Contents

Chapter 1
Introduction

Darcs is a revision control system, along the lines of CVS or arch. That means that it keeps track of various revisions and branches of your project, allows for changes to propagate from one branch to another. Darcs is intended to be an “advanced” revision control system. Darcs has two particularly distinctive features which differ from other revision control systems: 1) each copy of the source is a fully functional branch, and 2) underlying darcs is a consistent and powerful theory of patches.

Every source tree a branch The primary simplifying notion of darcs is that every copy of your source code is a full repository. This is dramatically different from CVS, in which the normal usage is for there to be one central repository from which source code will be checked out. It is closer to the notion of arch, since the ‘normal’ use of arch is for each developer to create his own repository. However, darcs makes it even easier, since simply checking out the code is all it takes to create a new repository. This has several advantages, since you can harness the full power of darcs in any scratch copy of your code, without committing your possibly destabilizing changes to a central repository.

Theory of patches The development of a simplified theory of patches is what originally motivated me to create darcs. This patch formalism means that darcs patches have a set of properties, which make possible manipulations that couldn’t be done in other revision control systems. First, every patch is invertible. Secondly, sequential patches (i.e. patches that are created in sequence, one after the other) can be reordered, although this reordering can fail, which means the second patch is dependent on the first. Thirdly, patches which are in parallel (i.e. both patches were created by modifying identical trees) can be merged, and the result of a set of merges is independent of the order in which the merges are performed. This last property is critical to darcs’ philosophy, as it means that a particular version of a source tree is fully defined by the list of patches that are in it, i.e. there is no issue regarding the order in which merges are performed. For a more thorough discussion of darcs’ theory of patches, see Appendix A.

A simple advanced tool Besides being “advanced” as discussed above, darcs is actually also quite simple. Versioning tools can be seen as three layers. At the foundation is the ability to manipulate changes. On top of that must be placed some kind of database system to keep track of the changes. Finally, at the very top is some sort of distribution system for getting changes from one place to another. Really, only the first of these three layers is of particular interest to me, so the other two are done as simply as possible. At the database layer, darcs just has an ordered list of patches along with the patches themselves, each stored as an individual file. Darcs’ distribution system is strongly inspired by that of arch. Like arch, darcs uses a dumb server, typically apache or just a local or network file system when pulling patches. Unlike arch, darcs can only use scp to write to a remote file system. The recommended method is to send patches through gpg-signed email messages, which has the advantage of being mostly asynchronous.

Keeping track of changes rather than versions In the last paragraph, I explained revision control systems in terms of three layers. One can also look at them as having two distinct uses. One is to provide a history of previous versions. The other is to keep track of changes that are made to the repository, and to allow these changes to be merged and moved from one repository to another. These two uses are distinct, and almost orthogonal, in the sense that a tool can support one of the two uses optimally while providing no support for the other. Darcs is not intended to maintain a history of versions, although it is possible to kludge together such a revision history, either by making each new patch depend on all previous patches, or by tagging regularly. In a sense, this is what the tag feature is for, but the intention is that tagging will be used only to mark particularly notable versions (e.g. released versions, or perhaps versions that pass a time consuming test suite).

Other revision control systems are centered upon the job of keeping track of a history of versions, with the ability to merge changes being added as it was seen that this would be desirable. But the fundamental object remained the versions themselves.

In such a system, a patch (I am using patch here to mean an encapsulated set of changes) is uniquely determined by two trees. Merging changes that are in two trees consists of finding a common parent tree, computing the diffs of each tree with their parent, and then cleverly combining those two diffs and applying the combined diff to the parent tree, possibly at some point in the process allowing human intervention, to allow for fixing up problems in the merge such as conflicts.

In the world of darcs, the source tree is not the fundamental object, but rather the patch is the fundamental object. Rather than a patch being defined in terms of the difference between two trees, a tree is defined as the result of applying a given set of patches to an empty tree. Moreover, these patches may be reordered (unless there are dependencies between the patches involved) without changing the tree. As a result, there is no need to find a common parent when performing a merge. Or, if you like, their common parent is defined by the set of common patches, and may not correspond to any version in the version history.

One useful consequence of darcs’ patch-oriented philosophy is that since a patch need not be uniquely defined by a pair of trees (old and new), we can have several ways of representing the same change, which differ only in how they commute and what the result of merging them is. Of course, creating such a patch will require some sort of user input. This is a Good Thing, since the user creating the patch should be the one forced to think about what they really want to change, rather than the user merging the patch. An example of this is the token replace patch (See Section A.5). This feature make it possible to create a patch, for example, which changes every instance of the variable “stupidly_named_var” with “better_var_name”, while leaving “other_stupidly_named_var” untouched. When this patch is merged with any other patch involving the “stupidly_named_var”, that instance will also be modified to “better_var_name”. This is in contrast to a more conventional merging method which would not only fail to change new instances of the variable, but would also involves conflicts when merging with any patch that modifies lines containing the variable. By more using additional information about the programmer’s intent, darcs is thus able to make the process of changing a variable name the trivial task that it really is, which is really just a trivial search and replace, modulo tokenizing the code appropriately.

The patch formalism discussed in Appendix A is what makes darcs’ approach possible. In order for a tree to consist of a set of patches, there must be a deterministic merge of any set patches, regardless of the order in which they must be merged. This requires that one be able to reorder patches. While I don’t know that the patches are required to be invertible as well, my implementation certainly requires invertibility. In particular, invertibility is required to make use of Theorem 2, which is used extensively in the manipulation of merges.

1.1 Features

Record changes locally In darcs, the equivalent of a cvs “commit” is called record, because it doesn’t put the change into any remote or centralized repository. Changes are always recorded locally, meaning no net access is required in order to work on your project and record changes as you make them. Moreover, this means that there is no need for a separate “disconnected operation” mode.

Interactive records You can choose to perform an interactive record, in which case darcs will prompt you for each change you have made and ask if you wish to record it. Of course, you can tell darcs to record all the changes in a given file, or to skip all the changes in a given file, or go back to a previous change, or whatever. There is also an experimental graphical interface, which allows you to view and choose changes even more easily, and in whichever order you like.

Unrecord local changes As a corrolary to the “local” nature of the record operation, if a change hasn’t yet been published to the world--that is, if the local repository isn’t accessible by others--you can safely unrecord a change (even if it wasn’t the most recently recorded change) and then re-record it differently, for example if you forgot to add a file, introduced a bug or realized that what you recorded as a single change was really two separate changes.

Interactive everything else Most darcs commands support an interactive interface. The “revert” command, for example, which undoes unrecorded changes has the same interface as record, so you can easily revert just a single change. Pull, push, send and apply all allow you to view and interactively select which changes you wish to pull, push, send or apply.

Test suites Darcs has support for integrating a test suite with a repository. If you choose to use this, you can define a test command (e.g. “make check”) and have darcs run that command on a clean copy of the project either prior to recording a change or prior to applying changes--and to reject changes that cause the test to fail.

Any old server Darcs does not require a specialized server in order to make a repository available for read access. You can use http, ftp, or even just a plain old ssh server to access your darcs repository.

You decide write permissions Darcs doesn’t try to manage write access. That’s your business. Supported push methods include direct ssh access (if you’re willing to give direct ssh access away), using sudo to allow users who already have shell access to only apply changes to the repository, or verification of gpg-signed changes sent via email against a list of allowed keys. In addition, there is good support for submission of patches via email that are not automatically applied, but can easily be applied via a shell escape from a mail reader (this is how I deal with contributions to darcs).

Symmetric repositories Every darcs repository is created equal (well, with the exception of a “partial” repository, which doesn’t contain a full history...), and every working directory has an associated repository. As a result, there is a symmetry between “uploading” and “downloading” changes--you can use the same commands (push or pull) for either purpose.

CGI script Darcs has a CGI script that allows browsing of the repositories.

Portable Darcs runs on UNIX (or UNIX-like) systems (which includes MacOS X) as well as on Microsoft Windows.

File and directory moves Renames or moves of files and directories, of course are handled properly, so when you rename a file or move it to a different directory, its history is unbroken, and merges with repositories that don’t have the file renamed will work as expected.

Token replace You can use the “darcs replace” command to modify all occurences of a particular token (defined by a configurable set of characters that are allowed in “tokens”) in a file. This has the advantage that merges with changes that introduce new copies of the old token will have the effect of changing it to the new token--which comes in handy when changing a variable or function name that is used throughout a project.

Configurable defaults You can easily configure the default flags passed to any command on either a per-repository or a per-user basis or a combination thereof.

1.2 Switching from CVS

Darcs is very different from CVS.

