1 Git User's Manual (for version 1.5.1 or newer)
2 ______________________________________________
4 This manual is designed to be readable by someone with basic unix
5 command-line skills, but no previous knowledge of git.
7 Chapter 1 gives a brief overview of git commands, without any
8 explanation; you may prefer to skip to chapter 2 on a first reading.
10 Chapters 2 and 3 explain how to fetch and study a project using
11 git--the tools you'd need to build and test a particular version of a
12 software project, to search for regressions, and so on.
14 Chapter 4 explains how to do development with git, and chapter 5 how
15 to share that development with others.
17 Further chapters cover more specialized topics.
19 Comprehensive reference documentation is available through the man
20 pages. For a command such as "git clone", just use
22 ------------------------------------------------
24 ------------------------------------------------
30 This is a quick summary of the major commands; the following chapters
31 will explain how these work in more detail.
33 [[quick-creating-a-new-repository]]
34 Creating a new repository
35 -------------------------
39 -----------------------------------------------
40 $ tar xzf project.tar.gz
43 Initialized empty Git repository in .git/
46 -----------------------------------------------
48 From a remote repository:
50 -----------------------------------------------
51 $ git clone git://example.com/pub/project.git
53 -----------------------------------------------
59 -----------------------------------------------
60 $ git branch # list all local branches in this repo
61 $ git checkout test # switch working directory to branch "test"
62 $ git branch new # create branch "new" starting at current HEAD
63 $ git branch -d new # delete branch "new"
64 -----------------------------------------------
66 Instead of basing new branch on current HEAD (the default), use:
68 -----------------------------------------------
69 $ git branch new test # branch named "test"
70 $ git branch new v2.6.15 # tag named v2.6.15
71 $ git branch new HEAD^ # commit before the most recent
72 $ git branch new HEAD^^ # commit before that
73 $ git branch new test~10 # ten commits before tip of branch "test"
74 -----------------------------------------------
76 Create and switch to a new branch at the same time:
78 -----------------------------------------------
79 $ git checkout -b new v2.6.15
80 -----------------------------------------------
82 Update and examine branches from the repository you cloned from:
84 -----------------------------------------------
86 $ git branch -r # list
90 $ git checkout -b masterwork origin/master
91 -----------------------------------------------
93 Fetch a branch from a different repository, and give it a new
94 name in your repository:
96 -----------------------------------------------
97 $ git fetch git://example.com/project.git theirbranch:mybranch
98 $ git fetch git://example.com/project.git v2.6.15:mybranch
99 -----------------------------------------------
101 Keep a list of repositories you work with regularly:
103 -----------------------------------------------
104 $ git remote add example git://example.com/project.git
105 $ git remote # list remote repositories
108 $ git remote show example # get details
110 URL: git://example.com/project.git
111 Tracked remote branches
113 $ git fetch example # update branches from example
114 $ git branch -r # list all remote branches
115 -----------------------------------------------
118 [[exploring-history]]
122 -----------------------------------------------
123 $ gitk # visualize and browse history
124 $ git log # list all commits
125 $ git log src/ # ...modifying src/
126 $ git log v2.6.15..v2.6.16 # ...in v2.6.16, not in v2.6.15
127 $ git log master..test # ...in branch test, not in branch master
128 $ git log test..master # ...in branch master, but not in test
129 $ git log test...master # ...in one branch, not in both
130 $ git log -S'foo()' # ...where difference contain "foo()"
131 $ git log --since="2 weeks ago"
132 $ git log -p # show patches as well
133 $ git show # most recent commit
134 $ git diff v2.6.15..v2.6.16 # diff between two tagged versions
135 $ git diff v2.6.15..HEAD # diff with current head
136 $ git grep "foo()" # search working directory for "foo()"
137 $ git grep v2.6.15 "foo()" # search old tree for "foo()"
138 $ git show v2.6.15:a.txt # look at old version of a.txt
139 -----------------------------------------------
141 Search for regressions:
143 -----------------------------------------------
145 $ git bisect bad # current version is bad
146 $ git bisect good v2.6.13-rc2 # last known good revision
147 Bisecting: 675 revisions left to test after this
149 $ git bisect good # if this revision is good, or
150 $ git bisect bad # if this revision is bad.
152 -----------------------------------------------
158 Make sure git knows who to blame:
160 ------------------------------------------------
161 $ cat >>~/.gitconfig <<\EOF
163 name = Your Name Comes Here
164 email = you@yourdomain.example.com
166 ------------------------------------------------
168 Select file contents to include in the next commit, then make the
171 -----------------------------------------------
172 $ git add a.txt # updated file
173 $ git add b.txt # new file
174 $ git rm c.txt # old file
176 -----------------------------------------------
178 Or, prepare and create the commit in one step:
180 -----------------------------------------------
181 $ git commit d.txt # use latest content only of d.txt
182 $ git commit -a # use latest content of all tracked files
183 -----------------------------------------------
189 -----------------------------------------------
190 $ git merge test # merge branch "test" into the current branch
191 $ git pull git://example.com/project.git master
192 # fetch and merge in remote branch
193 $ git pull . test # equivalent to git merge test
194 -----------------------------------------------
196 [[sharing-your-changes]]
200 Importing or exporting patches:
202 -----------------------------------------------
203 $ git format-patch origin..HEAD # format a patch for each commit
204 # in HEAD but not in origin
205 $ git am mbox # import patches from the mailbox "mbox"
206 -----------------------------------------------
208 Fetch a branch in a different git repository, then merge into the
211 -----------------------------------------------
212 $ git pull git://example.com/project.git theirbranch
213 -----------------------------------------------
215 Store the fetched branch into a local branch before merging into the
218 -----------------------------------------------
219 $ git pull git://example.com/project.git theirbranch:mybranch
220 -----------------------------------------------
222 After creating commits on a local branch, update the remote
223 branch with your commits:
225 -----------------------------------------------
226 $ git push ssh://example.com/project.git mybranch:theirbranch
227 -----------------------------------------------
229 When remote and local branch are both named "test":
231 -----------------------------------------------
232 $ git push ssh://example.com/project.git test
233 -----------------------------------------------
235 Shortcut version for a frequently used remote repository:
237 -----------------------------------------------
238 $ git remote add example ssh://example.com/project.git
239 $ git push example test
240 -----------------------------------------------
242 [[repository-maintenance]]
243 Repository maintenance
244 ----------------------
246 Check for corruption:
248 -----------------------------------------------
250 -----------------------------------------------
252 Recompress, remove unused cruft:
254 -----------------------------------------------
256 -----------------------------------------------
258 [[repositories-and-branches]]
259 Repositories and Branches
260 =========================
262 [[how-to-get-a-git-repository]]
263 How to get a git repository
264 ---------------------------
266 It will be useful to have a git repository to experiment with as you
269 The best way to get one is by using the gitlink:git-clone[1] command
270 to download a copy of an existing repository for a project that you
271 are interested in. If you don't already have a project in mind, here
272 are some interesting examples:
274 ------------------------------------------------
275 # git itself (approx. 10MB download):
276 $ git clone git://git.kernel.org/pub/scm/git/git.git
277 # the linux kernel (approx. 150MB download):
278 $ git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
279 ------------------------------------------------
281 The initial clone may be time-consuming for a large project, but you
282 will only need to clone once.
284 The clone command creates a new directory named after the project
285 ("git" or "linux-2.6" in the examples above). After you cd into this
286 directory, you will see that it contains a copy of the project files,
287 together with a special top-level directory named ".git", which
288 contains all the information about the history of the project.
290 In most of the following, examples will be taken from one of the two
294 How to check out a different version of a project
295 -------------------------------------------------
297 Git is best thought of as a tool for storing the history of a
298 collection of files. It stores the history as a compressed
299 collection of interrelated snapshots (versions) of the project's
302 A single git repository may contain multiple branches. It keeps track
303 of them by keeping a list of <<def_head,heads>> which reference the
304 latest version on each branch; the gitlink:git-branch[1] command shows
305 you the list of branch heads:
307 ------------------------------------------------
310 ------------------------------------------------
312 A freshly cloned repository contains a single branch head, by default
313 named "master", with the working directory initialized to the state of
314 the project referred to by that branch head.
316 Most projects also use <<def_tag,tags>>. Tags, like heads, are
317 references into the project's history, and can be listed using the
318 gitlink:git-tag[1] command:
320 ------------------------------------------------
332 ------------------------------------------------
334 Tags are expected to always point at the same version of a project,
335 while heads are expected to advance as development progresses.
337 Create a new branch head pointing to one of these versions and check it
338 out using gitlink:git-checkout[1]:
340 ------------------------------------------------
341 $ git checkout -b new v2.6.13
342 ------------------------------------------------
344 The working directory then reflects the contents that the project had
345 when it was tagged v2.6.13, and gitlink:git-branch[1] shows two
346 branches, with an asterisk marking the currently checked-out branch:
348 ------------------------------------------------
352 ------------------------------------------------
354 If you decide that you'd rather see version 2.6.17, you can modify
355 the current branch to point at v2.6.17 instead, with
357 ------------------------------------------------
358 $ git reset --hard v2.6.17
359 ------------------------------------------------
361 Note that if the current branch head was your only reference to a
362 particular point in history, then resetting that branch may leave you
363 with no way to find the history it used to point to; so use this command
366 [[understanding-commits]]
367 Understanding History: Commits
368 ------------------------------
370 Every change in the history of a project is represented by a commit.
371 The gitlink:git-show[1] command shows the most recent commit on the
374 ------------------------------------------------
376 commit 2b5f6dcce5bf94b9b119e9ed8d537098ec61c3d2
377 Author: Jamal Hadi Salim <hadi@cyberus.ca>
378 Date: Sat Dec 2 22:22:25 2006 -0800
380 [XFRM]: Fix aevent structuring to be more complete.
