VCSes – Elijah’s Blog

Git User’s Survey 2009

Git is running their annual user survey again.

We would like to ask you a few questions about your use of the Git version control system. This survey is mainly to understand who is using Git, how and why.

The results will be published to the Git wiki and discussed on the git mailing list.

The survey would be open from 15 July till 15 September 2009.

Please devote a few minutes of your time to fill this simple questionnaire, it will help a lot the git community to understand your needs, what you like of Git, and of course what you don’t like of it.

The survey can be found here:
http://tinyurl.com/GitSurvey2009
http://www.survs.com/survey?id=2PIMZGU0&channel=Q0EKJ3NF54

Limbo: Why users are more error-prone with git than other VCSes

Limbo is a term I use but VCS authors don’t. However, that’s because they tend to ignore a certain state that exists in all major VCSes (and give it no name because they tend to ignore it) despite the fact that this state seems to be the largest source of errors. I call this state limbo.

How to make git behave like other VCSes

Most potential git users probably don’t want to read this whole page, and would like their knowledge from usage of other VCSes to apply without learning how the index and limbo are different in git than their previous VCS (despite the really cool extra functionality it brings). This can be done by

Always using git diff HEAD instead of git diff

and

Always using git commit -a instead of git commit

Either make sure you always remember those extra arguments, or come back and read this page when you get a nasty surprise.

The concept of Limbo

VCS users are accustomed to thinking of using their VCS in terms of two states — a working copy where local changes are made, and the repository where the changes are saved. However, the working copy is split into three sets (see also VCS concepts):

(explicitly) ignored — files inside your working copy that you explicitly told the VCS system to not track
index — the content in your working copy that you asked the VCS to track; this is the portion of your working copy that will be saved when you commit (in CVS, this is done using the CVS/Entries files)
limbo — not explicitly ignored, and not explicitly added. This is stuff in your working copy that won’t be checked in when you commit but you haven’t told the VCS to ignore, which typically includes newly created files.

The first state is identical across all major VCSes. The second two states are identical across cvs, svn, bzr, hg, and likely others. But git splits the index and limbo differently.

One could imagine a VCS which just automatically saves all changes that aren’t in an explicitly ignored file (including newly created files) whenever a developer commits, i.e. a VCS where there is no limbo state. None of the major VCSes do this, however. There are various rationales for the existence of limbo: maybe developers are too lazy to add new files to the ignored list, perhaps they are unaware of some autogenerated files, or perhaps the VCS only has one ignore list and developers want to share it but not include their own temporary files in such a shared list. Whatever the reason, limbo is there in all major VCSes.

Changes in limbo are a large source of user error

The problem with limbo is that changes in this state are, in my experience, the cause of the most errors with users. If you create a new file and forget to explicitly add it, then it won’t be included in your commit (happens with all the major VCSes). Naturally, even those familiar with their VCS forget to do that from time to time. This always seems to happen when other changes were committed that depend on the new files, and it always happens just before the relevant developers go on vacation…leaving things in a broken state for me to deal with. (And sure, I return the favor on occasion when I simply forget to add new files.)

A powerful feature of git

Unlike other VCSes, git only commits what you explicitly tell it to. This means that without taking additional steps, the command “git commit” will commit nothing (in this particular case it typically complains that there’s nothing to commit and aborts). git also gives you a lot of fine-grained control over what to commit, more than most other VCSes. In particular, you can mark all the changes of a given file for subsequent committing, but unlike other VCSes this only means that you are marking the current contents of that file for commit; any further changes to the same file will not be included in subsequent commits unless separately added. Additionally, recent versions of git allow the developer to mark subsets of changes in an existing file for commit (pulling a handy feature from darcs). The power of this fine-grained choose-what-to-commit functionality is made possible due to the fact that git enables you to generate three different kinds of diffs: (1) just the changes marked for commit (git diff –cached), (2) just the changes you’ve made to files beyond what has been marked for commit (git diff), or (3) all the changes since the last commit (git diff HEAD).

This fine-grained control can come in handy in a variety of special cases:

When doing conflict resolution from large merges (or even just reviewing a largish patch from a new contributor), hunks of changes can be categorized into known-to-be-good and still-needs-review subsets.
It makes it easier to keep “dirty” changes in your working copy for a long time without committing them.
When adding a feature or refactoring (or otherwise making changes to several different sections of the code), you can mark some changes as known-to-be-good and then continue making further changes or even adding temporary debugging snippets.

These are features that would have helped me considerably in some GNOME development tasks I’ve done in the past.