CVS keeps version controlled data in a central repository, and requires that users check out a working directory whenever they wish to access the version-controlled sources. In order to modify the central repository, a user needs to have write access to the central repository; if he doesn’t, CVS merely becomes a tool to get the latest sources.

In darcs there is no distinction between working directories and repositories. In order to work on a project, a user makes a local copy of the repository he wants to work in; he may then harness the full power of version control locally. In order to distribute his changes, a user who has write access can push them to the remote repository; one who doesn’t can simply send them by e-mail in a format that makes them easy to apply on the remote system.

Darcs commands for CVS users Because of the different models used by arch and darcs, it is difficult to provide a complete equivalence between arch and darcs. A rough correspondence for the everyday commands follows:

1.3 Switching from arch

Although arch, like darcs, is a distributed system, and the two systems have many similarities (both require no special server, for example), their essential organization is very different.

Like CVS, arch keeps data in two types of data structures: repositories (called “archives”) and working directories. In order to modify a repository, one must first check out a corresponding working directory. This requires that users remember a number of different ways of pushing data around -- tla get, update, commit, archive-mirror and so on.

In darcs, on the other hand, there is no distinction between working directories and repositories, and just checking out your sources creates a local copy of a repository. This allows you to harness the full power of version control in any scratch copy of your sources, and also means that there are just two ways to push data around: darcs record, which stores edits into your local repository, and pull, which moves data between repositories. (darcs push is merely the opposite of pull; send and apply are just the two halves of push).

Darcs commands for arch users Because of the different models used by arch and darcs, it is difficult to provide a complete equivalence between arch and darcs. A rough correspondence for the everyday commands follows:

Chapter 2
Building darcs

This chapter should walk you through the steps necesary to build darcs for yourself. There are in general two ways to build darcs. One is for building released versions from tarballs, and the other is to build the latest and greatest darcs, from the darcs repo itself.

Please let me know if you have any problems building darcs, or don’t have problems described in this chapter and think there’s something obsolete here, so I can keep this page up-to-date.

Prerequisites To build darcs you will need to have ghc, the Glorious Glasgow Haskell Compiler. You should have at the very minimum version 5.04.2, but I myself only test it with 5.04.3. There are plans to phase out support for ghc5, so moving up to 6.0 is a good idea. It is a good idea (but not required) to have a recent version of libcurl installed. If not, you will at least need to have either wget or curl installed if you want to be able to grab repos remotely over normal network protocols (ftp or http). You also might want to have scp available if you want to grab your repos over ssh...

To send patches, you will also need to have a working /usr/sbin/sendmail or /usr/lib/sendmail, which is provided by most mail transport agents, and is generally available on linux and BSD systems. It’s also there on MacOS X. However, if you don’t have this, it won’t stop you from building darcs.

To use the diff command of darcs, a diff program supporting options -r (recursive diff) and -N (show new files as differences against an empty file) is required. The configure script will look for gdiff, gnudiff and diff in this order. You can force the use of another program by setting the DIFF environment variable before running configure.

To rebuild the documentation (which should not be necesary since it is included in html form with the tarballs), you will need to have latex installed, as well as latex2html if you want to build it in html form.

Building on MacOS X To build on MacOS X, you will need the Apple Developer Tools and the ghc 6.0 package installed. There seems to be a problem with ghc and Apple’s gcc 3.3, so you may need to run

Another recently discovered problem is that on MacOS, the curl-config from libcurl versions prior to 7.11.0 gives incorrect link flags, so you probably will want to make sure you have version 7.11.0 installed. I suspect one could also simply edit the curl-config of an older install to remove the offending “-arch i386”, but haven’t tried this.

Building on Microsoft Windows I don’t have Windows and don’t know how to build darcs on it. It has been done, but I may have broken it recently... If you try building darcs on windows, whether you succeed or fail, please post on the darcs-users mailing list so we’ll know how it went.

Building from tarball If you get darcs from a tarball, the procedure (after unpacking the tarball itself) is as follows:

There are options to configure that you may want to check out with

Building darcs from the repository To build the latest darcs from its repository, you will first need a working copy of darcs. You can get darcs using:

The repository doesn’t hold automatically generated files, which include the configure script and the HTML documentation. The makefile is smart enough to call autoconf to generate the configure script (and actually also to run the configure script), so you can just run

If you want to tweak the configure options, you’ll need to run ./configure yourself after the make, and then run make again.

Submitting patches to darcs I know, this doesn’t really belong in this chapter, but if you’re using the repository version of darcs it’s really easy to submit patches to me using darcs. In fact, even if you don’t know any haskell, you could submit fixes or additions to this document (by editing building_darcs.tex) based on your experience building darcs...

To do so, just record your changes (which you made in the darcs repository)

Chapter 3
Getting started

This chapter will lead you through an example use of darcs, which hopefully will allow you to get started using darcs with your project.

3.1 Creating your repository

Creating your repository in the first place just involves telling darcs to create the special directory (called _darcs) in your project tree, which will hold the revision information. This is done by simply calling from the root directory of your project:

3.2 Making changes

Now that we have created our repository, make a change to one or more of your files. After making the modification run:

Let’s say you have now made a change to your project. The next thing to do is to record a patch. Recording a patch consists of grouping together a set of related changes, and giving them a name. It also tags the patch with the date it was recorded and your email address.

To record a patch simply type:

You can now rerun whatsnew, and see that indeed the changes you have recorded are no longer marked as new.

3.3 Making your repository visible to others

How do you let the world know about these wonderful changes? Obviously, they must be able to see your repository. Currently the easiest way to do this is typically by http using any web server. The recommended way to do this (using apache in a UNIX environment) is to create a directory called /var/www/repos, and then put a symlink to your repo there:

As long as you’re running a web server and making your repo available to the world, you may as well make it easy for people to see what changes you’ve made. You can do this by running make installserver, which installs the program darcs_cgi at /usr/lib/cgi-bin/darcs. You also will need to create a cache directory named /var/cache/darcs_cgi, and make sure the owner of that directory is the same user that your web server runs its cgi scripts as. For me, this is www-data. Now your friends and enemies should be able to easily browse your repos by pointing their web browsers at http://your.server.org/cgi-bin/darcs. You can read more about this interface in Chapter 6.

3.4 Getting changes made to another repository

Ok, so I can now browse your repository using my web browser... so what? How do I get your changes into my repository, where they can do some good? It couldn’t be easier. I just cd into my repository, and there type:

Of course, maybe I don’t even have a copy of your repository. In that case I’d want to do a

Get, pull and push also work over ssh. Ssh-paths are of the same form accepted by scp, namely [username@]host:/path/to/repository.

3.5 Moving patches from one repo to another

Darcs is flexible as to how you move patches from one repo to another. This section will introduce all the ways you can get patches from one place to another, starting with the simplest and moving to the most complicated.

3.5.1 All pulls

The simplest method is the “all-pull” method. This involves making each repository readable (via http, ftp, nfs-mounted disk, whatever), and you run ‘darcs pull’ in the repo you want to move the patch to. This is nice, as it doesn’t require you to give write access to anyone else, and is reasonably simple.

3.5.2 Send and apply manually

Sometimes you have a machine on which it is not convenient to set up a web server, perhaps because it’s behind a firewall or perhaps for security reasons, or because it is often turned off. In this case you can use darcs send from that computer to generate a patch bundle destined for another repository. You can either let darcs email the patch for you, or save it as a file and transfer it by hand. Then in the destination repository you (or the owner of that repo) run darcs apply to apply the patches contained in the bundle. This is also quite a simple method since, like the all-pull method, it doesn’t require that you give anyone write access to your repository. But it’s less convenient, since you have to keep track of the patch bundle (in the email, or whatever).

If you use the send and apply method with email, you’ll probably want to create a _darcs/prefs/email file containing your email address. This way anyone who sends to your repository will automatically send the patch bundle to your email address.

If you receive many patches via email, you probably will benefit by running darcs apply directly from your mail program. I have in my .muttrc the following

3.5.3 Push

If you use ssh (and preferably also ssh-agent, so you won’t have to keep retyping your password), you can use the push method to transfer changes (using the scp protocol for communication). This method is again not very complicated, since you presumably already have the ssh permissions set up. Push can also be used when the target repository is local, in which case ssh isn’t needed. On the other hand, in this situation you could as easily run a pull, so there isn’t much benefit.

Note that you can use push to administer a multiple-user repository. You just need to create a user for the repository (or repositories), and give everyone with write access ssh access, perhaps via .ssh/authorized_keys. Then they run

3.5.4 Push --apply-as

Now we get more subtle. If you like the idea in the previous paragraph about creating a repository user to own a repository which is writable by a number of users, you have one other option.