382 aevents can not uniquely identify an SA. We break the ABI with this
383 patch, but consensus is that since it is not yet utilized by any
384 (known) application then it is fine (better do it now than later).
386 Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
387 Signed-off-by: David S. Miller <davem@davemloft.net>
389 diff --git a/Documentation/networking/xfrm_sync.txt b/Documentation/networking/xfrm_sync.txt
390 index 8be626f..d7aac9d 100644
391 --- a/Documentation/networking/xfrm_sync.txt
392 +++ b/Documentation/networking/xfrm_sync.txt
393 @@ -47,10 +47,13 @@ aevent_id structure looks like:
395 struct xfrm_aevent_id {
396 struct xfrm_usersa_id sa_id;
397 + xfrm_address_t saddr;
402 ------------------------------------------------
404 As you can see, a commit shows who made the latest change, what they
407 Every commit has a 40-hexdigit id, sometimes called the "object name" or the
408 "SHA1 id", shown on the first line of the "git show" output. You can usually
409 refer to a commit by a shorter name, such as a tag or a branch name, but this
410 longer name can also be useful. Most importantly, it is a globally unique
411 name for this commit: so if you tell somebody else the object name (for
412 example in email), then you are guaranteed that name will refer to the same
413 commit in their repository that it does in yours (assuming their repository
414 has that commit at all). Since the object name is computed as a hash over the
415 contents of the commit, you are guaranteed that the commit can never change
416 without its name also changing.
418 In fact, in <<git-internals>> we shall see that everything stored in git
419 history, including file data and directory contents, is stored in an object
420 with a name that is a hash of its contents.
422 [[understanding-reachability]]
423 Understanding history: commits, parents, and reachability
424 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
426 Every commit (except the very first commit in a project) also has a
427 parent commit which shows what happened before this commit.
428 Following the chain of parents will eventually take you back to the
429 beginning of the project.
431 However, the commits do not form a simple list; git allows lines of
432 development to diverge and then reconverge, and the point where two
433 lines of development reconverge is called a "merge". The commit
434 representing a merge can therefore have more than one parent, with
435 each parent representing the most recent commit on one of the lines
436 of development leading to that point.
438 The best way to see how this works is using the gitlink:gitk[1]
439 command; running gitk now on a git repository and looking for merge
440 commits will help understand how the git organizes history.
442 In the following, we say that commit X is "reachable" from commit Y
443 if commit X is an ancestor of commit Y. Equivalently, you could say
444 that Y is a descendent of X, or that there is a chain of parents
445 leading from commit Y to commit X.
448 Understanding history: History diagrams
449 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
451 We will sometimes represent git history using diagrams like the one
452 below. Commits are shown as "o", and the links between them with
453 lines drawn with - / and \. Time goes left to right:
456 ................................................
462 ................................................
464 If we need to talk about a particular commit, the character "o" may
465 be replaced with another letter or number.
468 Understanding history: What is a branch?
469 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
471 When we need to be precise, we will use the word "branch" to mean a line
472 of development, and "branch head" (or just "head") to mean a reference
473 to the most recent commit on a branch. In the example above, the branch
474 head named "A" is a pointer to one particular commit, but we refer to
475 the line of three commits leading up to that point as all being part of
478 However, when no confusion will result, we often just use the term
479 "branch" both for branches and for branch heads.
481 [[manipulating-branches]]
482 Manipulating branches
483 ---------------------
485 Creating, deleting, and modifying branches is quick and easy; here's
486 a summary of the commands:
490 git branch <branch>::
491 create a new branch named <branch>, referencing the same
492 point in history as the current branch
493 git branch <branch> <start-point>::
494 create a new branch named <branch>, referencing
495 <start-point>, which may be specified any way you like,
496 including using a branch name or a tag name
497 git branch -d <branch>::
498 delete the branch <branch>; if the branch you are deleting
499 points to a commit which is not reachable from the current
500 branch, this command will fail with a warning.
501 git branch -D <branch>::
502 even if the branch points to a commit not reachable
503 from the current branch, you may know that that commit
504 is still reachable from some other branch or tag. In that
505 case it is safe to use this command to force git to delete
507 git checkout <branch>::
508 make the current branch <branch>, updating the working
509 directory to reflect the version referenced by <branch>
510 git checkout -b <new> <start-point>::
511 create a new branch <new> referencing <start-point>, and
514 The special symbol "HEAD" can always be used to refer to the current
515 branch. In fact, git uses a file named "HEAD" in the .git directory to
516 remember which branch is current:
518 ------------------------------------------------
520 ref: refs/heads/master
521 ------------------------------------------------
524 Examining an old version without creating a new branch
525 ------------------------------------------------------
527 The git-checkout command normally expects a branch head, but will also
528 accept an arbitrary commit; for example, you can check out the commit
531 ------------------------------------------------
532 $ git checkout v2.6.17
533 Note: moving to "v2.6.17" which isn't a local branch
534 If you want to create a new branch from this checkout, you may do so
535 (now or later) by using -b with the checkout command again. Example:
536 git checkout -b <new_branch_name>
537 HEAD is now at 427abfa... Linux v2.6.17
538 ------------------------------------------------
540 The HEAD then refers to the SHA1 of the commit instead of to a branch,
541 and git branch shows that you are no longer on a branch:
543 ------------------------------------------------
545 427abfa28afedffadfca9dd8b067eb6d36bac53f
549 ------------------------------------------------
551 In this case we say that the HEAD is "detached".
553 This is an easy way to check out a particular version without having to
554 make up a name for the new branch. You can still create a new branch
555 (or tag) for this version later if you decide to.
557 [[examining-remote-branches]]
558 Examining branches from a remote repository
559 -------------------------------------------
561 The "master" branch that was created at the time you cloned is a copy
562 of the HEAD in the repository that you cloned from. That repository
563 may also have had other branches, though, and your local repository
564 keeps branches which track each of those remote branches, which you
565 can view using the "-r" option to gitlink:git-branch[1]:
567 ------------------------------------------------
577 ------------------------------------------------
579 You cannot check out these remote-tracking branches, but you can
580 examine them on a branch of your own, just as you would a tag:
582 ------------------------------------------------
583 $ git checkout -b my-todo-copy origin/todo
584 ------------------------------------------------
586 Note that the name "origin" is just the name that git uses by default
587 to refer to the repository that you cloned from.
589 [[how-git-stores-references]]
590 Naming branches, tags, and other references
591 -------------------------------------------
593 Branches, remote-tracking branches, and tags are all references to
594 commits. All references are named with a slash-separated path name
595 starting with "refs"; the names we've been using so far are actually
598 - The branch "test" is short for "refs/heads/test".
599 - The tag "v2.6.18" is short for "refs/tags/v2.6.18".
600 - "origin/master" is short for "refs/remotes/origin/master".
602 The full name is occasionally useful if, for example, there ever
603 exists a tag and a branch with the same name.
605 As another useful shortcut, the "HEAD" of a repository can be referred
606 to just using the name of that repository. So, for example, "origin"
607 is usually a shortcut for the HEAD branch in the repository "origin".
609 For the complete list of paths which git checks for references, and
610 the order it uses to decide which to choose when there are multiple
611 references with the same shorthand name, see the "SPECIFYING
612 REVISIONS" section of gitlink:git-rev-parse[1].
614 [[Updating-a-repository-with-git-fetch]]
615 Updating a repository with git fetch
616 ------------------------------------
618 Eventually the developer cloned from will do additional work in her
619 repository, creating new commits and advancing the branches to point
622 The command "git fetch", with no arguments, will update all of the
623 remote-tracking branches to the latest version found in her
624 repository. It will not touch any of your own branches--not even the
625 "master" branch that was created for you on clone.
627 [[fetching-branches]]
628 Fetching branches from other repositories
629 -----------------------------------------
631 You can also track branches from repositories other than the one you
632 cloned from, using gitlink:git-remote[1]:
634 -------------------------------------------------
635 $ git remote add linux-nfs git://linux-nfs.org/pub/nfs-2.6.git
636 $ git fetch linux-nfs
637 * refs/remotes/linux-nfs/master: storing branch 'master' ...
639 -------------------------------------------------
641 New remote-tracking branches will be stored under the shorthand name
642 that you gave "git remote add", in this case linux-nfs:
644 -------------------------------------------------
648 -------------------------------------------------
650 If you run "git fetch <remote>" later, the tracking branches for the
651 named <remote> will be updated.
653 If you examine the file .git/config, you will see that git has added
656 -------------------------------------------------
660 url = git://linux-nfs.org/pub/nfs-2.6.git
661 fetch = +refs/heads/*:refs/remotes/linux-nfs/*
663 -------------------------------------------------
665 This is what causes git to track the remote's branches; you may modify
666 or delete these configuration options by editing .git/config with a
667 text editor. (See the "CONFIGURATION FILE" section of
668 gitlink:git-config[1] for details.)
670 [[exploring-git-history]]
671 Exploring git history
672 =====================
674 Git is best thought of as a tool for storing the history of a
675 collection of files. It does this by storing compressed snapshots of
676 the contents of a file heirarchy, together with "commits" which show
677 the relationships between these snapshots.
679 Git provides extremely flexible and fast tools for exploring the
680 history of a project.
