Mercurial's decentralized development model can be confusing to new users. This page attempts to illustrate some of the basic concepts. See the ["Tutorial"] for step-by-step instructions.

(Translations: [:BrazilianPortugueseUnderstandingMercurial:Brazilian Portuguese], [:ChineseUnderstandingMercurial:Chinese], [:FrenchUnderstandingMercurial:French], [:GermanUnderstandingMercurial:German], [:ItalianUnderstandingMercurial:Italian], [:JapaneseUnderstandingMercurial:Japanese], [:KoreanUnderstandingMercurial:Korean], [:RussianUnderstandingMercurial:Russian], [:SpanishUnderstandingMercurial:Spanish] )

TableOfContents

What's in a Repository

Mercurial [:Repository:repositories] contain a [:WorkingDirectory:working directory] coupled with a store:

The store contains the complete history of the project. Unlike traditional [:SCM:SCMs], where there's only one central copy of this history, every working directory is paired with a private copy of the history. This allows development to go on in parallel.

The working directory contains a copy of the project's files at a given point in time (eg rev 2), ready for editing. Because [:Tag:tags] and [:.hgignore:ignored files] are revision-controlled, they are also included.

Committing Changes

When you [:Commit:commit], the state of the working directory relative to its [:Parent:parents] is recorded as a new [:Revision:revision]:

Note here that revision 4 is a [:Branch:branch] of revision 2, which was the revision in the working directory. Now revision 4 is the working directory's parent.

Revisions, Changesets, Heads, and Tip

Mercurial groups related changes to multiple files into single atomic [:ChangeSet:changesets], which are revisions of the whole project. These each get a sequential [:RevisionNumber:revision number]. Because Mercurial allows distributed parallel development, these revision numbers may disagree between users. So Mercurial also assigns each revision a global [:ChangeSetID:changeset ID]. Changeset IDs are 40-digit hexadecimal numbers, but they can be abbreviated to any unambiguous prefix, like "e38487".

Branches and [:Merge:merges] in the revision history can occur at any point. Each unmerged branch creates a new [:Head:head] of the revision history. Here, revisions 5 and 6 are heads. Mercurial considers revision 6 to be the [:Tip:tip] of the repository, the head with the highest revision number. Revision 4 is a merge changeset, as it has two parent changesets (revisions 2 and 3).

Cloning, Making Changes, Merging, and Pulling

Let's start with a user Alice, who has a store that looks like:

Bob [:Clone:clones] this repo, and ends up with a complete copy of Alice's store (though his working directory is independent!):

Bob then [:Commit:commits] a couple changes:

Alice then makes her own change in parallel:

Bob then [:Pull:pulls] Alice's repo to synchronize. This copies all of Alice's changes into Bob's repo:

Because Alice's g is the newest head in Bob's repository, it's now the tip. Bob then does a [:Merge:merge] which combines the last change he was working on (f) with the tip, commits the result, and ends up with:

Now if Alice pulls from Bob, she will get Bob's changes e, f, and h, and they will be fully synchronized:

A Decentralized System

Mercurial is a completely decentralized system, and thus has no internal notion of a central repository. Thus users are free to define their own topologies for sharing changes (see CommunicatingChanges):

What Mercurial can't do

Many SVN/CVS users expect to host related projects together in one repository. This is really not what hg was made for, so you should try a different way of working. This especially means, that you cannot check out only one directory of a repository. If you absolutely need to host multiple projects in a kind of meta-repository though, you could try the ForestExtension.

For a hands-on introduction to using Mercurial, see the ["Tutorial"].