Push --apply-as can run on either a local repository or one accessed via ssh, but uses sudo to run a darcs apply command (having created a patch bundle as in send) as another user. You can add the following line in your sudoers file to allow the users to apply their patches to a centralized repository:

3.5.5 Sending signed patches via email

Most of the previous methods are a bit clumsy if you don’t want to give each person with write access to a repo an account on your server. Darcs send can be configured to send a cryptographically signed patch via email. You can then set up your mail system to have darcs verify that patches were signed by an authorized user and apply them when a patch is received via email. The results of the apply can be returned to the user via email. Unsigned patches (or patches signed by unauthorized users) will be forwarded to the repository owner (or whoever you configure them to be forwarded to...).

This method is especially nice when combined with the --test option of darcs apply, since it allows you to run the test suite (assuming you have one) and reject patches that fail--and it’s all done on the server, so you can happily go on working on your development machine without slowdown while the server runs the tests.

Setting up darcs to run automatically in response to email is by far the most complicated way to get patches from one repo to another... so it’ll take a few sections to explain how to go about it.

Security considerations When you set up darcs to run apply on signed patches, you should assume that a user with write access can write to any file or directory that is writable by the user under which the apply process runs. Unless you specify the --no-test flag to darcs apply (and this is not the default), you are also allowing anyone with write access to that repository to run arbitrary code on your machine (since they can run a test suite--which they can modify however they like). This is quite a potential security hole.

For these reasons, if you don’t implicitly trust your users, it is recommended that you create a user for each repository to limit the damage an attacker can do via access to your repository. When considering who to trust, keep in mind that a security breach on any developer’s machine could give an attacker access to their private key and passphrase, and thus to your repository.

Installing necessary programs You also must install the following programs: gnupg, a mailer configured to receive mail (e.g. exim, sendmail or postfix), and a web server (usually apache). If you want to be able to browse your repository via the web you must also configure your web server to run cgi scripts and make sure the darcs cgi script was properly installed (via either a darcs-server package, or ‘make install-server’).

Setting up a repository with its own user To create a repository, as root run the ‘darcs-createrepo’. You will be prompted for the email address of the repository and the location of an existing copy of the repository. If your desired email is “myproject@my.url”, this will create a user named “myproject” with a home directory of /var/lib/darcs/repos/myproject. FIXME: I have no idea if the darcs-createrepo program will even run on any system other than debian. Success reports would be appreciated (or of course bug reports if it fails).

The “myproject” user will be configured to run the darcs patcher on any emails it receives. However, the patcher will bounce any emails which aren’t signed by a key in the /var/lib/darcs/repos/myproject/allowed_keys gpg keyring (which is empty). To give yourself access to this repository you will need to create a gpg key. If you don’t know about public key cryptography, take a look at the gnupg manual.

Granting access to a repository You create your gpg key by running (as your normal user):

You now need to export the public key so we can tell the patcher about it. You can do this with the following command (again as your normal user):

You should now be able to send a patch to the repository by running as your normal user, in a working copy of the repository:

Setting up a sendable repository using procmail If you don’t have root access on your machine, or perhaps simply don’t want to bother creating a separate user, you can set up a darcs repository using procmail to filter your mail. I will assume that you already use procmail to filter your email. If not, you will need to read up on it, or perhaps should use a different method for routing the email to darcs. To begin with, you must configure your repository so that a darcs send to your repository will know where to send the email. Do this by creating a file in /path/to/your/repo/_darcs/prefs called email containing your email address. As a trick (to be explained below), we will create the email address with “darcs repo” as your name, in an email address of the form “David Roundy <droundy@abridgegame.org>.”

The next step is to set up a gnupg keyring containing the public keys of people authorized to send to your repo. Here I’ll give a second way of going about this (see above for the first). This time I’ll assume you want to give me write access to your repository. You can do this by:

Finally, we add a few lines to your .procmailrc:

I find that I need the “umask 022” in order to keep procmail from setting the umask incorrectly, which causes the repository to no longer be world-readable.

Chapter 4
Configuring darcs

There are sevaral ways you can adjust darcs’ behavior to suit your needs. The first is to edit files in the _darcs/prefs/ directory of a repository. Such configuration only applies when working with that repository. To configure darcs on a per-user rather than per-repository basis (but with essentially the same methods), you can edit (or create) files in the ~/.darcs/ directory. Finally, the behavior of some darcs commands can be modified by setting appropriate environment variables.

4.1 prefs

The _darcs directory contains a prefs directory. This directory exists simply to hold user configuration settings specific to this repository. The contents of this directory are intended to be modifiable by the user, although in some cases a mistake in such a modification may cause darcs to behave strangely.

4.1.1 defaults

Default values for darcs commands can be configured on a per-repository basis by editing (and possibly creating) the _darcs/prefs/defaults file. Each line of this file has the following form:

For example, if your system clock is bizarre, you could instruct darcs to always ignore the file modification times by adding the following line to your _darcs/prefs/defaults file. (Note that this would have to be done for each repository!)

If you never want to run a test when recording to a particular repository (but still want to do so when running check on that repo), and like to name all your patches “Stupid patch”, you could use the following:

Also, a global preferences file can be created with the name .darcs/defaults in your home directory. Options present there will be added to the repository-specific preferences.

4.1.2 repos

The _darcs/prefs/repos file contains a list of repositories you have pulled from or pushed to, and is used for autocompletion of pull and push commands in bash. Feel free to delete any lines from this list that might get in there, or to delete the file as a whole.

4.1.3 author

The _darcs/prefs/author file contains the email address (or name) to be used as the author when patches are recorded in this repository. This file overrides the contents of the environment variables $DARCS_EMAIL and $EMAIL.

4.1.4 boring

The _darcs/prefs/boring file may contain a list of regular expressions describing files, such as object files, that you do not expect to add to your project. As an example, the boring file that I use with my darcs repository is:

Any file whose full path (such as manual/index.html) matches any of the boring regular expressions is considered boring. The boring file is used to filter the files provided to darcs add, to allow you to use a simple “darcs add newdir newdir/*” without accidentally adding a bunch of object files. It is also used when the --look-for-adds flag is given to whatsnew or record.

4.1.5 binaries

The _darcs/prefs/binaries file may contain a list of regular expressions describing files that should be treated as binary files rather than text files. You probably will want to have the binaries file under version control. To do this you can use darcs setpref to set the value “binariesfile” to the name of your desired binaries file (e.g. “darcs setpref binariesfile ./.binaries”, where .binaries is a file that has been darcs added to your repository). As with the boring file, you can also set up a ~/.darcs/binaries file if you like.

4.2 Environment variables

There are a few environment variables whose contents affect darcs’ behavior.

4.2.1 DARCS_EMAIL

The DARCS_EMAIL environment variable determines the “author” name used by darcs when recording if no _darcs/prefs/author exists. If DARCS_EMAIL is undefined, the contents of the EMAIL environment variable are used.

4.2.2 DARCS_EDITOR

When pulling up an editor (for example, when adding a long comment in record), darcs uses the contents of DARCS_EDITOR if it is defined. If not, it tries the contents of VISUAL, and if that isn’t defined (or fails for some reason), it tries EDITOR. If none of those environment variables are defined, darcs tries vi, emacs, emacs -nw and nano in that order.

4.2.3 DARCS_TMPDIR

If the environment variable DARCS_TMPDIR is defined, darcs will use that directory for its temporaries. Otherwise it will use TMPDIR, if that is defined, and if not that then /tmp and if /tmp doesn’t exist, it’ll put the temporaries in _darcs.

This is very helpful, for example, when recording with a test suite that uses MPI, in which case using /tmp to hold the test copy is no good, as /tmp isn’t shared over NFS and thus the mpirun call will fail, since the binary isn’t present on the compute nodes.

4.2.4 HOME

HOME is used to find the per-user prefs directory, which is located at $HOME/.darcs.

4.2.5 TERM

If darcs is compiled with libcurses support and support for color output, it uses the environment variable TERM to decide whether or not color is supported on the output terminal.

4.2.6 SSH_PORT

When using ssh, if the SSH_PORT enviroment variable is defined, darcs will use that port rather than the default ssh port (which is 22).

4.2.7 DARCS_SSH

The DARCS_SSH environment variable defines the command that darcs will use when asked to run ssh. This command is not interpereted by a shell, so you cannot use shell metacharacters, and the first word in the command must be the name of an executable located in your path.

4.2.8 DARCS_SCP

The DARCS_SCP environment variable defines the command that darcs will use when asked to run scp. Note that scp is how darcs accesses repositories whose URL is of the form user@foo.org:foo or foo.org:foo. This command is not interpereted by a shell, so you cannot use shell metacharacters, and the first word in the command must be the name of an executable located in your path.