682 We start with one specialized tool that is useful for finding the
683 commit that introduced a bug into a project.
686 How to use bisect to find a regression
687 --------------------------------------
689 Suppose version 2.6.18 of your project worked, but the version at
690 "master" crashes. Sometimes the best way to find the cause of such a
691 regression is to perform a brute-force search through the project's
692 history to find the particular commit that caused the problem. The
693 gitlink:git-bisect[1] command can help you do this:
695 -------------------------------------------------
697 $ git bisect good v2.6.18
698 $ git bisect bad master
699 Bisecting: 3537 revisions left to test after this
700 [65934a9a028b88e83e2b0f8b36618fe503349f8e] BLOCK: Make USB storage depend on SCSI rather than selecting it [try #6]
701 -------------------------------------------------
703 If you run "git branch" at this point, you'll see that git has
704 temporarily moved you to a new branch named "bisect". This branch
705 points to a commit (with commit id 65934...) that is reachable from
706 v2.6.19 but not from v2.6.18. Compile and test it, and see whether
707 it crashes. Assume it does crash. Then:
709 -------------------------------------------------
711 Bisecting: 1769 revisions left to test after this
712 [7eff82c8b1511017ae605f0c99ac275a7e21b867] i2c-core: Drop useless bitmaskings
713 -------------------------------------------------
715 checks out an older version. Continue like this, telling git at each
716 stage whether the version it gives you is good or bad, and notice
717 that the number of revisions left to test is cut approximately in
720 After about 13 tests (in this case), it will output the commit id of
721 the guilty commit. You can then examine the commit with
722 gitlink:git-show[1], find out who wrote it, and mail them your bug
723 report with the commit id. Finally, run
725 -------------------------------------------------
727 -------------------------------------------------
729 to return you to the branch you were on before and delete the
730 temporary "bisect" branch.
732 Note that the version which git-bisect checks out for you at each
733 point is just a suggestion, and you're free to try a different
734 version if you think it would be a good idea. For example,
735 occasionally you may land on a commit that broke something unrelated;
738 -------------------------------------------------
739 $ git bisect visualize
740 -------------------------------------------------
742 which will run gitk and label the commit it chose with a marker that
743 says "bisect". Chose a safe-looking commit nearby, note its commit
744 id, and check it out with:
746 -------------------------------------------------
747 $ git reset --hard fb47ddb2db...
748 -------------------------------------------------
750 then test, run "bisect good" or "bisect bad" as appropriate, and
757 We have seen several ways of naming commits already:
759 - 40-hexdigit object name
760 - branch name: refers to the commit at the head of the given
762 - tag name: refers to the commit pointed to by the given tag
763 (we've seen branches and tags are special cases of
764 <<how-git-stores-references,references>>).
765 - HEAD: refers to the head of the current branch
767 There are many more; see the "SPECIFYING REVISIONS" section of the
768 gitlink:git-rev-parse[1] man page for the complete list of ways to
769 name revisions. Some examples:
771 -------------------------------------------------
772 $ git show fb47ddb2 # the first few characters of the object name
773 # are usually enough to specify it uniquely
774 $ git show HEAD^ # the parent of the HEAD commit
775 $ git show HEAD^^ # the grandparent
776 $ git show HEAD~4 # the great-great-grandparent
777 -------------------------------------------------
779 Recall that merge commits may have more than one parent; by default,
780 ^ and ~ follow the first parent listed in the commit, but you can
783 -------------------------------------------------
784 $ git show HEAD^1 # show the first parent of HEAD
785 $ git show HEAD^2 # show the second parent of HEAD
786 -------------------------------------------------
788 In addition to HEAD, there are several other special names for
791 Merges (to be discussed later), as well as operations such as
792 git-reset, which change the currently checked-out commit, generally
793 set ORIG_HEAD to the value HEAD had before the current operation.
795 The git-fetch operation always stores the head of the last fetched
796 branch in FETCH_HEAD. For example, if you run git fetch without
797 specifying a local branch as the target of the operation
799 -------------------------------------------------
800 $ git fetch git://example.com/proj.git theirbranch
801 -------------------------------------------------
803 the fetched commits will still be available from FETCH_HEAD.
805 When we discuss merges we'll also see the special name MERGE_HEAD,
806 which refers to the other branch that we're merging in to the current
809 The gitlink:git-rev-parse[1] command is a low-level command that is
810 occasionally useful for translating some name for a commit to the object
811 name for that commit:
813 -------------------------------------------------
814 $ git rev-parse origin
815 e05db0fd4f31dde7005f075a84f96b360d05984b
816 -------------------------------------------------
822 We can also create a tag to refer to a particular commit; after
825 -------------------------------------------------
826 $ git tag stable-1 1b2e1d63ff
827 -------------------------------------------------
829 You can use stable-1 to refer to the commit 1b2e1d63ff.
831 This creates a "lightweight" tag. If you would also like to include a
832 comment with the tag, and possibly sign it cryptographically, then you
833 should create a tag object instead; see the gitlink:git-tag[1] man page
836 [[browsing-revisions]]
840 The gitlink:git-log[1] command can show lists of commits. On its
841 own, it shows all commits reachable from the parent commit; but you
842 can also make more specific requests:
844 -------------------------------------------------
845 $ git log v2.5.. # commits since (not reachable from) v2.5
846 $ git log test..master # commits reachable from master but not test
847 $ git log master..test # ...reachable from test but not master
848 $ git log master...test # ...reachable from either test or master,
850 $ git log --since="2 weeks ago" # commits from the last 2 weeks
851 $ git log Makefile # commits which modify Makefile
852 $ git log fs/ # ... which modify any file under fs/
853 $ git log -S'foo()' # commits which add or remove any file data
854 # matching the string 'foo()'
855 -------------------------------------------------
857 And of course you can combine all of these; the following finds
858 commits since v2.5 which touch the Makefile or any file under fs:
860 -------------------------------------------------
861 $ git log v2.5.. Makefile fs/
862 -------------------------------------------------
864 You can also ask git log to show patches:
866 -------------------------------------------------
868 -------------------------------------------------
870 See the "--pretty" option in the gitlink:git-log[1] man page for more
873 Note that git log starts with the most recent commit and works
874 backwards through the parents; however, since git history can contain
875 multiple independent lines of development, the particular order that
876 commits are listed in may be somewhat arbitrary.
882 You can generate diffs between any two versions using
885 -------------------------------------------------
886 $ git diff master..test
887 -------------------------------------------------
889 Sometimes what you want instead is a set of patches:
891 -------------------------------------------------
892 $ git format-patch master..test
893 -------------------------------------------------
895 will generate a file with a patch for each commit reachable from test
896 but not from master. Note that if master also has commits which are
897 not reachable from test, then the combined result of these patches
898 will not be the same as the diff produced by the git-diff example.
900 [[viewing-old-file-versions]]
901 Viewing old file versions
902 -------------------------
904 You can always view an old version of a file by just checking out the
905 correct revision first. But sometimes it is more convenient to be
906 able to view an old version of a single file without checking
907 anything out; this command does that:
909 -------------------------------------------------
910 $ git show v2.5:fs/locks.c
911 -------------------------------------------------
913 Before the colon may be anything that names a commit, and after it
914 may be any path to a file tracked by git.
920 [[checking-for-equal-branches]]
921 Check whether two branches point at the same history
922 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
924 Suppose you want to check whether two branches point at the same point
927 -------------------------------------------------
928 $ git diff origin..master
929 -------------------------------------------------
931 will tell you whether the contents of the project are the same at the
932 two branches; in theory, however, it's possible that the same project
933 contents could have been arrived at by two different historical
934 routes. You could compare the object names:
936 -------------------------------------------------
937 $ git rev-list origin
938 e05db0fd4f31dde7005f075a84f96b360d05984b
939 $ git rev-list master
940 e05db0fd4f31dde7005f075a84f96b360d05984b
941 -------------------------------------------------
943 Or you could recall that the ... operator selects all commits
944 contained reachable from either one reference or the other but not
947 -------------------------------------------------
948 $ git log origin...master
949 -------------------------------------------------
951 will return no commits when the two branches are equal.
953 [[finding-tagged-descendants]]
954 Find first tagged version including a given fix
955 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
957 Suppose you know that the commit e05db0fd fixed a certain problem.
958 You'd like to find the earliest tagged release that contains that
961 Of course, there may be more than one answer--if the history branched
962 after commit e05db0fd, then there could be multiple "earliest" tagged
965 You could just visually inspect the commits since e05db0fd:
967 -------------------------------------------------
969 -------------------------------------------------
971 Or you can use gitlink:git-name-rev[1], which will give the commit a
972 name based on any tag it finds pointing to one of the commit's
975 -------------------------------------------------
976 $ git name-rev --tags e05db0fd
977 e05db0fd tags/v1.5.0-rc1^0~23
978 -------------------------------------------------
980 The gitlink:git-describe[1] command does the opposite, naming the
981 revision using a tag on which the given commit is based:
983 -------------------------------------------------
984 $ git describe e05db0fd
985 v1.5.0-rc0-260-ge05db0f
986 -------------------------------------------------
988 but that may sometimes help you guess which tags might come after the
991 If you just want to verify whether a given tagged version contains a
992 given commit, you could use gitlink:git-merge-base[1]:
994 -------------------------------------------------
995 $ git merge-base e05db0fd v1.5.0-rc1
996 e05db0fd4f31dde7005f075a84f96b360d05984b
997 -------------------------------------------------
999 The merge-base command finds a common ancestor of the given commits,
1000 and always returns one or the other in the case where one is a
1001 descendant of the other; so the above output shows that e05db0fd
1002 actually is an ancestor of v1.5.0-rc1.