How git is more problematic

This decision to only commit changes that are explicitly added, and doing so at content boundaries rather than file boundaries, amounts to a shift in the boundary between the index and limbo. With limbo being much larger in git, there is also more room for user error. In particular, while this allows for a powerful feature in git noted above, it also comes with some nasty gotchas in common use cases as can be seen in the following scenarios:

Only new files included in the commit
1. Edit bar
2. Create foo
3. Run git add foo
4. Run git commit
In this set of steps, users of other VCSes will be surprised that after step 4 the changes to bar were not included in the commit. git only commits changes when explicitly asked. (This can be avoided by either running git add bar before committing, or running git commit -a. The -a flag to commit means “Act like other VCSes — commit all changes in any files included in the previous commit”.)
Missing changes in the commit
1. Create/edit the file foo
2. Run git add foo
3. Edit foo some more
4. Run git commit
In this set of steps, users of other VCSes will be surprised that after step 4 the version of foo that was commited was the version that existed at the time step 2 was run; not the version that existed when step 4 was run. That’s because step 2 is translated to mean mark the changes currently in the file foo for commit. (This can be avoided by running git add foo again before committing, or running git commit -a for step 4.)
Missing edits in the generated patch
1. Edit the tracked file foo
2. Run git add foo
3. Edit foo some more
4. Run git diff
In this set of steps, users of other VCSes will be surprised that at step 4 they only get a list of changes to foo made in step 3. To get a list of changes to foo made since the last commit, run git diff HEAD instead.
Missing file in the generated patch
1. Create a new file called foo
2. Run git add foo
3. Run git diff
In this set of steps, users of other VCSes will be surprised that at step 3 the file foo is not included in the diff (unless changes have been made to foo since step 2, but then only those additional changes will be shown). To get foo included in the diff, run git diff HEAD instead.

These gotchas are there in addition to the standard gotcha exhibited in all the major VCSes:

How all the major VCSes are problematic

Missing file in the commit
1. Edit bar
2. Create a new file called foo
3. Run vcs commit (where vcs is cvs, svn, hg, bzr…see below about git)
In this set of steps, the edits in step 1 will be included in the commit, but the file foo will not be. The user must first run vcs add foo (again, replacing vcs with the relevant VCS being used) before committing in order to get foo included in the commit.

It turns out that git actually can help the user in this case due to its default to only commit what it is explicitly told to commit; meaning that in this case it won’t commit anything and tell the user that it wasn’t told to commit anything. However, since nearly every tutorial on git[*] says to use git commit -a, users include that flag most the time (60% of the time? 98%?). Due to that training, they’ll still get this nasty bug. However, they’re going to forget or neglect this flag sometimes, so they also get the new gotchas above.

[*] Recent versions of the official git tutorial being the only exception I’ve run across. It’s fairly thorough (make sure to also read part two), though it isn’t quite as explicit about the potential gotchas in certain situations.

How bzr, hg, and git mitigate these gotchas (and cvs and svn don’t)

These gotchas can be avoided by always running vcs status (again, replace vcs with the relevant VCS being used) and looking closely at the states the VCS lists files in. It turns out bzr, hg, and git are smart here and try to help the user avoid problems by showing the output of the status command when running a plain vcs commit (at the end of the commit message they are given to edit). This helps, but isn’t foolproof; I’ve somehow glossed over this extra bit of info in the past and still been bit. Also, I’ll often either use the -m flag to specify the commit message on the command line (for tiny personal projects) or a flag to specify taking the commit message from a file (i.e. using -F in most vcses, -l in hg).

The concepts a user must learn to understand existing VCSes

Note: Most will not find this post as interesting as my previous posts or my next one. It was intended to help explain questions like “How much knowledge transfers to a new VCS if you’ve learned another?” and “Why do some claim that certain *types* of VCSes are easier to learn than others, while others claim that they are all pretty much equal?”, questions I mentioned in my first post. Most probably aren’t interested in those questions and thus not this entry. I’m including it anyway.

Editors as an analogy

I sometimes see people arguing about whether text editors and word processors ought to automatically save with every change. While almost every existing editor has two states (the version being edited, plus the version on disk when last saved), some argue that it would simplify things to save on every change. Most editors stick with the two state model, which from a darwinian point of view would suggest it is the more superior model overall. However, it is interesting to note that the multi-state model does come with its complications even for simple cases like this. The multi-state model for editors has stung just about everyone at least once in the past before they learned the appropriate habits. For example, many have lost data in the past due to exiting the app before saving, due to power outages, due to application crashes, or even due to OS and hardware failures. (These days, most editors have workarounds which mitigate these problems.) Also, users can’t use separate programs to copy or print or import the file on disk and use it unless they rememebered to first save their latest changes. And users may be confused at first by extra files (foo.autosave, foo.bak, foo~, .#foo, etc.) that show up on their hard disk.