Chapter 5
Darcs commands

The general format of a darcs command is

If you wish, you may use any unambiguous beginning of a command name as a shortcut: for darcs record, you could type darcs recor or darcs rec, but not darcs re since that could be confused with darcs replace, darcs revert and darcs remove.

Command overview Not all commands modify the ”patches” of your repository (that is, the named patches which other users can pull), some commands only affect the copy of the source tree you’re working on (your ”working directory”), and some affect both. This table summarizes what you should expect from each one and will hopefully serve as guide when you’re having doubts about which command to use.

5.1 Common options to darcs commands

5.1.1 Help

Every COMMAND accepts --help as an argument, which tells it to provide a bit of help. Among other things, this help always provides an accurate listing of the options available with that command, and is guaranteed never to be out of sync with the version of darcs you actually have installed (unlike this manual, which could be for an entirely different version of darcs).

5.1.2 Verbose

Most commands also accept the --verbose option, which tells darcs to provide additional output. The amount of verbosity varies from command to command.

5.1.3 Repository directory

Another common option is the --repodir option, which allows you to specify the directory of the repository in which to perform the command. This option is used with commands, such as whatsnew, that ordinarily would be performed within a repository directory, and allows you to use those commands without actually being in the repo directory when calling the command. This is useful when running darcs as a pipe, as might be the case when running apply from a mailer.

5.1.4 Match

The --match flag is accepted by commands that would prompt you for a patch that has already been recorded. It allows you to specify a matching criteria in order to filter the patch choices. I won’t discuss here the meaning of such a match, but simply the syntax which --match accepts.

Currently --match accepts just two primitive match types, although there are plans to expand it to match more patterns. Also, note that the syntax is still preliminary and subject to change.

The first match type accepts a regular expression which is checked against the patch name. The syntax is

The second match type matches the darcs hash for each patch:

The --match patter can include the logical operators &&, || and not, as well as grouping of patters with parentheses. For example

5.1.5 Ignore file modification times

Darcs optimizes its operations by keeping track of the modification times of your files. This dramatically speeds up commands such as whatsnew and record which would otherwise require reading every file in the repo and comparing it with a reference version. However, there are times when this can cause problems, such as when running a series of darcs commands from a script, in which case often a file will be modified twice in the same second, which can lead to the second modification going unnoticed. The solution to such predicaments is the --ignore-times option, which instructs darcs not to trust the file modification times, but instead to check each file’s contents explicitely.

5.1.6 Author

Several commands need to be able to identify you. Conventionally, you provide an email address for this purpose. The easiest way to do this is to define an environment variable EMAIL or DARCS_EMAIL (with the latter overriding the former). You can also override this using the --author flag to any command. Alternatively, you could set your email address on a per-repository basis using the “defaults” mechanism for “ALL” commands, as described in Appendix B. Or, you could specify the author on a per-repository basis using the _darcs/prefs/author file as described in section 4.1.3.

5.1.7 Patch compression

By default, darcs commands that write patches to disk will compress the patch files. If you don’t want this, you can choose the --dont-compress option, which causes darcs not to compress the patch file.

5.1.8 Optional graphical interface

Certain commands may have an optional graphical user interface. If such commands are supported, you can activate the graphical user interface by calling darcs with the --gui flag.

5.1.9 Resolution of conflicts

To resolve conflicts using an external tool, you need to specify a command to use via something like

Note that if you do use an external merge tool, most likely you will want to your defaults file _darcs/prefs/defaults (see 4.1.1) a line such as

Options apart from darcs commands Calling darcs with just “--help” as an argument gives a brief summary of what commands are available. The “--extended-help” option gives a more technical summary of what the commands actually do. Calling darcs with the flag “--version” tells you the version of darcs you are using. Similarly calling darcs with only “--commands” gives a simple list of available commands. This latter arrangement is primarily intended for the use of command-line autocompletion facilities, as are available in bash.

5.2 darcs initialize

No description available!

Generally you will only call initialize once for each project you work on, and calling it is just about the first thing you do. Just make sure you are in the main directory of the project, and initialize will set up all the directories and files darcs needs in order to start keeping track of revisions for your project.

initialize actually follows a very simple procedure. It simply creates the directories _darcs, _darcs/current and _darcs/patches, and then creates an empty file, _darcs/inventory. However, it is strongly recommended that you use darcs initialize to do this, as this procedure may change in a future version of darcs.

5.3 darcs add

Add needs to be called whenever you add a new file or directory to your project. Of course, it also needs to be called when you first create the project, to let darcs know which files should be kept track of.

Darcs will refuse to add a file or directory that differs from an existing one only in case. This is because the HFS+ file system used on under MacOS treats such files as being one and the same.

5.4 darcs remove

Remove should be called when you want to remove a file from your project, but don’t actually want to delete the file. Otherwise just delete the file or directory, and darcs will notice that it has been removed.

5.5 darcs whatsnew

Display unrecorded changes in the working directory.

What’s-new gives you a view of what changes you’ve made in your working directory that haven’t yet been recorded. The changes are displayed in darcs patch format. Darcs whatsnew will return a non-zero value if there are no changes, which can be useful if you just want to see in a script if anything has been modified. If you want to see some context around your changes, you can use the -u option, to get output similar to the unidiff format.

If you give one or more file or directory names as an argument to whatsnew, darcs whatsnew will output only changes to those files or to files in those directories.

5.6 darcs check

Check the repository for consistency.

Check verifies that the patches stored in the repository, when successively applied to an empty tree, properly recreate the stored current tree.

If you have a checkpoint of the repository (as is the case if you got the repo originally using darcs get --partial), by default darcs check will only verify the contents since the most recent checkpoint. You can change this behavior using the --complete flag.

If you like, you can configure your repository to be able to run a test suite of some sort. You can do this by using “setpref” to set the “test” value to be a command to run. e.g.

Normally darcs deletes the directory in which the test was run afterwards. Sometimes (especially when the test fails) you’d prefer to be able to be able to examine the test directory after the test is run. You can do this by specifying the --leave-test-directory flag. Alas, there is no way to make darcs leave the test directory only if the test fails. The opposite of --leave-test-directory is --remove-test-directory, which could come in handy if you choose to make --leave-test-directory the default (see section 4.1.1).

If you just want to check the consistency of your repository without running the test, you can call darcs check with the --no-test option.

5.7 darcs record

Record is used to name a set of changes and record the patch to the repository. If you provide one or more files or directories as additional arguments to record, you will only be prompted to changes in those files or directories. Each patch is given a name, which typically would consist of a brief description of the changes. This name is later used to describe the patch. The name must fit on one line (i.e. cannot have any embedded newlines). If you have more to say, stick it in the log. The patch is also flagged with the author of the change, taken by default from the DARCS_EMAIL environment variable, and if that doesn’t exist, from the EMAIL environment variable. The date on which the patch was recorded is also included. Currently there is no provision for keeping track of when a patch enters a given repository. Finally, each changeset should have a full log (which may be empty). This log is for detailed notes which are too lengthy to fit in the name. If you answer that you do want to create a comment file, darcs will open an editor so that you can enter the comment in. The choice of editor proceeds as follows. If one of the $DARCS_EDITOR, $VISUAL or $EDITOR environment variables is defined, its value is used (with precedence proceeding in the order listed). If not, “vi”, “emacs”, “emacs -nw” and “nano” are tried in that order.

There is a problem that shows up with at least some versions of vi which keep them from working properly with darcs. An (ugly) solution to this is to tell vi to use the tty as stdin and stdout:

If you wish, you may specify the patch name and log using the --logfile flag. If you do so, the first line of the specified file will be taken to be the patch name, and the remainder will be the “long comment”. This feature can be especially handy if you have a test that fails several times on the record (thus aborting the record), so you don’t have to type in the long comment multiple times.

Each patch may depend on any number of previous patches. If you choose to make your patch depend on a previous patch, that patch is required to be applied before your patch can be applied to a repo. This can be used, for example, if a piece of code requires a function to be defined, which was defined in an earlier patch.

If you want to manually define any dependencies for your patch, you can use the --ask-deps flag, and darcs will ask you for the patch’s dependencies.

If you configure darcs to run a test suite, darcs will run this test on the recorded repo to make sure it is valid. Darcs first creates a pristine copy of the source tree (in /tmp), then it runs the test, using its return value to decide if the record is valid. If it is not valid, the record will be aborted. This is a handy way to avoid making stupid mistakes like forgetting to ‘darcs add’ a new file. It also can be tediously slow, so there is an option (--no-test) to skip the test.