1004 Alternatively, note that
1006 -------------------------------------------------
1007 $ git log v1.5.0-rc1..e05db0fd
1008 -------------------------------------------------
1010 will produce empty output if and only if v1.5.0-rc1 includes e05db0fd,
1011 because it outputs only commits that are not reachable from v1.5.0-rc1.
1013 As yet another alternative, the gitlink:git-show-branch[1] command lists
1014 the commits reachable from its arguments with a display on the left-hand
1015 side that indicates which arguments that commit is reachable from. So,
1016 you can run something like
1018 -------------------------------------------------
1019 $ git show-branch e05db0fd v1.5.0-rc0 v1.5.0-rc1 v1.5.0-rc2
1020 ! [e05db0fd] Fix warnings in sha1_file.c - use C99 printf format if
1022 ! [v1.5.0-rc0] GIT v1.5.0 preview
1023 ! [v1.5.0-rc1] GIT v1.5.0-rc1
1024 ! [v1.5.0-rc2] GIT v1.5.0-rc2
1026 -------------------------------------------------
1028 then search for a line that looks like
1030 -------------------------------------------------
1031 + ++ [e05db0fd] Fix warnings in sha1_file.c - use C99 printf format if
1033 -------------------------------------------------
1035 Which shows that e05db0fd is reachable from itself, from v1.5.0-rc1, and
1036 from v1.5.0-rc2, but not from v1.5.0-rc0.
1039 [[Developing-with-git]]
1043 [[telling-git-your-name]]
1044 Telling git your name
1045 ---------------------
1047 Before creating any commits, you should introduce yourself to git. The
1048 easiest way to do so is to make sure the following lines appear in a
1049 file named .gitconfig in your home directory:
1051 ------------------------------------------------
1053 name = Your Name Comes Here
1054 email = you@yourdomain.example.com
1055 ------------------------------------------------
1057 (See the "CONFIGURATION FILE" section of gitlink:git-config[1] for
1058 details on the configuration file.)
1061 [[creating-a-new-repository]]
1062 Creating a new repository
1063 -------------------------
1065 Creating a new repository from scratch is very easy:
1067 -------------------------------------------------
1071 -------------------------------------------------
1073 If you have some initial content (say, a tarball):
1075 -------------------------------------------------
1076 $ tar -xzvf project.tar.gz
1079 $ git add . # include everything below ./ in the first commit:
1081 -------------------------------------------------
1083 [[how-to-make-a-commit]]
1084 How to make a commit
1085 --------------------
1087 Creating a new commit takes three steps:
1089 1. Making some changes to the working directory using your
1091 2. Telling git about your changes.
1092 3. Creating the commit using the content you told git about
1095 In practice, you can interleave and repeat steps 1 and 2 as many
1096 times as you want: in order to keep track of what you want committed
1097 at step 3, git maintains a snapshot of the tree's contents in a
1098 special staging area called "the index."
1100 At the beginning, the content of the index will be identical to
1101 that of the HEAD. The command "git diff --cached", which shows
1102 the difference between the HEAD and the index, should therefore
1103 produce no output at that point.
1105 Modifying the index is easy:
1107 To update the index with the new contents of a modified file, use
1109 -------------------------------------------------
1110 $ git add path/to/file
1111 -------------------------------------------------
1113 To add the contents of a new file to the index, use
1115 -------------------------------------------------
1116 $ git add path/to/file
1117 -------------------------------------------------
1119 To remove a file from the index and from the working tree,
1121 -------------------------------------------------
1122 $ git rm path/to/file
1123 -------------------------------------------------
1125 After each step you can verify that
1127 -------------------------------------------------
1129 -------------------------------------------------
1131 always shows the difference between the HEAD and the index file--this
1132 is what you'd commit if you created the commit now--and that
1134 -------------------------------------------------
1136 -------------------------------------------------
1138 shows the difference between the working tree and the index file.
1140 Note that "git add" always adds just the current contents of a file
1141 to the index; further changes to the same file will be ignored unless
1142 you run git-add on the file again.
1144 When you're ready, just run
1146 -------------------------------------------------
1148 -------------------------------------------------
1150 and git will prompt you for a commit message and then create the new
1151 commit. Check to make sure it looks like what you expected with
1153 -------------------------------------------------
1155 -------------------------------------------------
1157 As a special shortcut,
1159 -------------------------------------------------
1161 -------------------------------------------------
1163 will update the index with any files that you've modified or removed
1164 and create a commit, all in one step.
1166 A number of commands are useful for keeping track of what you're
1169 -------------------------------------------------
1170 $ git diff --cached # difference between HEAD and the index; what
1171 # would be commited if you ran "commit" now.
1172 $ git diff # difference between the index file and your
1173 # working directory; changes that would not
1174 # be included if you ran "commit" now.
1175 $ git diff HEAD # difference between HEAD and working tree; what
1176 # would be committed if you ran "commit -a" now.
1177 $ git status # a brief per-file summary of the above.
1178 -------------------------------------------------
1180 [[creating-good-commit-messages]]
1181 Creating good commit messages
1182 -----------------------------
1184 Though not required, it's a good idea to begin the commit message
1185 with a single short (less than 50 character) line summarizing the
1186 change, followed by a blank line and then a more thorough
1187 description. Tools that turn commits into email, for example, use
1188 the first line on the Subject line and the rest of the commit in the
1195 You can rejoin two diverging branches of development using
1196 gitlink:git-merge[1]:
1198 -------------------------------------------------
1199 $ git merge branchname
1200 -------------------------------------------------
1202 merges the development in the branch "branchname" into the current
1203 branch. If there are conflicts--for example, if the same file is
1204 modified in two different ways in the remote branch and the local
1205 branch--then you are warned; the output may look something like this:
1207 -------------------------------------------------
1210 Auto-merged file.txt
1211 CONFLICT (content): Merge conflict in file.txt
1212 Automatic merge failed; fix conflicts and then commit the result.
1213 -------------------------------------------------
1215 Conflict markers are left in the problematic files, and after
1216 you resolve the conflicts manually, you can update the index
1217 with the contents and run git commit, as you normally would when
1218 creating a new file.
1220 If you examine the resulting commit using gitk, you will see that it
1221 has two parents, one pointing to the top of the current branch, and
1222 one to the top of the other branch.
1224 [[resolving-a-merge]]
1228 When a merge isn't resolved automatically, git leaves the index and
1229 the working tree in a special state that gives you all the
1230 information you need to help resolve the merge.
1232 Files with conflicts are marked specially in the index, so until you
1233 resolve the problem and update the index, gitlink:git-commit[1] will
1236 -------------------------------------------------
1238 file.txt: needs merge
1239 -------------------------------------------------
1241 Also, gitlink:git-status[1] will list those files as "unmerged", and the
1242 files with conflicts will have conflict markers added, like this:
1244 -------------------------------------------------
1245 <<<<<<< HEAD:file.txt
1249 >>>>>>> 77976da35a11db4580b80ae27e8d65caf5208086:file.txt
1250 -------------------------------------------------
1252 All you need to do is edit the files to resolve the conflicts, and then
1254 -------------------------------------------------
1257 -------------------------------------------------
1259 Note that the commit message will already be filled in for you with
1260 some information about the merge. Normally you can just use this
1261 default message unchanged, but you may add additional commentary of
1262 your own if desired.
1264 The above is all you need to know to resolve a simple merge. But git
1265 also provides more information to help resolve conflicts:
1267 [[conflict-resolution]]
1268 Getting conflict-resolution help during a merge
1269 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1271 All of the changes that git was able to merge automatically are
1272 already added to the index file, so gitlink:git-diff[1] shows only
1273 the conflicts. It uses an unusual syntax:
1275 -------------------------------------------------
1278 index 802992c,2b60207..0000000
1281 @@@ -1,1 -1,1 +1,5 @@@
1282 ++<<<<<<< HEAD:file.txt
1286 ++>>>>>>> 77976da35a11db4580b80ae27e8d65caf5208086:file.txt
1287 -------------------------------------------------
1289 Recall that the commit which will be commited after we resolve this
1290 conflict will have two parents instead of the usual one: one parent
1291 will be HEAD, the tip of the current branch; the other will be the
1292 tip of the other branch, which is stored temporarily in MERGE_HEAD.
1294 During the merge, the index holds three versions of each file. Each of
1295 these three "file stages" represents a different version of the file:
1297 -------------------------------------------------
1298 $ git show :1:file.txt # the file in a common ancestor of both branches
1299 $ git show :2:file.txt # the version from HEAD, but including any
1300 # nonconflicting changes from MERGE_HEAD
1301 $ git show :3:file.txt # the version from MERGE_HEAD, but including any
1302 # nonconflicting changes from HEAD.
1303 -------------------------------------------------
1305 Since the stage 2 and stage 3 versions have already been updated with
1306 nonconflicting changes, the only remaining differences between them are
1307 the important ones; thus gitlink:git-diff[1] can use the information in
1308 the index to show only those conflicts.
1310 The diff above shows the differences between the working-tree version of
1311 file.txt and the stage 2 and stage 3 versions. So instead of preceding
1312 each line by a single "+" or "-", it now uses two columns: the first
1313 column is used for differences between the first parent and the working
1314 directory copy, and the second for differences between the second parent
1315 and the working directory copy. (See the "COMBINED DIFF FORMAT" section
1316 of gitlink:git-diff-files[1] for a details of the format.)