Virtually all editors use this two-state model (current edits, plus last version saved on disk), and nearly all computer users seem to have mastered it. At a basic level, VCSes use a similar model.

The multiple states of all major VCSes

All the major VCSes provide developers with their own little sandbox, or working copy, as well as a place for changes that are ready to be saved, called a repository. This maps almost directly to the concepts of standard editors — changes you make locally, and what version you last saved. Most any VCS guide will say that these are the two states you need to learn (I particularly remember reading several about CVS which said this.) It’s a convenient lie though. There are more than two states.

Compiling source code can create files that don’t need to be saved in the repository (others can regenerate them with the source). So, all the major VCSes have the concept of an ignore list; any files in the ignore list will not be saved in the repository. So we have three states so far: ignored files, local changes, and the repository.

Sadly, there’s another state that the working copy is split into. The major VCSes seem to have decided that developers may be too lazy to add files that shouldn’t be saved to the ignore list…or that they may be unaware of such files (editor autosave files, for example), or that developers want to have shared ignore lists but don’t want to add some personal files to such shared lists. Whatever the reason, the major VCSes have another state which I call “limbo”, whose existence everyone seems to forget about. This state is changes which aren’t explicitly added to the index (think CVS/Entries files in CVS) and thus will not be saved, but are not explicitly ignored either. This state causes the most bugs in my experience, even with advanced users, because people simply forget to explicitly add new files to the index and thus they don’t get saved with the rest of the changes. So we have four states so far (three being subsets of the working copy): explicitly ignored files, limbo, local changes that will be saved in the next commit, and the repository.

It turns out that the repository side also is split into multiple states. Developers want to be able to track what changes they themselves have made to their working copy, regardless of commits that have since been recorded in the (remote) repository. So, if you want to get a list of changes you’ve made, or the history that led up to your current working copy, it needs to be relative to the version of the repository that existed when you got your copy. That may not be the current version, because other developers could have recorded their changes in the (remote) repository. This also affects your ability to push your changes to the (remote) repository (by a “commit” in cvs or svn terminology), potentially requiring you to merge the various changes together. So, we have five states:

Substates of the working copy:
1. (explicitly) ignored — files inside your working copy that you explicitly told the VCS system to not track
2. index — the content in your working copy that you asked the VCS to track; this is the portion of your working copy that will be saved when you commit (in CVS, this is done using the CVS/Entries files)
3. limbo — not explicitly ignored, and not explicitly added. This is stuff in your working copy that won’t be checked in when you commit but you haven’t told the VCS to ignore, which typically includes newly created files.
Substates of the repository:
1. “checkout” version — the version of the code in your working copy before you started modifying it
2. remote version — the version of the code currently saved in the remote repository

Not understanding these multiple states and the differences between them for the VCS you are using has varying consquences: not being able to take full advantage of your system, being unable to do some basic operations, or (worst case) introducing erroneous or incomplete changes.

Similarities and differences between the major VCSes with these states

Most of these five states are similar between the major VCSes. State 1 (ignored files) is essentially identical between the systems. (The only difference is in the details of setting it up; for cvs it means editing .cvsignore, for svn it means modifying the svn:ignore property, for mercurial it means editing .hgingore, etc.) State 5 is also essentially identical. States 2 and 3 always sum up to everything in the working copy other than explicitly ignored files, so extending state 3 means shrinking state 2. Thus, we can get a feel for the differences between VCSes by looking at their differences in states 3 and 4.

The major distinction between inherently centralized VCSes (e.g. cvs and svn), and so called distributed (I prefer the term “multi-centered”) VCSes comes with state 4. The differences in this state can be thought of as different choices along a continuum rather than a binary difference, however. The difference here is in how much information gets cached when one gets a copy. With CVS, you only get a working copy plus info about what version you checked out and where the repository is located. With SVN, you get the same as with CVS, but also an extra copy of all the files. Most distributed systems go a few steps farther than svn and by default cache a copy of all the versions on the given branch. git, by default, caches a copy of all versions on all branches.

Caching extra information as part of state 4 can allow additional work to be done offline. cvs and svn are very limited in this respect, but the additional offline capabilities of the other systems come with the understanding that the local cache itself is a repository and thus users need to understand both how to sync changes with the local repository as well as between the local repository and the remote one(s). In cvs and svn, it’s not useful to “sync with the local cache”; instead those systems just automatically synchronize the local cache and the remote repository to the indexed local changes all at
once. Thus, cvs and svn users only need to learn a smaller set of “synchronization” commands (limited to “commit” and “update”.)