If you run record with the --pipe option, you will be prompted for the patch name, patch date and the long comment. The long comment will extend until the end of file of stdin is reached (ctrl-D on unixy systems). This interface is intended for scripting darcs, in particular for writing repository conversion scripts. The prompts are intended mostly as useful guide (since scripts won’t need them), to help you understand the format in which to provide the input.

5.8 darcs get

Get is used to get a local copy of a repository. I recommend using the --verbose flag to get, as this command can take a while, and with no feedback, that can be rather boring.

If the remote repo and the current directory are in the same filesystem and that filesystem supports hard links, get will create hard links for the patch files, which means that the additional storage space needed will be minimal. This is very good for your disk usage (and for the speed of running get), so if you want multiple copies of a repository, I strongly recommend first running darcs get to get yourself one copy, and then running darcs get on that copy to make any more you like. The only catch is that the first time you run darcs push or darcs pull from any of these second copies, by default they will access your first copy--which may not be what you want.

If you want to get a specific version of a repository, you have a few of options. You either use the --tag, --patch or --match options, which are also accepted by a number of other commands, or you can use the --context=FILENAME option, which specifies a file containing a context generated with darcs changes --context. This allows you (for example) to include in your compiled program an option to output the precise version of the repository from which it was generated, and then perhaps ask users to include this information in bug reports.

Note that when specifying --patch or --match, you may get a version of your code that has never before been seen, if the patches have gotten themselves reordered. If you ever want to be able to precisely reproduce a given version, you need either to tag it or create a context file.

5.9 darcs pull

Pull is used to bring changes made in another repo into the current repo (that is, the one that is the current directory). Pull allows you to bring over all or some of the patches that are in that repo but not in the current one. Pull accepts an argument, which is the URL from which to pull, and when called without an argument, pull will use the repository from which you have most recently either pushed or pulled.

You can use an external interactive merge tool to resolve conflicts via the flag --external-merge. For more details see section 5.1.9.

The --patch-name argument can be used to specify a regexp, which should be of the extended type used by egrep. If this option is used, only patches which match this regexp (along with their dependencies) are considered. Similarly, --tag-name can be used along with a regexp to pull all patches which are in versions with a matching tag.

If you give a --patch-name argument, darcs will silently pull along any other patches upon which the patches which match the patch-name depend. So --patch-name bugfix mean “pull all the patches with ‘bugfix’ in their name, along with any patches they require.” If you really only want the patches with ‘bugfix’ in their name, you should use the --no-deps option, which is only useful in combination with --patch-name, and makes darcs only pull in those matching patches which have no dependencies (apart from other matching patches).

If you specify the --test option, pull will run the test (if a test exists) on a scratch copy of the repo contents prior to actually performing the pull. If the test fails, the pull will be aborted.

5.10 darcs push

Push is the opposite of pull. Push allows you to move changes from the current repository into another repository.

For obvious reasons, you can only push to repositories to which you have write access. In addition, you can only push to repos that you access either via the local file system or via ssh. In order to apply via ssh, darcs must also be installed on the remote computer. The command invoked to run ssh may be configured via the DARCS_SSH environment variable (see section ??.

If you give the --apply-as flag, darcs will use sudo to apply the changes as a different user. This can be useful if you want to set up a system where several users can modify the same repository, but you don’t want to allow them full write access. This isn’t secure against skilled malicious attackers, but at least can protect your repository from clumsy, inept or lazy users.

The --patchname argument can be used to specify a regexp, which should be of the extended type used by egrep. If this option is used, only patches which match this regexp (along with their dependencies) are considered for pushing.

5.11 darcs send

Send is used to prepare a bundle of patches to send to another repository. Send accepts the URL of the repository as an argument. When called without an argument, send will use the most recent repository that was either pushed to, pulled from or sent to.

If you want to create a patches having context, you can use the --unified option, which create output vaguely reminiscent of “diff -u”.

The --output and --to flags determine what darcs does with the patch bundle after creating it. If you provide an --output argument, the patch bundle is saved to that file. If you give one or more --to arguments, the bundle of patches is emailed to those addresses.

If you don’t provide either a --output or a --to flag, darcs will look at the contents of the _darcs/prefs/email file in the target repository (if it exists), and send the patch by email to that address. In this case, you may use the --cc option to specify additional recipients without overriding the default repository email address.

If there is no email address associated with the repository, darcs will prompt you for an email address.

The --patchname argument can be used to specify a regexp, which should be of the extended type used by egrep. If this option is used, only patches which match this regexp (along with their dependencies) are considered for sending.

If you want to include a description or explanation along with the bundle of patches, you need to specify the --edit-description flag, which will cause darcs to open up an editor with which you can compose an email to go along with your patches.

5.12 darcs apply

Apply is used to apply a bundle of patches to this repository. Such a bundle may be created using send.

Darcs apply accepts a single argument, which is the name of the patch file to be applied. If you omit this argument, the patch is read from standard input.³ This allows you to use apply with a pipe from your email program, for example.

If you specify the --verify PUBRING option, darcs will check that the patch was gpg-signed by a key which is in PUBRING, and will refuse to apply the patch otherwise.

If you give the --reply FROM option to darcs apply, it will send the results of the application to the sender of the patch. This only works if the patch is in the form of an email with its headers intact, so that darcs can actually know the origin of the patch. The reply email will indicate whether or not the patch was successfully applied. The FROM flag is the email address that will be used as the “from” address when replying. If the darcs apply is being done automatically, it is important that this address not be the same as the address at which the patch was received in order to avoid automatic email loops.

If you want to also send the apply email to another address (for example, to create something like a “commits” mailing list), you can use the --cc option to specify additional recipients.

The --reply feature of apply is intended primarily for two uses. When used by itself, it is handy for when you want to apply patches sent to you by other developers so that they will know when their patch has been applied. For example, in my .muttrc (the config file for my mailer) I have:

You can use an external interactive merge tool to resolve conflicts via the flag --external-merge. For more details see section 5.1.9.

If you provide the --interactive or --gui flag, darcs will ask you for each change in the patch bundle whether or not you wish to apply that change. The opposite is the --all flag, which can be used to override an interactive or gui which might be set in your “defaults” file.

When used in combination with the --verify option, the --reply option allows for a nice pushable repository. When these two options are used together, any patches that don’t pass the verify will be forwarded to the FROM address of the --reply option. This allows you to set up a repository so that anyone who is authorized can push to it and have it automatically applied, but if a stranger pushes to it, the patch will be forwarded to you. Please (for your own sake!) be certain that the --reply FROM address is different from the one used to send patches to a pushable repository, since otherwise an unsigned patch will be forwarded to the repository in an infinite loop.

If you use ‘darcs apply --verify PUBRING --reply’ to create a pushable repo by applying patches automatically as they are recieved via email, you will also want to use the --no-resolve-conflicts option, which will keep the local and recorded versions in sync on the repo.

If you specify the --test option, apply will run the test (if a test exists) prior to applying the patch. If the test fails, the patch is not applied. In this case, if the --reply option was used, the results of the test are send in the reply email. You can also specify the --no-test option, which will override the --test option, and prevent the test from being run. This is helpful when setting up a pushable repository, to keep users from running code.

5.13 darcs rollback

Rollback is used to undo the effects of a single patch without actually deleting that patch. Instead, it applies the inverse patch as a new patch. Unlike unpull and unrecord (which accomplish a similar goal) rollback is perfectly safe, since it leaves in the repository a record of the patch it is removing. If you decide you didn’t want to roll back a patch after all, you probably should use unrecord to undo the rollback, since like rollback, unrecord doesn’t affect the working directory.

5.14 darcs unrecord

Unrecord is used to undo a single recorded patch. It will prompt you for which patch to unrecord, and then will undo that patch. Unrecord actually removes the unrecorded patch from your repository, so there is no way to “rerecord” an unrecorded patch⁴. Note that unrecord doesn’t affect your working copy of the tree at all, so if you want to completely undo the change, you’ll also need to darcs revert, or do an unpull.

If you don’t revert after unrecording, then the changes made by the unrecorded patches are left in your working tree. If these patches are actually from another repository, interaction (either pushes or pulls) with that repository may be massively slowed down, as darcs tries to cope with the fact that you appear to have made a large number of changes that conflict with those present on the other repository. So if you really want to undo the result of a pull operation, use unpull! Unrecord is primarily intended for when you record a patch, realize it needs just one more change, but would rather not have a separate patch for just that one change.