1318 After resolving the conflict in the obvious way (but before updating the
1319 index), the diff will look like:
1321 -------------------------------------------------
1324 index 802992c,2b60207..0000000
1327 @@@ -1,1 -1,1 +1,1 @@@
1331 -------------------------------------------------
1333 This shows that our resolved version deleted "Hello world" from the
1334 first parent, deleted "Goodbye" from the second parent, and added
1335 "Goodbye world", which was previously absent from both.
1337 Some special diff options allow diffing the working directory against
1338 any of these stages:
1340 -------------------------------------------------
1341 $ git diff -1 file.txt # diff against stage 1
1342 $ git diff --base file.txt # same as the above
1343 $ git diff -2 file.txt # diff against stage 2
1344 $ git diff --ours file.txt # same as the above
1345 $ git diff -3 file.txt # diff against stage 3
1346 $ git diff --theirs file.txt # same as the above.
1347 -------------------------------------------------
1349 The gitlink:git-log[1] and gitk[1] commands also provide special help
1352 -------------------------------------------------
1355 -------------------------------------------------
1357 These will display all commits which exist only on HEAD or on
1358 MERGE_HEAD, and which touch an unmerged file.
1360 You may also use gitlink:git-mergetool[1], which lets you merge the
1361 unmerged files using external tools such as emacs or kdiff3.
1363 Each time you resolve the conflicts in a file and update the index:
1365 -------------------------------------------------
1367 -------------------------------------------------
1369 the different stages of that file will be "collapsed", after which
1370 git-diff will (by default) no longer show diffs for that file.
1376 If you get stuck and decide to just give up and throw the whole mess
1377 away, you can always return to the pre-merge state with
1379 -------------------------------------------------
1380 $ git reset --hard HEAD
1381 -------------------------------------------------
1383 Or, if you've already commited the merge that you want to throw away,
1385 -------------------------------------------------
1386 $ git reset --hard ORIG_HEAD
1387 -------------------------------------------------
1389 However, this last command can be dangerous in some cases--never
1390 throw away a commit you have already committed if that commit may
1391 itself have been merged into another branch, as doing so may confuse
1398 There is one special case not mentioned above, which is treated
1399 differently. Normally, a merge results in a merge commit, with two
1400 parents, one pointing at each of the two lines of development that
1403 However, if the current branch is a descendant of the other--so every
1404 commit present in the one is already contained in the other--then git
1405 just performs a "fast forward"; the head of the current branch is moved
1406 forward to point at the head of the merged-in branch, without any new
1407 commits being created.
1413 If you've messed up the working tree, but haven't yet committed your
1414 mistake, you can return the entire working tree to the last committed
1417 -------------------------------------------------
1418 $ git reset --hard HEAD
1419 -------------------------------------------------
1421 If you make a commit that you later wish you hadn't, there are two
1422 fundamentally different ways to fix the problem:
1424 1. You can create a new commit that undoes whatever was done
1425 by the previous commit. This is the correct thing if your
1426 mistake has already been made public.
1428 2. You can go back and modify the old commit. You should
1429 never do this if you have already made the history public;
1430 git does not normally expect the "history" of a project to
1431 change, and cannot correctly perform repeated merges from
1432 a branch that has had its history changed.
1434 [[reverting-a-commit]]
1435 Fixing a mistake with a new commit
1436 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1438 Creating a new commit that reverts an earlier change is very easy;
1439 just pass the gitlink:git-revert[1] command a reference to the bad
1440 commit; for example, to revert the most recent commit:
1442 -------------------------------------------------
1444 -------------------------------------------------
1446 This will create a new commit which undoes the change in HEAD. You
1447 will be given a chance to edit the commit message for the new commit.
1449 You can also revert an earlier change, for example, the next-to-last:
1451 -------------------------------------------------
1453 -------------------------------------------------
1455 In this case git will attempt to undo the old change while leaving
1456 intact any changes made since then. If more recent changes overlap
1457 with the changes to be reverted, then you will be asked to fix
1458 conflicts manually, just as in the case of <<resolving-a-merge,
1459 resolving a merge>>.
1461 [[fixing-a-mistake-by-editing-history]]
1462 Fixing a mistake by editing history
1463 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1465 If the problematic commit is the most recent commit, and you have not
1466 yet made that commit public, then you may just
1467 <<undoing-a-merge,destroy it using git-reset>>.
1470 can edit the working directory and update the index to fix your
1471 mistake, just as if you were going to <<how-to-make-a-commit,create a
1472 new commit>>, then run
1474 -------------------------------------------------
1475 $ git commit --amend
1476 -------------------------------------------------
1478 which will replace the old commit by a new commit incorporating your
1479 changes, giving you a chance to edit the old commit message first.
1481 Again, you should never do this to a commit that may already have
1482 been merged into another branch; use gitlink:git-revert[1] instead in
1485 It is also possible to edit commits further back in the history, but
1486 this is an advanced topic to be left for
1487 <<cleaning-up-history,another chapter>>.
1489 [[checkout-of-path]]
1490 Checking out an old version of a file
1491 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1493 In the process of undoing a previous bad change, you may find it
1494 useful to check out an older version of a particular file using
1495 gitlink:git-checkout[1]. We've used git checkout before to switch
1496 branches, but it has quite different behavior if it is given a path
1499 -------------------------------------------------
1500 $ git checkout HEAD^ path/to/file
1501 -------------------------------------------------
1503 replaces path/to/file by the contents it had in the commit HEAD^, and
1504 also updates the index to match. It does not change branches.
1506 If you just want to look at an old version of the file, without
1507 modifying the working directory, you can do that with
1508 gitlink:git-show[1]:
1510 -------------------------------------------------
1511 $ git show HEAD^:path/to/file
1512 -------------------------------------------------
1514 which will display the given version of the file.
1516 [[ensuring-good-performance]]
1517 Ensuring good performance
1518 -------------------------
1520 On large repositories, git depends on compression to keep the history
1521 information from taking up to much space on disk or in memory.
1523 This compression is not performed automatically. Therefore you
1524 should occasionally run gitlink:git-gc[1]:
1526 -------------------------------------------------
1528 -------------------------------------------------
1530 to recompress the archive. This can be very time-consuming, so
1531 you may prefer to run git-gc when you are not doing other work.
1534 [[ensuring-reliability]]
1535 Ensuring reliability
1536 --------------------
1538 [[checking-for-corruption]]
1539 Checking the repository for corruption
1540 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1542 The gitlink:git-fsck[1] command runs a number of self-consistency checks
1543 on the repository, and reports on any problems. This may take some
1544 time. The most common warning by far is about "dangling" objects:
1546 -------------------------------------------------
1548 dangling commit 7281251ddd2a61e38657c827739c57015671a6b3
1549 dangling commit 2706a059f258c6b245f298dc4ff2ccd30ec21a63
1550 dangling commit 13472b7c4b80851a1bc551779171dcb03655e9b5
1551 dangling blob 218761f9d90712d37a9c5e36f406f92202db07eb
1552 dangling commit bf093535a34a4d35731aa2bd90fe6b176302f14f
1553 dangling commit 8e4bec7f2ddaa268bef999853c25755452100f8e
1554 dangling tree d50bb86186bf27b681d25af89d3b5b68382e4085
1555 dangling tree b24c2473f1fd3d91352a624795be026d64c8841f
1557 -------------------------------------------------
1559 Dangling objects are not a problem. At worst they may take up a little
1560 extra disk space. They can sometimes provide a last-resort method of
1561 recovery lost work--see <<dangling-objects>> for details. However, if
1562 you want, you may remove them with gitlink:git-prune[1] or the --prune
1563 option to gitlink:git-gc[1]:
1565 -------------------------------------------------
1567 -------------------------------------------------
1569 This may be time-consuming. Unlike most other git operations (including
1570 git-gc when run without any options), it is not safe to prune while
1571 other git operations are in progress in the same repository.
1573 [[recovering-lost-changes]]
1574 Recovering lost changes
1575 ~~~~~~~~~~~~~~~~~~~~~~~
1581 Say you modify a branch with gitlink:git-reset[1] --hard, and then
1582 realize that the branch was the only reference you had to that point in
1585 Fortunately, git also keeps a log, called a "reflog", of all the
1586 previous values of each branch. So in this case you can still find the
1587 old history using, for example,
1589 -------------------------------------------------
1590 $ git log master@{1}
1591 -------------------------------------------------
1593 This lists the commits reachable from the previous version of the head.
1594 This syntax can be used to with any git command that accepts a commit,
1595 not just with git log. Some other examples:
1597 -------------------------------------------------
1598 $ git show master@{2} # See where the branch pointed 2,
1599 $ git show master@{3} # 3, ... changes ago.
1600 $ gitk master@{yesterday} # See where it pointed yesterday,
1601 $ gitk master@{"1 week ago"} # ... or last week
1602 $ git log --walk-reflogs master # show reflog entries for master
1603 -------------------------------------------------
1605 A separate reflog is kept for the HEAD, so
1607 -------------------------------------------------
1608 $ git show HEAD@{"1 week ago"}
1609 -------------------------------------------------
1611 will show what HEAD pointed to one week ago, not what the current branch
1612 pointed to one week ago. This allows you to see the history of what
1615 The reflogs are kept by default for 30 days, after which they may be
1616 pruned. See gitlink:git-reflog[1] and gitlink:git-gc[1] to learn
1617 how to control this pruning, and see the "SPECIFYING REVISIONS"
1618 section of gitlink:git-rev-parse[1] for details.