There is also a potential difference between VCSes in state 3. Having changes in state 3 is the place that in my experience causes the most errors. Users simply forget that their changes are in this state and forget to add them. Now, it turns out that all VCSes I’ve looked at close enough are identical here, except for git. (So if you know one of them you already also understand this aspect in all the other VCS systems other than git.) git extends the concept of limbo, turning the index into a high-level (and in your face) concept with some really cool features, but unfortunately it has the side-effect of making git even more error-prone for users. I’ll discuss this in more detail in my next post.

Local caching: A major distinguishing difference between VCSes

An interesting difference between the major VCSes concerns how much information is cached locally when one obtains a copy of the source code using the VCS. The amount of information obtained by default when one performs a checkout or clone with each of the five major VCSes is:

cvs – a working copy of the specified version of the source code files, plus information about which revision was checked out and where the repository is located.
svn – same as cvs, plus an extra copy of the specified version of the source code files
bzr, hg – same as svn, plus the remainder of the history of the current branch (i.e. cvs, plus a copy of the complete history of the current branch)
git – same as bzr & hg, plus the full history of all other branches in the repository as well.

Note that some systems have options to cache less than the default.

Benefits of local caching

The additional cached information can serve multiple purposes; for example, making operations faster (by using the local disk instead of the network), or allowing offline use. For example, nearly all operations in cvs other than edits of the working copy require network connectivity to the repository. In subversion, diffs between the version you checked out and your current working copy is fast due to the extra copy that was checked out, but other operations still require network connectivity. In bzr & hg, diffs against versions older than the checkout version, reverting to an older version, and getting a log of the changes on the branch can all be fast operations and can be done offline. In git, even comparing to a different branch, switching to a different branch, or getting a log of changes in any branch can be done quickly and offline.

This local caching also pushes another point: cvs and svn have limited utility when working offline. bzr, hg, and git allow quite a bit of offline use…in fact, it even makes sense to commit while offline (and then merge the local commit(s) and remote repositories later). Thus, one thinks of the local cache in such cases as being a repository itself. This has ramifications as well. Since the local cache is a repository, it means that it makes sense to think of updating from a different remote repository than you got your checkout/clone from, and of merging/pushing your changes to yet another location. This is the essence of being a VCS with distributed capabilities. This can be taken to the pathological extreme (resulting in the kernel development model), or one can use a more standard centralized model that simply has impressive offline capabilities (which is how Xorg runs), or one can pick something inbetween that suits them. One common case where someone might want to pick something in the middle is when an organization has multiple development sites (perhaps one in the US and one in Europe) and developers at the remote site would like to avoid the penalties associated with slow network connections. In such a case, there can be two “central” repositories which developers update from and commit to, with occasional merges between these centers. It can also be useful with developers gone to a conference wanting to work on the software and collaborate even when they don’t have connectivity to the “real” repository.

Another side effect of local caches being a repository is that it becomes extremely simple to mirror repositories.

Another interesting observation to make is that git allows the most offline use. There have been many times where I’ve wanted to work offline with cvs or svn projects (I’ve even resorted to rsyncing cvs repositories when I had access as a cheap hack to try to achieve this), and many times that I wished I had a copy of other branches and older versions while offline. bzr and hg are leaps and bounds better than cvs and svn in this regard, but they only partially solve this problem; using them would mean that I’d either need to manually do a checkout for every branch, that I’ll have to be online, or that I’ll have to do without information potentailly useful to me when I don’t have network connectivity. This is especially important considering that VCSes with distributed capabilities make merging easy, which encourages the use of more branches. Looking at the comparison this way, I’d really have to say that the extensive offline capabilities of git is a killer feature. I’m confused why other VCSes haven’t adopted as much local caching as git does (though I read somewhere that bzr may be considering it).

Disk usage — Client

When people see this list of varying amounts of local caching, they typically assume that disk usage is proportional to the amount of history cached, and thus believe that git will require hundreds of times the amount of diskspace to get a copy of the source code…with bzr and hg being somewhere inbetween. Reality is somewhat surprising; from my tests, the size of a checkout or clone from the various VCSes would rank in this order (with some approximate relative sizes to cvs checkouts in parentheses):

cvs (1)
git (1.92)
svn (2)
hg (2.05)
bzr (3.2) [*]

The main reason for git, hg, and bzr being so small relative to expectations is that source code packs well and these systems tend to be smart about handling metadata (information about the checkout and how to contact the server). However, there are some caveats here: my numbers (particularly for hg and bzr) aren’t based off as thorough studies as they should be, and the numbers have a higher than you’d expect variance (depends a lot on how well history of your project can pack, whether you have large files in the history that are no longer in the project, etc.) Also, while bzr and hg do automatic packing for the user, git requires the user to state when packing should be done. If the user never packs (i.e. never runs ‘git gc’) then the local repository can be much larger than a cvs or svn checkout. A basic rule of thumb is to just run ‘git gc’ after several commits, or whenever .git is larger than you think it should be.