5.15 darcs unpull

Unpull is used to undo a single patch that has been pulled from another repository. It will prompt you for which patch to unpull, and then will undo that patch. Beware that unpull undoes the patch both from the repo records AND from the current working directory, and does NOT check that the patch originated with a pull. In otherwords, you could lose precious code by unpulling!

Like unrecord, unpull does not just apply an inverse patch to the repository, it actually deletes the patch from the repository. This makes unpull a particularly dangerous command, as it not only deletes the patch from the repo, but also removes the changes from the working directory. It is equivalent to an unrecord followed by a revert, except that revert can be unreverted.

Contrary to what its name suggests, there is nothing in unpull that requires that the “unpulled” patch originate from different repository. The name was chosen simply to suggest a situation in which it is “safe” to used unpull. If the patch was originally from another repo, then unpulling is safe, because you can always pull the patch again if you decide you want it after all. If you unpull a locally recorded patch, all record of that change is lost, which is what makes this a “dangerous” command, and thus deserving of a obscure name which is more suggestive of when it is safe to use than precisely what it does.

5.16 darcs revert

Revert is used to undo changes make to the local tree which have not yet been recorded. You will be prompted for which changes you wish to undo. The actions of a revert may be reversed using the unrevert command (see section 5.17). However, if you’ve made changes since the revert your mileage may vary, so please be careful.

You can give revert optional arguments indicating files or directories. If you do so it will only prompt you to revert changes in those files or in files in those directories.

5.17 darcs unrevert

Unrevert is used to undo the results of a revert command.

While this is only guaranteed to work properly if you haven’t made any changes since the revert was performed, it makes a best effort to merge the unreversion with any changes you have since made. In fact, unrevert should even work if you’ve recorded changes since reverting.

5.18 darcs dist

Create a distribution tarball.

Dist is a handy tool for implementing a “make dist” target in your makefile. It creates a tarball of the recorded edition of your tree. Basically, you will typically use it via a makefile rule such as

5.19 darcs mv

Darcs mv needs to be called whenever you want to rename or move a file or directory. Unlike remove, mv actually performs the move itself in your working directory. This is why “mv” isn’t called “move”, since it is really almost equivalent to the unix command “mv”. I could add an equivalent command named “move” for those who like vowels.

Darcs mv will by default refuse to rename a file if there already exists a file having the same name apart from case. This is because doing so could create a repository that could not be used on file systems that are case insensitive (such as Apples HFS+). You can override this by with the flag --case-ok.

5.20 darcs replace

Replace allows you to change a specified token whereever it occurs in the specified files. Tokens here are defined by a regexp specifying the characters which are allowed. By default a token corresponds to a C identifier.

The default regexp is [A-Za-z_0-9]), and if one of your tokens contains a ‘-’ or ‘.’, you will then (by default) get the “filename” regexp, which is [A-Za-z_0-9\-\.]. If you prefer to choose a different set of characters to define your token (perhaps because you are programming in some other language), you may do so with the --token-chars option. You may prefer to define tokens in terms of delimiting characters instead of allowed characters using a flag such as --token-chars '[^ \n\t]', which would define a token as being white-space delimited.

If you do choose a non-default token definition, I recommend using _darcs/prefs/defaults to always specify the same --token-chars, since your replace patches will be better behaved (in terms of commutation and merges) if they have tokens defined in the same way.

When using darcs replace, the “new” token may not already appear in the file--if that is the case, the replace change would not be invertible. This limitation holds both on the working copy of the file and on the already recorded version of the file.

There is a potentially confusing difference, however, when a replace is used to make another replace possible:

5.21 darcs tag

Tag is used to name a version of the tree. Tag differs from record in that it doesn’t record any new changes, and it always depends on all patches residing in the repository when it is tagged. This means that one can later reproduce this version of the repository by calling, for example:

Each tagged version has a version name. The version is also flagged with the person who tagged it (taken by default from the ‘DARCS_EMAIL’ or ‘EMAIL’ environment variable). The date is also included in the version information.

A tagged version automatically depends on all patches in the repo. This allows you to later reproduce precisely that version. The tag does this by depending on all patches in the repo, except for those which are depended upon by other tags already in the repo. In the common case of a sequential series of tags, this means that the tag depends on all patches since the last tag, plus that tag itself.

If you run tag with the --pipe option, you will be prompted for the tag name and date. This interface is intended for scripting darcs, in particular for writing repository conversion scripts. The prompts are intended mostly as useful guide (since scripts won’t need them), to help you understand the format in which to provide the input.

5.22 darcs changes

Changes gives a human-readable list of changes between versions suitable for use as a changelog.

When given one or more files or directories as an argument, changes lists only those patches which affect those files or the contents of those directories, or of course the directories themselves.

If changes is given one --tag-name or --patch-name flag, it outputs only those changes since that tag or patch. If two such flags are given, the changes between the two versions are output.

When given the --context flag, darcs changes outputs sufficient information which will allow the current state of the repository to be recreated at a later date. This information should generally be piped to a file, and then can be used later in conjunction with darcs get --context to recreate the current version. Note that while the --context flag may be used in conjunction with --xml-output or --human-readable, in neither case will darcs get be able to read the output. On the other hand, sufficient information will be output for a knowledgeable human to recreate the current state of the repository.

5.23 darcs diff

Diff can be used to create a diff between two versions which are in your repository. Specifying just one version will get you a diff against the version in your working directory. If you give diff no version arguments, it gives you the same information as whatsnew except that the patch is formatted as the output of a diff command.

Diff calls an external “diff” command to do the actual work, and passes any unrecognized flags to this diff command. Thus you can call

If you want to view only the differences to one or more files, you can do so with a command such as

FIXME: I should allow the user to specify the external diff command. Currently it is hardwired to ”diff”.

5.24 darcs setpref

Usage example:

Setpref allows you to set a preferences value in a way that will propagate to other repositories. Valid preferences are: test predist boringfile binariesfile. If you just want to set the pref value in your repository only, you can just edit “_darcs/prefs/prefs”. Changes you make in that file will be preserved.

The “_darcs/prefs/prefs” holds the only preferences information that can propagate between repositories via pushes and pulls, and the only way this happens is when the setprefs command is used. Note that although prefs settings are included in patches, they are not fully version controlled. In particular, depending on the order in which a series of merges is perform, you may end up with a different final prefs configuration. In practice I don’t expect this to be a problem, as the prefs usually won’t be changed very often.

The following values are valid preferences options which can be configured using setpref:

• “test” -- the command to run as a test script.
• “predist” -- a command to run prior to tarring up a distribution tarball. Typically this would consist of autoconf and/or automake.
• “boringfile” -- the name of a file to read instead of the “boring” prefs file.
• “binariesfile” -- the name of a file to read instead of the “binaries” prefs file.

5.25 darcs trackdown

Trackdown tries to find the most recent version in the repository which passes a test. Given no arguments, it uses the default repository test. Given one argument, it treats it as a test command. Given two arguments, the first is an initialization command with is run only once, and the second is the test command. When given no options, trackdown tries to find the latest version that passes the test (i.e. the test that is run under ’darcs record’). If you give it a single argument, it is interpereted as a shell command to be run as a test. If you give it two arguments, the first is a shell command that is run only once (e.g. autoconf, perhaps) and the second is the “test command”.

Trackdown is helpful for locating when something was broken. FIXME: It is still rather primitive. Currently it just goes back over the history in reverse order trying each version. I’d like for it to explore different patch combinations, to try to find the minimum number of patches that you would need to unpull in order to make the test succeed.

FIXME: I also would like to add an interface by which you can tell it which patches it should consider not including. Without such a feature, the following command:

Example usage If you want to find the last version of darcs that had a FIXME note in the file Record.lhs, you could run

To find the latest version that compiles, you can run

5.26 darcs optimize

Optimize reorganizes your repository data to make it more efficent to access

Optimize will always write out a fresh copy of the inventory that minimizes the amount of inventory that need be downloaded when people pull from the repo.

If you use the --checkpoint flag optimize creates a checkpoint patch for a tag. You can specify the tag with the --tag-name option, or just let darcs choose the most recent tag.

If you give the compress flag, optimize will compress all the patches in the repository. Similarly, if you give the uncompress or dont-compress flag, optimize will decompress all the patches in the repository. You might want to do this on large repositories if you have plenty of disk space but run into memory pressure, since it will allow darcs to open the patch files via mmap.

5.27 darcs annotate

Display useful information about the repository history.

Annotate allows you to extract all sorts of interesting information from your repository.

When called with just a patch name, annotate outputs the patch in darcs format. The --summary, --human-readable or --unified options may be used to modify the format with which it is displayed.