1620 Note that the reflog history is very different from normal git history.
1621 While normal history is shared by every repository that works on the
1622 same project, the reflog history is not shared: it tells you only about
1623 how the branches in your local repository have changed over time.
1625 [[dangling-object-recovery]]
1626 Examining dangling objects
1627 ^^^^^^^^^^^^^^^^^^^^^^^^^^
1629 In some situations the reflog may not be able to save you. For example,
1630 suppose you delete a branch, then realize you need the history it
1631 contained. The reflog is also deleted; however, if you have not yet
1632 pruned the repository, then you may still be able to find the lost
1633 commits in the dangling objects that git-fsck reports. See
1634 <<dangling-objects>> for the details.
1636 -------------------------------------------------
1638 dangling commit 7281251ddd2a61e38657c827739c57015671a6b3
1639 dangling commit 2706a059f258c6b245f298dc4ff2ccd30ec21a63
1640 dangling commit 13472b7c4b80851a1bc551779171dcb03655e9b5
1642 -------------------------------------------------
1645 one of those dangling commits with, for example,
1647 ------------------------------------------------
1648 $ gitk 7281251ddd --not --all
1649 ------------------------------------------------
1651 which does what it sounds like: it says that you want to see the commit
1652 history that is described by the dangling commit(s), but not the
1653 history that is described by all your existing branches and tags. Thus
1654 you get exactly the history reachable from that commit that is lost.
1655 (And notice that it might not be just one commit: we only report the
1656 "tip of the line" as being dangling, but there might be a whole deep
1657 and complex commit history that was dropped.)
1659 If you decide you want the history back, you can always create a new
1660 reference pointing to it, for example, a new branch:
1662 ------------------------------------------------
1663 $ git branch recovered-branch 7281251ddd
1664 ------------------------------------------------
1666 Other types of dangling objects (blobs and trees) are also possible, and
1667 dangling objects can arise in other situations.
1670 [[sharing-development]]
1671 Sharing development with others
1672 ===============================
1674 [[getting-updates-with-git-pull]]
1675 Getting updates with git pull
1676 -----------------------------
1678 After you clone a repository and make a few changes of your own, you
1679 may wish to check the original repository for updates and merge them
1682 We have already seen <<Updating-a-repository-with-git-fetch,how to
1683 keep remote tracking branches up to date>> with gitlink:git-fetch[1],
1684 and how to merge two branches. So you can merge in changes from the
1685 original repository's master branch with:
1687 -------------------------------------------------
1689 $ git merge origin/master
1690 -------------------------------------------------
1692 However, the gitlink:git-pull[1] command provides a way to do this in
1695 -------------------------------------------------
1696 $ git pull origin master
1697 -------------------------------------------------
1699 In fact, "origin" is normally the default repository to pull from,
1700 and the default branch is normally the HEAD of the remote repository,
1701 so often you can accomplish the above with just
1703 -------------------------------------------------
1705 -------------------------------------------------
1707 See the descriptions of the branch.<name>.remote and branch.<name>.merge
1708 options in gitlink:git-config[1] to learn how to control these defaults
1709 depending on the current branch. Also note that the --track option to
1710 gitlink:git-branch[1] and gitlink:git-checkout[1] can be used to
1711 automatically set the default remote branch to pull from at the time
1712 that a branch is created:
1714 -------------------------------------------------
1715 $ git checkout --track -b origin/maint maint
1716 -------------------------------------------------
1718 In addition to saving you keystrokes, "git pull" also helps you by
1719 producing a default commit message documenting the branch and
1720 repository that you pulled from.
1722 (But note that no such commit will be created in the case of a
1723 <<fast-forwards,fast forward>>; instead, your branch will just be
1724 updated to point to the latest commit from the upstream branch.)
1726 The git-pull command can also be given "." as the "remote" repository,
1727 in which case it just merges in a branch from the current repository; so
1730 -------------------------------------------------
1733 -------------------------------------------------
1735 are roughly equivalent. The former is actually very commonly used.
1737 [[submitting-patches]]
1738 Submitting patches to a project
1739 -------------------------------
1741 If you just have a few changes, the simplest way to submit them may
1742 just be to send them as patches in email:
1744 First, use gitlink:git-format-patch[1]; for example:
1746 -------------------------------------------------
1747 $ git format-patch origin
1748 -------------------------------------------------
1750 will produce a numbered series of files in the current directory, one
1751 for each patch in the current branch but not in origin/HEAD.
1753 You can then import these into your mail client and send them by
1754 hand. However, if you have a lot to send at once, you may prefer to
1755 use the gitlink:git-send-email[1] script to automate the process.
1756 Consult the mailing list for your project first to determine how they
1757 prefer such patches be handled.
1759 [[importing-patches]]
1760 Importing patches to a project
1761 ------------------------------
1763 Git also provides a tool called gitlink:git-am[1] (am stands for
1764 "apply mailbox"), for importing such an emailed series of patches.
1765 Just save all of the patch-containing messages, in order, into a
1766 single mailbox file, say "patches.mbox", then run
1768 -------------------------------------------------
1769 $ git am -3 patches.mbox
1770 -------------------------------------------------
1772 Git will apply each patch in order; if any conflicts are found, it
1773 will stop, and you can fix the conflicts as described in
1774 "<<resolving-a-merge,Resolving a merge>>". (The "-3" option tells
1775 git to perform a merge; if you would prefer it just to abort and
1776 leave your tree and index untouched, you may omit that option.)
1778 Once the index is updated with the results of the conflict
1779 resolution, instead of creating a new commit, just run
1781 -------------------------------------------------
1783 -------------------------------------------------
1785 and git will create the commit for you and continue applying the
1786 remaining patches from the mailbox.
1788 The final result will be a series of commits, one for each patch in
1789 the original mailbox, with authorship and commit log message each
1790 taken from the message containing each patch.
1792 [[setting-up-a-public-repository]]
1793 Setting up a public repository
1794 ------------------------------
1796 Another way to submit changes to a project is to simply tell the
1797 maintainer of that project to pull from your repository, exactly as
1798 you did in the section "<<getting-updates-with-git-pull, Getting
1799 updates with git pull>>".
1801 If you and maintainer both have accounts on the same machine, then
1802 then you can just pull changes from each other's repositories
1803 directly; note that all of the commands (gitlink:git-clone[1],
1804 git-fetch[1], git-pull[1], etc.) that accept a URL as an argument
1805 will also accept a local directory name; so, for example, you can
1808 -------------------------------------------------
1809 $ git clone /path/to/repository
1810 $ git pull /path/to/other/repository
1811 -------------------------------------------------
1813 If this sort of setup is inconvenient or impossible, another (more
1814 common) option is to set up a public repository on a public server.
1815 This also allows you to cleanly separate private work in progress
1816 from publicly visible work.
1818 You will continue to do your day-to-day work in your personal
1819 repository, but periodically "push" changes from your personal
1820 repository into your public repository, allowing other developers to
1821 pull from that repository. So the flow of changes, in a situation
1822 where there is one other developer with a public repository, looks
1826 your personal repo ------------------> your public repo
1829 | you pull | they pull
1833 their public repo <------------------- their repo
1835 Now, assume your personal repository is in the directory ~/proj. We
1836 first create a new clone of the repository:
1838 -------------------------------------------------
1839 $ git clone --bare ~/proj proj.git
1840 -------------------------------------------------
1842 The resulting directory proj.git contains a "bare" git repository--it is
1843 just the contents of the ".git" directory, without a checked-out copy of
1844 a working directory.
1846 Next, copy proj.git to the server where you plan to host the
1847 public repository. You can use scp, rsync, or whatever is most
1850 If somebody else maintains the public server, they may already have
1851 set up a git service for you, and you may skip to the section
1852 "<<pushing-changes-to-a-public-repository,Pushing changes to a public
1853 repository>>", below.
1855 Otherwise, the following sections explain how to export your newly
1856 created public repository:
1858 [[exporting-via-http]]
1859 Exporting a git repository via http
1860 -----------------------------------
1862 The git protocol gives better performance and reliability, but on a
1863 host with a web server set up, http exports may be simpler to set up.
1865 All you need to do is place the newly created bare git repository in
1866 a directory that is exported by the web server, and make some
1867 adjustments to give web clients some extra information they need:
1869 -------------------------------------------------
1870 $ mv proj.git /home/you/public_html/proj.git
1872 $ git --bare update-server-info
1873 $ chmod a+x hooks/post-update
1874 -------------------------------------------------
1876 (For an explanation of the last two lines, see
1877 gitlink:git-update-server-info[1], and the documentation
1878 link:hooks.txt[Hooks used by git].)
1880 Advertise the url of proj.git. Anybody else should then be able to
1881 clone or pull from that url, for example with a commandline like:
1883 -------------------------------------------------
1884 $ git clone http://yourserver.com/~you/proj.git
1885 -------------------------------------------------
1888 link:howto/setup-git-server-over-http.txt[setup-git-server-over-http]
1889 for a slightly more sophisticated setup using WebDAV which also
1890 allows pushing over http.)
1892 [[exporting-via-git]]
1893 Exporting a git repository via the git protocol
1894 -----------------------------------------------
1896 This is the preferred method.
1898 For now, we refer you to the gitlink:git-daemon[1] man page for
1899 instructions. (See especially the examples section.)
1901 [[pushing-changes-to-a-public-repository]]
1902 Pushing changes to a public repository
1903 --------------------------------------
1905 Note that the two techniques outline above (exporting via
1906 <<exporting-via-http,http>> or <<exporting-via-git,git>>) allow other
1907 maintainers to fetch your latest changes, but they do not allow write
1908 access, which you will need to update the public repository with the
1909 latest changes created in your private repository.