I’ve done lots of git imports (git-cvsimport and git-svn make this easy), comparing dozens of cvs and svn repository checkouts to git ones. So I feel fairly confident about my number for git above. It does vary pretty wildly, though; e.g. for metacity it’d be 1.51 while for gtk+ it’d be 2.56[**]; I’ve seen ranges between about 0.3 and 6.0 on real world projects, so the 1.92 is just an overall mean. The hg numbers were based strictly off of converting git imports of both metacity and gtk+ to hg and taking an average of the relative difference of those (using the recent ‘hg convert’ command). My bzr number was based off importing metacity with bzr-svn and with git-svn and comparing those relative sizes (bzr-svn choked on gtk+, and I couldn’t get tailor to convert the existing git gtk+ repo to bzr).

[*] I did these tests before bzr-0.92 was out, which has a new experimental (and non-default) format that claims to drop this number significantly. I hear this new format is planned to become the default (with possibly a few tweaks) in a few months, so this is a benchmark that should be redone early next year. However, the existing number does show that bzr is already very close to an svn checkout in size despite bringing lots more information.

[**] For those wanting to duplicate, I ignored the space taken by the .git/svn directory, since that information is not representative of how much space a native git repository would take. It is interesting to note, though, that .git/svn/tags is ridiculously huge; to the point that I think it’s got to be a bug in the git-svn bridge.

Disk usage — “Central” Server

If what concerns you is the size of the repository on the central server, then the numbers are more dramatic. Benchmarks I’ve seen put git at about 1/3 the size of CVS and 1/10 the size of svn.

UPDATE: A number of people pointed me to the new named branches feature in hg that I was unaware of, which looks like it puts hg in the same category as git. Cool!

Adoption of various VCSes

There are a lot of Version Control Systems out there, and one of the biggest criteria in selecting one to use is who else uses it. I’ll try to quickly summarize what I have learned about the adoption of various VCSes. There are many people who know more than me, but here’s some of the bits that I’ve picked up.

Perceived adoption from lots of reading

I have read many blog posts, comparisons, tutorials, news articles, reader comments (in blogs and at news sites), and emails (including various VCS project archives) about version control systems. In doing so, it is clear to me that some are frequently mentioned and deemed worthy of comparison by others, while many VCSes seem so obscure that they only appear in comparisons at sites that attempt to be exhaustive or completely objective (e.g. at wikipedia). Here are the ones I hear mentioned more frequently than others:

First rung: cvs, subversion, bazaar-ng,
mercurial, tla/baz, and
git.

Though bazaar perhaps belongs in a rung below (more on that in a minute). There are also several VCSes that are still mentioned often, but not as often as the ones above:

Second rung: svk, monotone, darcs,
codeville, perforce, clearcase,
and bitkeeper.

tla/baz died a few years ago (with both developers and users mostly abandoning it for other systems, though I hear tla got revived for maintenance-only changes). Also, bazaar-ng really straddles these two levels rather than being in the upper one, but I was one of the early adopters and it has relatively strong support in the GNOME community so it’s more relevant to me. Perforce, clearcase, and bitkeeper are proprietary and thus irrelevant to me (other than as a comparison point).

Adoption according to project records

Of the non-dead open source systems, here’s a list of links to who uses them plus some comments on the links:

bazaar-ng – WhoUsesBzr – wiki page name is inconsistent; it should be “ProjectsUsingBzr” (compare to wiki page names below) :-). The page is also slightly misleading; they claim drupal as a user but my searches show otherwise (turns out to just be a developer with an unofficial mirror). Hopefully there aren’t other cases like this.
codeville – NoPage – I wasn’t able to find any list of projects using codeville anywhere. In fact, I wasn’t able to find any projects claiming to use it either. It must have shown up in other peoples’ comparisons on the basis of its interesting merge algorithm.
cvs – NoPage – I don’t have a good reference page, and it’d likely go out-of-date quickly. However, while CVS is no longer developed and projects are switching from CVS in droves these days, it wasn’t very many years ago that cvs was ubiquitous and a near universal standard. Nearly everyone familiar with at least one vcs is familiar with cvs, making it a useful reference point. Also, it still has a pretty impressive installed base; I’m even forced to use it occasionally in the open source world as well as every day at work.
darcs – ProjectsUsingDarcs – I strongly appreciate the included list of projects that stopped using their VCS (and why). Bonus points to darcs for not hiding anything.
git – ProjectsUsingGit
mercurial – ProjectsUsingMercurial – I like how they make a separate list for projects with synchronized repositories (bzr and svk ought to adopt this practice, and maybe others)
monotone – ProjectsUsingMonotone – I really like the project stats provided.
subversion – open-source-projects-using-svn – wiki page name isn’t ProjectsUsingSvn; couldn’t they read everyone else’s minds and realize that they needed such a name to fit in with the standard naming scheme? 😉
svk – ProjectsUsingSVK – claims WINE, KDE, and Ruby on Rails as users; my simple searches showed otherwise (likely svk developers just knew of developers from those projects hosting their own unofficial svk mirrors). I don’t know if their other claimed users are are accurate or not; I only checked these three.

Some adoption pages point to both the project home page and the project repositories, which is very helpful. The other adoption wiki pages should adopt that practice too, IMHO.

Adoption by “Big” users

Looking at the adoption pages listed above, each of the projects other than svk and codeville seem to have lots of users. Mostly small projects, but most projects probably are are small and it is also easier for small projects to switch to a new VCS. The real test is whether VCSes are also capable of supporting large projects. I’d like to compare on that basis, but I’m unwilling to investigate how big each listed project is. So, I’ll instead compare based on (a) if I’ve heard of the project before and know at least a little about it, and (b) I think of the project as big. This results in the following list of “big” users of various VCSes:

bazaar-ng – This is kind of surprising, but Ubuntu is the only case matching my definition above. As an added surprise, they aren’t in bzr’s list of users. (samba and drupal only have some unofficial users; and in the case of samba, I know they also have unofficial git users. Official adoption only for my comparison purposes; otherwise GNOME and KDE would be in lots of lists.)
codeville – none
cvs – Used to be used by virtually everything. Many projects still haven’t moved on yet.
darcs – none of the projects listed match my definition of “big” above
git – linux kernel (and many related projects), much of freedesktop.org (including, Xorg. HAL, DBUS, cairo, compiz), OLPC, and WINE
mercurial – opensolaris, mozilla (update: apparently mozilla hasn’t converted quite yet)
monotone – tough case. I would have possibly said none here, noting gaim, er, pidgin, as the closest but their stats suggest two projects (Xaraya and OpenEmbedded) are big…and that pidgin is bigger than I realized. I guess I’m changing my rules due to their cool use of stats.
subversion – KDE, GNOME, GCC, Samba, Python, and others
svk – none

Brief notes about each system

As a quick additional comparison point for those considering adoption, I’ll add some very brief notes about each system that I’ve gathered from my reading or experience with the system. I’ll try to list both a good point and a bad point for each.