If a directory name is given, annotate will output the contents of that directory. If a patch name is given, the contents of that directory after that patch was applied will be output. If a tag name is given, the contents of that directory in the specified tagged version will be output.

If a file name is given, the contents of that file will be output, along with markup indicating when each line was last (and perhaps next) modified.

5.28 darcs repair

Repair attempts to fix corruption that may have entered your repository.

Repair currently will only repair damage to the _darcs/current directory. Fortunately this is just the sort of corruption that is most likely to happen.

Chapter 6
Web interface

The darcs web interface allows you to conveniently browse the information stored in a repository. You can view the repo by file and see the history of that file as various patches were applied, or you can browse in patch view mode, seeing which files were modified in each patch.

The darcs_cgi cgi script allows you to browse changes made in your darcs repository via the web. To use it with apache, you can install it using make installserver, and create a cache directory at /var/cache/darcs_cgi. This cache directory must be writeable by the cgi script, which for me means chowning it to the user and group www-data. Finally, you should create a directory named repos in /var/www, in which you will place symlinks to the repos themselves. Once all this is done, the user can then browse the repos at http://your.site/cgi-bin/darcs.

The repos directory is configurable via the configuration file /etc/darcs/cgi.conf. This file can contain comments (any line starting with a ‘#’ char) and key value pairs with an equal sign in between. For example:

The page, http://your.site/cgi-bin/darcs, displays a listing of all repos available on the server. From this page, the user can get to any available repository.

Clicking on a given repository will take you to the file view page of that repository, which shows a listing of all the files in the repo. This page also has a link to the patch view page, which is a bit more interesting.

In patch view mode, the web interface displays a listing of all the patches in the repo. Clicking on a patch gives a listing of all files that were in the repo at the time that patch was applied.

Clicking on one of the files shows the file contents, with added lines shown in green, and removed ones in red. To the left of each line is a small ‘+’ and ‘-’. These are links to the patch which added or removed that line.

Appendix A
Theory of patches

A.1 Background

I think a little background on the author is in order. I am a physicist, and think like a physicist. The proofs and theorems given here are what I would call “physicist” proofs and theorems, which is to say that while the proofs may not be rigorous, they are practical, and the theorems are intended to give physical insight. It would be great to have a mathematician work on this, but I am not a mathematician, and don’t care for math.

From the beginning of this theory, which originated as the result of a series of email discussions with Tom Lord, I have looked at patches as being analagous to the operators of quantum mechanics. I include in this appendix footnotes explaining the theory of patches in terms of the theory of quantum mechanics. I know that for most people this won’t help at all, but many of my friends (and as I write this all three of darcs’ users) are physicists, and this will be helpful to them. To nonphysicists, perhaps it will provide some insight into how at least this physicist thinks.

A.2 Introduction

A patch describes a change to the tree. It could be either a primitive patch (such as a file add/remove, a directory rename, or a hunk replacement within a file), or a compostive patch describing many such changes. Every patch type must satisfy the conditions described in this appendix. The theory of patches is independent of the data which the patches manipulate, which is what makes it both powerful and useful, as it provides a framework upon which one can build a revision control system in a sane manner.

Although in a sense, the defining property of any patch is that it can be applied to a certain tree, and thus make a certain change, this change does not wholly define the patch. A patch is defined by a representation, together with a set of rules for how it behaves (which it has in common with its patch type). The representation of a patch defines what change that particular patch makes, and must be defined in the context of a specific tree. The theory of patches is a theory of the many ways one can change the representation of a patch to place it in the context of a different tree. The patch itself is not changed, since it describes a single change, which must be the same regardless of its representation¹.

So how does one define a tree, or the context of a patch? The simplest way to define a tree is as the result of a series of patches applied to the empty tree². Thus, the context of a patch consists of the set of patches that precede it.

The simplest relationship between two patches is that of “sequential” patches, which means that the context of the second patch (the one on the left) consists of the first patch (on the right) plus the context of the first patch. The composition of two patches (which is also a patch) refers to the patch which is formed by first applying one and then the other. The composition of two patches, P₁ and P₂ is represented as P₂P₁, where P₁ is to be applied first, then P₂³

There is one other very useful relationship that two patches can have, which is to be parallel patches, which means that the two patches have an identical context (i.e. their representation applies to identical trees). This is represented by P₁ || P₂. Of course, two patches may also have no simple relationship to one another. In that case, if you want to do something with them, you’ll have to manipulate them with respect to other patches until they are either in sequence or in parallel.

The most fundamental and simple property of patches is that they must be invertible. The inverse of a patch is decribed by: P^-1. In the darcs implementation, the inverse is required to be computable from knowledge of the patch only, without knowledge of its context, but that (although convenient) is not required by the theory of patches.

Using this definition, it is trivial to prove the following theorem relating to the inverse of a composition of two patches.

Theorem 1 The inverse of the composition of two patches is

Moreover, it is possible to show that the right inverse of a patch is equal to its left inverse. In this respect, patches continue to be analagous to square matrices, and indeed the proofs relating to these properties of the inverse are entirely analagous to the proofs in the case of matrix multiplication. The compositions proofs can also readily be extended to the composition of more than two patches.

The first way (of only two) to change the context of a patch is by commutation, which is the process of changing the order of two sequential patches.

Definition 2 The commutation of patches P₁ and P₂ is represented by

Here P₁' is intended to describe the same change as P₁, with the only difference being that P₁' is applied after P₂' rather than before P₂.

The above definition is obviously rather vague, the reason being that what is the “same change” has not been defined, and we simply assume (and hope) that the code’s view of what is the “same change” will match those of its human users. The ‘’ operator should be read as something like the == operator in C, indicating that the right hand side performs identical changes to the left hand side, but the two patches are in reversed order. When read in this manner, it is clear that commutation must be a reversible process, and indeed this means that commutation can fail, and must fail in certain cases. For example, the creation and deletion of the same file cannot be commuted. When two patches fail to commute, it is said that the second patch depends on the first, meaning that it must have the first patch in its context (remembering that the context of a patch is a set of patches, which is how we represent a tree). ⁴

Merge The second way one can change the context of a patch is by a merge operation. A merge is an operation that takes two parallel patches and gives a pair of sequenctial patches. The merge operation is represented by the arrow “”.

Definition 3 The result of a merge of two patches, P₁ and P₂ is one of two patches, P₁' and P₂', which satisfy the relationship:

Note that the sequential patches resulting from a merge are required to commute. This is an important consideration, as without it most of the manipulations we would like to perform would not be possible. The other important fact is that a merge cannot fail. Naively, those two requirements seem contradictory. In reality, what it means is that the result of a merge may be a patch which is much more complex than any we have yet considered⁵.

A.3 How merges are actually performed

The constraint that any two compatible patches (patches which can successfully be applied to the same tree) can be merged is actually quite difficult to apply. The above merge constraints also imply that the result of a series of merges must be independent of the order of the merges. So I’m putting a whole section here for the interested to see what algorithms I use to actually perform the merges (as this is pretty close to being the most difficult part of the code).

The first case is that in which the two merges don’t actually conflict, but don’t trivially merge either (e.g. hunk patches on the same file, where the line number has to be shifted as they are merged). This kind of merge can actually be very elegantly dealt with using only commutation and inversion.

There is a handy little theorem which is immensely useful when trying to merge two patches.

This can easily be proven by multiplying both sides of the first commutation by P₁'^-1 on the left, and by P₁^-1 on the right. Besides being used in merging, this theorem is also useful in the recursive commutations of mergers. From Theorem 2, we see that the merge of P₁ and P₂' is simply the commutation of P₂ with P₁^-1 (making sure to do the commutation the right way). Of course, if this commutation fails, the patches conflict. Moreover, one must check that the merged result actually commutes with P₁, as the theorem applies only when both commutations are successful.

Of couse, there are patches that actually conflict, meaning a merge where the two patches truly cannot both be applied (e.g. trying to create a file and a directory with the same name). We deal with this case by creating a special kind of patch to support the merge, which we will call a “merger”. Basically, a merger is a patch that contains the two patches that conflicted, and instructs darcs basically to resolve the conflict. By construction a merger will satisfy the commutation property (see Definition 3) that characterizes all merges. Moreover the merger’s properties are what makes the order of merges unimportant (which is a rather critical property for darcs as a whole).

The job of a merger is basically to undo the two conflicting patches, and then apply some sort of a “resolution” of the two instead. In the case of two conflicting hunks, this will look much like what CVS does, where it inserts both versions into the file. In general, of course, the two conflicting patches may both be mergers themselves, in which case the situation is considerably more complicated.

