structure of CVS commands, and describes some commands in detail (others are described elsewhere; for a quick reference to CVS commands, see section Quick reference to CVS commands).

A.1 Overall structure of CVS commands

The overall format of all CVS commands is:

cvs [ cvs_options ] cvs_command [ command_options ] [ command_args ]

cvs

The name of the CVS program.

cvs_options

Some options that affect all sub-commands of CVS. These are described below.

cvs_command

One of several different sub-commands. Some of the commands have aliases that can be used instead; those aliases are noted in the reference manual for that command. There are only two situations where you may omit `cvs_command': `cvs -H' elicits a list of available commands, and `cvs -v' displays version information on CVS itself.

command_options

Options that are specific for the command.

command_args

Arguments to the commands.

There is unfortunately some confusion between cvs_options and command_options. When given as a cvs_option, some options only affect some of the commands. When given as a command_option it may have a different meaning, and be accepted by more commands. In other words, do not take the above categorization too seriously. Look at the documentation instead.

A.2 CVS's exit status

CVS can indicate to the calling environment whether it succeeded or failed by setting its exit status. The exact way of testing the exit status will vary from one operating system to another. For example in a unix shell script the `$?' variable will be 0 if the last command returned a successful exit status, or greater than 0 if the exit status indicated failure.

If CVS is successful, it returns a successful status; if there is an error, it prints an error message and returns a failure status. The one exception to this is the cvs diff command. It will return a successful status if it found no differences, or a failure status if there were differences or if there was an error. Because this behavior provides no good way to detect errors, in the future it is possible that cvs diff will be changed to behave like the other CVS commands.