1911 The simplest way to do this is using gitlink:git-push[1] and ssh; to
1912 update the remote branch named "master" with the latest state of your
1913 branch named "master", run
1915 -------------------------------------------------
1916 $ git push ssh://yourserver.com/~you/proj.git master:master
1917 -------------------------------------------------
1921 -------------------------------------------------
1922 $ git push ssh://yourserver.com/~you/proj.git master
1923 -------------------------------------------------
1925 As with git-fetch, git-push will complain if this does not result in
1926 a <<fast-forwards,fast forward>>. Normally this is a sign of
1927 something wrong. However, if you are sure you know what you're
1928 doing, you may force git-push to perform the update anyway by
1929 proceeding the branch name by a plus sign:
1931 -------------------------------------------------
1932 $ git push ssh://yourserver.com/~you/proj.git +master
1933 -------------------------------------------------
1935 As with git-fetch, you may also set up configuration options to
1936 save typing; so, for example, after
1938 -------------------------------------------------
1939 $ cat >>.git/config <<EOF
1940 [remote "public-repo"]
1941 url = ssh://yourserver.com/~you/proj.git
1943 -------------------------------------------------
1945 you should be able to perform the above push with just
1947 -------------------------------------------------
1948 $ git push public-repo master
1949 -------------------------------------------------
1951 See the explanations of the remote.<name>.url, branch.<name>.remote,
1952 and remote.<name>.push options in gitlink:git-config[1] for
1955 [[setting-up-a-shared-repository]]
1956 Setting up a shared repository
1957 ------------------------------
1959 Another way to collaborate is by using a model similar to that
1960 commonly used in CVS, where several developers with special rights
1961 all push to and pull from a single shared repository. See
1962 link:cvs-migration.txt[git for CVS users] for instructions on how to
1965 [[setting-up-gitweb]]
1966 Allow web browsing of a repository
1967 ----------------------------------
1969 The gitweb cgi script provides users an easy way to browse your
1970 project's files and history without having to install git; see the file
1971 gitweb/INSTALL in the git source tree for instructions on setting it up.
1973 [[sharing-development-examples]]
1977 TODO: topic branches, typical roles as in everyday.txt, ?
1980 [[cleaning-up-history]]
1981 Rewriting history and maintaining patch series
1982 ==============================================
1984 Normally commits are only added to a project, never taken away or
1985 replaced. Git is designed with this assumption, and violating it will
1986 cause git's merge machinery (for example) to do the wrong thing.
1988 However, there is a situation in which it can be useful to violate this
1992 Creating the perfect patch series
1993 ---------------------------------
1995 Suppose you are a contributor to a large project, and you want to add a
1996 complicated feature, and to present it to the other developers in a way
1997 that makes it easy for them to read your changes, verify that they are
1998 correct, and understand why you made each change.
2000 If you present all of your changes as a single patch (or commit), they
2001 may find that it is too much to digest all at once.
2003 If you present them with the entire history of your work, complete with
2004 mistakes, corrections, and dead ends, they may be overwhelmed.
2006 So the ideal is usually to produce a series of patches such that:
2008 1. Each patch can be applied in order.
2010 2. Each patch includes a single logical change, together with a
2011 message explaining the change.
2013 3. No patch introduces a regression: after applying any initial
2014 part of the series, the resulting project still compiles and
2015 works, and has no bugs that it didn't have before.
2017 4. The complete series produces the same end result as your own
2018 (probably much messier!) development process did.
2020 We will introduce some tools that can help you do this, explain how to
2021 use them, and then explain some of the problems that can arise because
2022 you are rewriting history.
2024 [[using-git-rebase]]
2025 Keeping a patch series up to date using git-rebase
2026 --------------------------------------------------
2028 Suppose that you create a branch "mywork" on a remote-tracking branch
2029 "origin", and create some commits on top of it:
2031 -------------------------------------------------
2032 $ git checkout -b mywork origin
2038 -------------------------------------------------
2040 You have performed no merges into mywork, so it is just a simple linear
2041 sequence of patches on top of "origin":
2043 ................................................
2047 ................................................
2049 Some more interesting work has been done in the upstream project, and
2050 "origin" has advanced:
2052 ................................................
2053 o--o--O--o--o--o <-- origin
2056 ................................................
2058 At this point, you could use "pull" to merge your changes back in;
2059 the result would create a new merge commit, like this:
2061 ................................................
2062 o--o--O--o--o--o <-- origin
2064 a--b--c--m <-- mywork
2065 ................................................
2067 However, if you prefer to keep the history in mywork a simple series of
2068 commits without any merges, you may instead choose to use
2069 gitlink:git-rebase[1]:
2071 -------------------------------------------------
2072 $ git checkout mywork
2074 -------------------------------------------------
2076 This will remove each of your commits from mywork, temporarily saving
2077 them as patches (in a directory named ".dotest"), update mywork to
2078 point at the latest version of origin, then apply each of the saved
2079 patches to the new mywork. The result will look like:
2082 ................................................
2083 o--o--O--o--o--o <-- origin
2085 a'--b'--c' <-- mywork
2086 ................................................
2088 In the process, it may discover conflicts. In that case it will stop
2089 and allow you to fix the conflicts; after fixing conflicts, use "git
2090 add" to update the index with those contents, and then, instead of
2091 running git-commit, just run
2093 -------------------------------------------------
2094 $ git rebase --continue
2095 -------------------------------------------------
2097 and git will continue applying the rest of the patches.
2099 At any point you may use the --abort option to abort this process and
2100 return mywork to the state it had before you started the rebase:
2102 -------------------------------------------------
2103 $ git rebase --abort
2104 -------------------------------------------------
2106 [[modifying-one-commit]]
2107 Modifying a single commit
2108 -------------------------
2110 We saw in <<fixing-a-mistake-by-editing-history>> that you can replace the
2111 most recent commit using
2113 -------------------------------------------------
2114 $ git commit --amend
2115 -------------------------------------------------
2117 which will replace the old commit by a new commit incorporating your
2118 changes, giving you a chance to edit the old commit message first.
2120 You can also use a combination of this and gitlink:git-rebase[1] to edit
2121 commits further back in your history. First, tag the problematic commit with
2123 -------------------------------------------------
2124 $ git tag bad mywork~5
2125 -------------------------------------------------
2127 (Either gitk or git-log may be useful for finding the commit.)
2129 Then check out that commit, edit it, and rebase the rest of the series
2130 on top of it (note that we could check out the commit on a temporary
2131 branch, but instead we're using a <<detached-head,detached head>>):
2133 -------------------------------------------------
2135 $ # make changes here and update the index
2136 $ git commit --amend
2137 $ git rebase --onto HEAD bad mywork
2138 -------------------------------------------------
2140 When you're done, you'll be left with mywork checked out, with the top
2141 patches on mywork reapplied on top of your modified commit. You can
2144 -------------------------------------------------
2146 -------------------------------------------------
2148 Note that the immutable nature of git history means that you haven't really
2149 "modified" existing commits; instead, you have replaced the old commits with
2150 new commits having new object names.
2152 [[reordering-patch-series]]
2153 Reordering or selecting from a patch series
2154 -------------------------------------------
2156 Given one existing commit, the gitlink:git-cherry-pick[1] command
2157 allows you to apply the change introduced by that commit and create a
2158 new commit that records it. So, for example, if "mywork" points to a
2159 series of patches on top of "origin", you might do something like:
2161 -------------------------------------------------
2162 $ git checkout -b mywork-new origin
2163 $ gitk origin..mywork &
2164 -------------------------------------------------
2166 And browse through the list of patches in the mywork branch using gitk,
2167 applying them (possibly in a different order) to mywork-new using
2168 cherry-pick, and possibly modifying them as you go using commit
2171 Another technique is to use git-format-patch to create a series of
2172 patches, then reset the state to before the patches:
2174 -------------------------------------------------
2175 $ git format-patch origin
2176 $ git reset --hard origin
2177 -------------------------------------------------
2179 Then modify, reorder, or eliminate patches as preferred before applying
2180 them again with gitlink:git-am[1].
2182 [[patch-series-tools]]
2186 There are numerous other tools, such as stgit, which exist for the
2187 purpose of maintaining a patch series. These are outside of the scope of
2190 [[problems-with-rewriting-history]]
2191 Problems with rewriting history
2192 -------------------------------
2194 The primary problem with rewriting the history of a branch has to do
2195 with merging. Suppose somebody fetches your branch and merges it into
2196 their branch, with a result something like this:
2198 ................................................
2199 o--o--O--o--o--o <-- origin
2201 t--t--t--m <-- their branch:
2202 ................................................
2204 Then suppose you modify the last three commits:
2206 ................................................
2207 o--o--o <-- new head of origin
2209 o--o--O--o--o--o <-- old head of origin
2210 ................................................
2212 If we examined all this history together in one repository, it will
2215 ................................................
2216 o--o--o <-- new head of origin
2218 o--o--O--o--o--o <-- old head of origin
2220 t--t--t--m <-- their branch:
2221 ................................................
2223 Git has no way of knowing that the new head is an updated version of
2224 the old head; it treats this situation exactly the same as it would if
2225 two developers had independently done the work on the old and new heads
2226 in parallel. At this point, if someone attempts to merge the new head
2227 in to their branch, git will attempt to merge together the two (old and
2228 new) lines of development, instead of trying to replace the old by the
2229 new. The results are likely to be unexpected.