Much of the merger code depends on a routine which recreates from a single merger the entire sequence of patches which led up to that merger (this is, of course, assuming that this is the complicated general case of a merger of mergers of mergers). This “unwind” procedure is rather complicated, but absolutely critical to the merger code, as without it we wouldn’t even be able to undo the effects of the patches involved in the merger, since we wouldn’t know what patches were all involved in it.

Basically, unwind takes a merger such as

The catch is that things don’t always turn out this easily. There is no guarantee that the two A’s would come out at the same time, and if they didn’t, we’d have to rearrange things until they did. Or if there was no way to rearrange things so that they would agree, we have to go on to plan B, which I will explain now.

Consider the case of M( M(A,B), M(C,D)). We can easily unwind the two subpatches

Giving an unwinding of

There are a couple of simple constraints on the routine which determines how to resolve two conflicting patches (which is called ‘glump’). These must be satisfied in order that the result of a series of merges is always independent of their order. Firstly, the output of glump cannot change when the order of the two conflicting patches is switched. If it did, then commuting the merger could change the resulting patch, which would be bad. Secondly, the result of the merge of three (or more) conflicting patches cannot depend on the order in which the merges are performed.

The conflict resolution code (glump) begins by “unravelling” the merger into a set of sequences of patches. Each sequence of patches corresponds to one non-conflicted patch that got merged together with the others. The result of the unravelling of a series of merges must obviously be independent of the order in which those merges are performed. This unravelling code (which uses the unwind code mentioned above) uses probably the second most complicated algorithm. Fortunately, if we can successfully unravel the merger, almost any function of the unravelled merger satisfies the two constraints mentioned above that the conflict resolution code must satisfy.

It can sometimes be handy to have a canonical representation of a given patch. We achieve this by defining a canonical form for each patch type, and a function “canonize” which takes a patch and puts it into canonical form. This routine is used by the diff function to create an optimal patch (based on an LCS algorithm) from a simple hunk describing the old and new version of a file. Note that canonization may fail, if the patch is internally inconsistent.

A simpler, faster (and more generally useful) cousin of canonize is the coalescing function. This takes two sequential patches, and tries to turn them into one patch. This function is used to deal with “split” patches, which are created when the commutation of a primitive patch can only be represented by a composite patch. In this case the resulting composite patch must return to the original primitive patch when the commutation is reversed, which a split patch accomplishes by trying to coalesce its contents each time it is commuted.

A.4 File patches

A file patch is a patch which only modifies a single file. There are some rules which can be made about file patches in general, which makes them a handy class. For example, commutation of two filepatches is trivial if they modify different files. There is an exception when one of the files has a name ending with “-conflict”, in which case it may not commute with a file having the same name, but without the “-conflict.” If they happen to modify the same file, we’ll have to check whether or not they commute.

There is another handy function, which primarily affects file patches (although it can also affect other patches, such as rename patches or dir add/remove patches), which is the submerge-in-directory function. This function changes the patch to act on a patch within a subdirectory rather than in the current directory, and is useful when performing the recursive diff.

Hunks are an example of a complex filepatch. A hunk is a set of lines of a text file to be replaced by a different set of lines. Either of these sets may be empty, which would mean a deletion or insertion of lines. The hunk is the simplest patch that has a commuting pattern in which the commuted patches differ from the originals (rather than simple success or failure). This makes commuting or merging two hunks a tad tedious. Hunks, of course, can be coalesced if they have any overlap. Note that coalesce code doesn’t check if the two patches are conflicting. If you are coalescing two conflicting hunks, you’ve already got a bug somewhere.

One of the most important pieces of code is the canonization of a hunk, which is where the “diff” algorithm is performed. This algorithm begins with chopping off the identical beginnings and endings of the old and new hunks. This isn’t strictly necesary, but is a good idea, since this process is O(n), while the primary diff algorithm is something considerably more painful than that... actually the head would be dealt with all right, but with more space complexity. I think it’s more efficient to just chop the head and tail off first.

A.5 Token replace patches

Although most filepatches will be hunks, darcs is clever enough to support other types of changes as well. A “token replace” patch replaces all instances of a given token with some other version. A token, here, is defined by a regular expression, which must be of the simple [a-z...] type, indicating which characters are allowed in a token, with all other characters acting as delimiters. For example, a C identifier would be a token with the flag [A-Za-z_0-9].

What makes the token replace patch special is the fact that a token replace can be merged with almost any ordinary hunk, giving exactly what you would want. For example, you might want to change the patch type TokReplace to TokenReplace (if you decided that saving two characters of space was stupid). If you did this using hunks, it would modify every line where TokReplace occurred, and quite likely provoke a conflict with another patch modifying those lines. On the other hand, if you did is using a token replace patch, the only change that it could conflict with would be if someone else had used the token “TokenReplace” in their patch rather than TokReplace--and that actually would be a real conflict!

A.6 Composite patches

Composite patches are made up of a series of patches intended to be applied sequentially. They are represented by a list of patches, with the first patch in the list being applied first.

A.7 Patch string formatting

Of course, in order to store our patches in a file, we’ll have to save them as some sort of strings. The convention is that each patch string will end with a newline, but on parsing we skip any amount of whitespace between patches.

Composite patch A patch made up of a few other patches.

Split patch A split patch is similar to a composite patch (identical in how it’s stored), but rather than being composed of several patches grouped together, it is created from one patch that has been split apart, typically through a merge or commutation.

Hunk Replace a hunk (set of contiguous lines) of text with a new hunk.

Token replace Replace a token with a new token. Note that this format means that the white space must not be allowed within a token. If you know of a practical application of whitespace within a token, let me know and I may change this.

Binary file modification Modify a binary file

Add file Add an empty file to the tree. addfile filename

Remove file Delete a file from the tree. rmfile filename

Move Rename a file or directory. move oldname newname

Change Pref Change one of the preference settings. Darcs stores a number of simple string settings. Among these are the name of the test script and the name of the script that must be called prior to packing in a make dist.

Add dir Add an empty directory to the tree. adddir filename

Remove dir Delete a directory from the tree. rmdir filename

Merger patches Merge two patches. The MERGERVERSION is included to allow some degree of backwards compatibility if the merger algorithm needs to be changed.

Named patches Named patches are diplayed as a ‘patch id’ which is in square brackets, followed by a patch. Optionally, after the patch id (but before the patch itself) can come a list of dependencies surrounded by angle brackets. Each dependency consists of a patch id.

Appendix B
Repository format

A repository consists of a working directory, which has within it a directory called _darcs. There must also be subdirectories within _darcs named current and patches. The current directory contains the version of the tree which has been recorded, while patches contains the actual patches which are in the repository.

WARNING! Viewing files in current is perfectly acceptable, but if you view them with an editor (e.g. vi or emacs), that editor may create temporary files in the _darcs/current dirctory, which will temporarily cause your repository to be inconsistent. So don’t record any patches while viewing files in _darcs/current with an editor! A better plan would be to restrict yourself to viewing these files with a pager such as more or less.

Also within _darcs is the inventory file, which lists all the patches that are in the repo. Moreover, it also gives the order of the representation of the patches as they are stored. Given a source of patches, i.e. any other set of repositories which have between them all the patches contained in a given repo, that repo can be reproduced based on only the information in the inventory file. Under those circumstances, the order of the patches specified in the inventory file would be unimportant, as this order is only needed to provide context for the interperetation of the stored patches in this repository.

There is a very special patch which may be stored in patches which is called ‘pending’. This patch describes any changes which have not yet been recorded, and cannot be determined by a simple diff. For example file additions or renames are placed in pending until they are recorded. Similarly, token replaces are stored in pending until they are recorded. The _darcs directory also contains a directory called “prefs”, which is described in Chapter 4.

Appendix C
GNU General Public License

Version 2, June 1991

Preamble

The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation’s software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Library General Public License instead.) You can apply it to your programs, too.

When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things.

To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it.

For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights.

We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software.

Also, for each author’s protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors’ reputations.

Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone’s free use or not licensed at all.

The precise terms and conditions for copying, distribution and modification follow.

0. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The “Program”, below, refers to any such program or work, and a “work based on the Program” means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term “modification”.) Each licensee is addressed as “you”.

   Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does.

1. You may copy and distribute verbatim copies of the Program’s source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program.

   You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee.

2. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions:
   1. You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change.
   2. You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.
   3. If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.)

   These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.

   Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program.

   In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License.

3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:
   1. Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
   2. Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
   3. Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.)

   The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable.

   If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code.

4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.
5. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it.
6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients’ exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License.
7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.

   If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances.

   It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice.

   This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License.

8. If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License.
9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns.

   Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and “any later version”, you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.

10. If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally.
    NO WARRANTY

11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.