Free/Open source VCSes
- bazaar-ng (bzr) – Developed and Evangelized by Canonical (backers of the Ubuntu distribution). Designed to be easy to use and distributed, and often gets praise for those features. It received a bit of a black eye in the early days for being horribly slow (it made cvs look like a speed demon in its early days), though I hear that the speed issues have received lots of attention and changes (and brief recent usage seems to suggest that it’s a lot better). Annoyingly, it provides misleading and less-than-useful results when passing a date to diff (the implemented behavior is well documented and apparently intentional, it’s just crap).
- codeville – Designed by Bram Cohen (inventor of bittorrent). People seem to find the merge algorithm introduced by codeville interesting. Doesn’t seem to have been adopted much, though, and it even appeared to have died for a while (going a year and a half between releases, with other updates hard to find as well). Seems to be picking back up again.
- cvs – The VCS that all other VCSes compare to, both because of its recent ubiquity and because its well known flaws are easy to leverage in promoting new alternatives. The developers working on cvs decided its existing flaws could not be fixed without a rewrite, and thus created a new system called subversion. cvs is inherently centralized.
- darcs – Really interesting and claimed easy to use research system written by David Roundy (some physicist at OSU) that is based on patches rather than source trees. I believe this allows, for example, merging between source trees that do not necessarily have common history (touted as an advanced cherry-picking algorithm that no other VCS can yet match). However, this design has an associated “doppelganger” bug that can cause darcs to become wedged and which requires care from the user to avoid. From the descriptions of this bug, it sounds like something any big project would trigger all the time (it’s an operation I’ve seen happen lots in my GNOME maintainence even on modestly sized projects like metacity.) However, developers apparently can avoid this bug if they know about it and take steps to actively avoid triggering it. I think this is related to “the conflict bug”, which can cause darcs to be slow on large repository merging, but am not sure.
- git – Invented by Linus Torvalds (inventor of the linux kernel). It has amazed a lot of people (including me) with its speed, and there are many benchmarks out there that are pretty impressive. I’ve heard/seen people claim that it is at least an order of magnitude faster than all other VCSes they’ve tried (from people who then list most all the major VCSes people think of as fast among the list of VCSes they’ve tried). It also has lots of interesting advanced features. However, versions prior to 1.5 were effectively unusable, requiring superhuman ability to learn how to use. The UI warts are being hammered away and git > 1.5 is much better usability-wise; it’s now becoming a usable system once users first learn and understand a few differences from other systems, despite its few remaining warts here and there. The online tutorials have transformed into something welcoming for new users, though the man pages (which double as the built in “–help” system) still remind me more of academic research articles written for a community of existing experts rather than user documentation. Also, no official port to windows (without cygwin) exists yet, though one is apparently getting close. Interestingly, git seems to be highly preferred as a VCS among those I consider low-level hackers.
- GNU Arch (tla/baz) – Invented by Tom Lord (who also tried to replace libc with his own rewrite). Both tla and baz are dead now with developers and users having moved on, for the most part. Proponents of these systems (particularly Tom) loudly evangelized the merits of distributed version control systems, which probably backfired since tla/baz were so horribly awful in terms of usability, complexity, quirkiness, and speed that these particular distributed VCSes really didn’t have any redeeming qualities or even salvagable pieces. (baz was written as a fork designed to make a usable tla which was backward compatible to tla; the developers eventually gave up and switched to bzr since this was an impossible goal.) I really wish I had the part of my life back I wasted learning and using these systems. And no, I don’t care about impartiality when it comes to them.
- mercurial (hg) – Written by Matt Mackall (linux kernel developer). Started two days after git, it was designed to replace bitkeeper as the VCS for the kernel. Thus, like git, it focused on speed. While not as fast as git in most benchmarks I’ve seen, it has received lots of praise for being easier to learn, having more accessible documentation, working on Windows, and still being faster than most other VCSes. The community behind mercurial seems to be a bit smaller, however: it doesn’t have nearly as many plugins as bzr or git (let alone cvs or svn). Also, it annoyingly doesn’t accept a date as an argument to diff, unlike all the other major VCSes.
- monotone (mtn) – Maintained by Nathaniel Smith and Graydon Hoare (who I don’t know of from elsewhere). The main thing I hear about this system is about it’s ideas to focus on authentication of history to verify repository contents and changes. These ideas influenced and were adopted by git and mercurial. On the con side, it appears getting an initial copy can take an extraordinarily large amount of time; for example, if you look at the developer site for pidgin you’ll note that they provide detailed steps on how to get a checkout of pidgin that involves bypassing monotone since it’s too slow to handle this on its own.
- subversion (svn) – Designed by former cvs maintainers to “be a better cvs”. It doesn’t suffer from many of the same warts as CVS; e.g. commits are atomic, files can be renamed without messing up project history, changes are per-commit rather than per-commit-per-file, and a number of operations are much faster than in cvs. Most users (myself included) feel that it is much nicer than CVS. Like CVS, svn remains inherently centralized and has no useful merge feature. Unlike CVS, half the point of tagging is inherently broken in svn as far as I can tell[*] (you can’t pass a tag to svn diff; you have to search the log trying to find the revision from which the tag was created and then use whatever revision you think is right as the revision number in svn diff).
- svk – Invented by Chia-liang Kao and now developed by Best Practical Solutions (some random company). Designed to use the subversion repository format but allow decentralized actions. I know little about their system and am hesitant to comment as I can’t think of any good comments I’ve heard (nor more than a couple bad ones.) However, on the light side of things, I absolutely love their SVKAntiFUD page. On that page, in response to the question “svk is built on top of subversion, isn’t it over-engineered and fragile?” an additional note to the answer (claimed to have been added in 2005) states that “Spaghetti code can certainly not be called over-engineered.” While the history page of their wiki suggests it has been there for at least a year, I’m guessing the maintainers don’t know about this comment and will remove it as soon as someone points it out to them.
Proprietary (i.e. included only for comparison purposes) VCSes
- bitkeeper – A system developed by BitMover Inc., founded by Larry McVoy. Gained prominence from its usage for a few years by the linux kernel. “Free Use” (as in no monetary cost) of the system by open source projects was revoked when Andrew Tridgell started reverse engineering the protocol (by telnetting to a server and typing “help”). Most users of this system seem to like it technically, but the free/open source crowd understandably often disliked its proprietary nature. I haven’t used the system, but think of it as being similar to mercurial (though I don’t know for sure if that’s the best match).
- clearcase – Developed by (the Rational Software division of) IBM. Clearcase is an exceptionally unusual VCS in that I’ve never heard anyone I know mention a positive word about it. Literally. They all seem to have stories about how it seems to hinder progress far more than it helps. There has to be someone out there that likes it (it seems to have quite a number of users for a proprietary VCS despite being exceptionally expensive), but for some reason I haven’t run across them. Very weird. I believe it is actually lock-based instead of either distributed or inherently centralized, meaning that only one person can edit any given file at a time on a given branch. Sounds mind-bogglingly crazy to me.
- perforce – Developed by Perforce Software, Inc. It seems that users of the system generally like it technically, and it has a free-of-charge clause for open source software development. My rough feeling is that Perforce is like CVS or subversion, but has a number of speed optimizations over those two. It is apparently even worse than cvs or svn for offline working, making editing not-already-opened files in the working copy problematic and error-prone unless online.

The major VCSes

Based on everything above, I consider the following VCSes to be the “major” ones:

cvs, svn, bzr, hg, and git.

I’ll add an “honorable mention” category for monotone and darcs (which bzr nearly belongs in as well, but passes based on the Canonical backing and much higher than average support by developers within the GNOME community). These five VCSes are the ones that I’ll predominantly be comparing between in my subsequent posts.

Update

[*] Kalle Vahlman in the comments points out that you can diff against a tag in svn, though it requires using atrocious syntax and a store of patience:

As much as I agree with [the claim] that SVN is just a prettier CVS, [it] isn’t really true. You can [run]:

svn diff http://svn.gnome.org/svn/metacity/tags/METACITY_2_21_1 http://svn.gnome.org/svn/metacity/trunk

to get differences between the tag and current trunk. If it looks horribly slow to you, it’s because you are on a very fast connection. IT IS SO SLOW IT MAKES LITTLE KITTENS WEEP. But it is possible anyway.

There are a number of other good posts in the comments too, pointing out project adoption cases I potentially missed and noting additional issues with some systems that I won’t be comparing later.

Starting to compare Version Control Systems

As I blogged about some time ago, I decided to spend some time learning and comparing various version control systems (VCSes for short). Of course, there are many version control system comparisons out there already, and I’ve read countless other sources as well (blogs, articles + comments, archived mailing list messages found in random google searches, etc.). While some of these sources have very interesting information, they still don’t answer all the questions I had; in fact, even the types of comparisons typically performed in these comparisons don’t cover everything I wanted to see. Here are some of the questions I have been considering:

What are the most important VCSes to consider?
Why are VCSes hard to learn? If someone learns one VCS, how much lower is their learning curve for switching to another?
What are the most common pitfalls that users experience with each of the major VCSes? Are there similarities across systems in the mistakes that users make?
Why are some systems more widely adopted than others? Are there certain qualities that make some systems more likely to be adopted by certain groups and less likely by others?
Why do some users of inherently centralized systems claim that “distributed”[1] systems are harder to learn? Why do users of distributed systems claim that they are *not* harder to learn? Why are there similar questions between the various “distributed” systems?
Which VCS is the “best” for a given individual/group? More importantly, what are the important criteria and where do various VCSes shine?
Why is there so much misunderstanding between users of different systems?
To what extent does the truism that “all software sucks” apply to VCSes?
Typical stuff: Which is the fastest at operation X (and by how much)? Which provides the most useful output (why is it more useful)? Which has the best add-ons? Which has the most relevant features? Which has the best documentation (how much better is it)? Which has killer features missing in others? etc.

I’m still far from answering all of them. However, I have learned a few things, and I figured it’d be a useful exercise to bore everyone to death by writing up some of my thoughts on the subject. So I’ll be doing that occasionally. Some of the things I write up will have comparisons similar to what you’d see elsewhere (but with my own slant which focuses on what seems relevant to me), while a few will analyze the subject from an angle different than what I have been able to find in other comparisons. I have a few posts mostly written up already, and may find time to write up a couple more after those.

Obvious Disclaimers: I’m no expert and am additionally error-prone, so I’ll likely make mistakes in my posts. I also won’t make any claims to be objective, as I don’t think it’s even possible to fully achieve. I will aim to be “reasonably” objective…but note that I have an opinion that placing too high a priority on objectivity makes it impossible to achieve much of the full usefulness of a comparison, limiting what “reasonable” can mean.

[1] As I have mentioned before, I think this is a somewhat misleading term; unfortunately, I don’t have a good replacement. Maybe something like “multi-centered”?