2231 You may still choose to publish branches whose history is rewritten,
2232 and it may be useful for others to be able to fetch those branches in
2233 order to examine or test them, but they should not attempt to pull such
2234 branches into their own work.
2236 For true distributed development that supports proper merging,
2237 published branches should never be rewritten.
2239 [[advanced-branch-management]]
2240 Advanced branch management
2241 ==========================
2243 [[fetching-individual-branches]]
2244 Fetching individual branches
2245 ----------------------------
2247 Instead of using gitlink:git-remote[1], you can also choose just
2248 to update one branch at a time, and to store it locally under an
2251 -------------------------------------------------
2252 $ git fetch origin todo:my-todo-work
2253 -------------------------------------------------
2255 The first argument, "origin", just tells git to fetch from the
2256 repository you originally cloned from. The second argument tells git
2257 to fetch the branch named "todo" from the remote repository, and to
2258 store it locally under the name refs/heads/my-todo-work.
2260 You can also fetch branches from other repositories; so
2262 -------------------------------------------------
2263 $ git fetch git://example.com/proj.git master:example-master
2264 -------------------------------------------------
2266 will create a new branch named "example-master" and store in it the
2267 branch named "master" from the repository at the given URL. If you
2268 already have a branch named example-master, it will attempt to
2269 <<fast-forwards,fast-forward>> to the commit given by example.com's
2270 master branch. In more detail:
2272 [[fetch-fast-forwards]]
2273 git fetch and fast-forwards
2274 ---------------------------
2276 In the previous example, when updating an existing branch, "git
2277 fetch" checks to make sure that the most recent commit on the remote
2278 branch is a descendant of the most recent commit on your copy of the
2279 branch before updating your copy of the branch to point at the new
2280 commit. Git calls this process a <<fast-forwards,fast forward>>.
2282 A fast forward looks something like this:
2284 ................................................
2285 o--o--o--o <-- old head of the branch
2287 o--o--o <-- new head of the branch
2288 ................................................
2291 In some cases it is possible that the new head will *not* actually be
2292 a descendant of the old head. For example, the developer may have
2293 realized she made a serious mistake, and decided to backtrack,
2294 resulting in a situation like:
2296 ................................................
2297 o--o--o--o--a--b <-- old head of the branch
2299 o--o--o <-- new head of the branch
2300 ................................................
2302 In this case, "git fetch" will fail, and print out a warning.
2304 In that case, you can still force git to update to the new head, as
2305 described in the following section. However, note that in the
2306 situation above this may mean losing the commits labeled "a" and "b",
2307 unless you've already created a reference of your own pointing to
2311 Forcing git fetch to do non-fast-forward updates
2312 ------------------------------------------------
2314 If git fetch fails because the new head of a branch is not a
2315 descendant of the old head, you may force the update with:
2317 -------------------------------------------------
2318 $ git fetch git://example.com/proj.git +master:refs/remotes/example/master
2319 -------------------------------------------------
2321 Note the addition of the "+" sign. Alternatively, you can use the "-f"
2322 flag to force updates of all the fetched branches, as in:
2324 -------------------------------------------------
2325 $ git fetch -f origin
2326 -------------------------------------------------
2328 Be aware that commits that the old version of example/master pointed at
2329 may be lost, as we saw in the previous section.
2331 [[remote-branch-configuration]]
2332 Configuring remote branches
2333 ---------------------------
2335 We saw above that "origin" is just a shortcut to refer to the
2336 repository that you originally cloned from. This information is
2337 stored in git configuration variables, which you can see using
2338 gitlink:git-config[1]:
2340 -------------------------------------------------
2342 core.repositoryformatversion=0
2344 core.logallrefupdates=true
2345 remote.origin.url=git://git.kernel.org/pub/scm/git/git.git
2346 remote.origin.fetch=+refs/heads/*:refs/remotes/origin/*
2347 branch.master.remote=origin
2348 branch.master.merge=refs/heads/master
2349 -------------------------------------------------
2351 If there are other repositories that you also use frequently, you can
2352 create similar configuration options to save typing; for example,
2355 -------------------------------------------------
2356 $ git config remote.example.url git://example.com/proj.git
2357 -------------------------------------------------
2359 then the following two commands will do the same thing:
2361 -------------------------------------------------
2362 $ git fetch git://example.com/proj.git master:refs/remotes/example/master
2363 $ git fetch example master:refs/remotes/example/master
2364 -------------------------------------------------
2366 Even better, if you add one more option:
2368 -------------------------------------------------
2369 $ git config remote.example.fetch master:refs/remotes/example/master
2370 -------------------------------------------------
2372 then the following commands will all do the same thing:
2374 -------------------------------------------------
2375 $ git fetch git://example.com/proj.git master:refs/remotes/example/master
2376 $ git fetch example master:refs/remotes/example/master
2378 -------------------------------------------------
2380 You can also add a "+" to force the update each time:
2382 -------------------------------------------------
2383 $ git config remote.example.fetch +master:ref/remotes/example/master
2384 -------------------------------------------------
2386 Don't do this unless you're sure you won't mind "git fetch" possibly
2387 throwing away commits on mybranch.
2389 Also note that all of the above configuration can be performed by
2390 directly editing the file .git/config instead of using
2391 gitlink:git-config[1].
2393 See gitlink:git-config[1] for more details on the configuration
2394 options mentioned above.
2401 Git depends on two fundamental abstractions: the "object database", and
2402 the "current directory cache" aka "index".
2404 [[the-object-database]]
2408 The object database is literally just a content-addressable collection
2409 of objects. All objects are named by their content, which is
2410 approximated by the SHA1 hash of the object itself. Objects may refer
2411 to other objects (by referencing their SHA1 hash), and so you can
2412 build up a hierarchy of objects.
2414 All objects have a statically determined "type" which is
2415 determined at object creation time, and which identifies the format of
2416 the object (i.e. how it is used, and how it can refer to other
2417 objects). There are currently four different object types: "blob",
2418 "tree", "commit", and "tag".
2420 A <<def_blob_object,"blob" object>> cannot refer to any other object,
2421 and is, as the name implies, a pure storage object containing some
2422 user data. It is used to actually store the file data, i.e. a blob
2423 object is associated with some particular version of some file.
2425 A <<def_tree_object,"tree" object>> is an object that ties one or more
2426 "blob" objects into a directory structure. In addition, a tree object
2427 can refer to other tree objects, thus creating a directory hierarchy.
2429 A <<def_commit_object,"commit" object>> ties such directory hierarchies
2430 together into a <<def_DAG,directed acyclic graph>> of revisions - each
2431 "commit" is associated with exactly one tree (the directory hierarchy at
2432 the time of the commit). In addition, a "commit" refers to one or more
2433 "parent" commit objects that describe the history of how we arrived at
2434 that directory hierarchy.
2436 As a special case, a commit object with no parents is called the "root"
2437 commit, and is the point of an initial project commit. Each project
2438 must have at least one root, and while you can tie several different
2439 root objects together into one project by creating a commit object which
2440 has two or more separate roots as its ultimate parents, that's probably
2441 just going to confuse people. So aim for the notion of "one root object
2442 per project", even if git itself does not enforce that.
2444 A <<def_tag_object,"tag" object>> symbolically identifies and can be
2445 used to sign other objects. It contains the identifier and type of
2446 another object, a symbolic name (of course!) and, optionally, a
2449 Regardless of object type, all objects share the following
2450 characteristics: they are all deflated with zlib, and have a header
2451 that not only specifies their type, but also provides size information
2452 about the data in the object. It's worth noting that the SHA1 hash
2453 that is used to name the object is the hash of the original data
2454 plus this header, so `sha1sum` 'file' does not match the object name
2456 (Historical note: in the dawn of the age of git the hash
2457 was the sha1 of the 'compressed' object.)
2459 As a result, the general consistency of an object can always be tested
2460 independently of the contents or the type of the object: all objects can
2461 be validated by verifying that (a) their hashes match the content of the
2462 file and (b) the object successfully inflates to a stream of bytes that
2463 forms a sequence of <ascii type without space> + <space> + <ascii decimal
2464 size> + <byte\0> + <binary object data>.
2466 The structured objects can further have their structure and
2467 connectivity to other objects verified. This is generally done with
2468 the `git-fsck` program, which generates a full dependency graph
2469 of all objects, and verifies their internal consistency (in addition
2470 to just verifying their superficial consistency through the hash).
2472 The object types in some more detail:
2478 A "blob" object is nothing but a binary blob of data, and doesn't
2479 refer to anything else. There is no signature or any other
2480 verification of the data, so while the object is consistent (it 'is'
2481 indexed by its sha1 hash, so the data itself is certainly correct), it
2482 has absolutely no other attributes. No name associations, no
2483 permissions. It is purely a blob of data (i.e. normally "file
2486 In particular, since the blob is entirely defined by its data, if two
2487 files in a directory tree (or in multiple different versions of the
2488 repository) have the same contents, they will share the same blob
2489 object. The object is totally independent of its location in the
2490 directory tree, and renaming a file does not change the object that
2491 file is associated with in any way.
2493 A blob is typically created when gitlink:git-update-index[1]
2494 is run, and its data can be accessed by gitlink:git-cat-file[1].
2500 The next hierarchical object type is the "tree" object. A tree object
2501 is a list of mode/name/blob data, sorted by name. Alternatively, the
2502 mode data may specify a directory mode, in which case instead of
2503 naming a blob, that name is associated with another TREE object.
2505 Like the "blob" object, a tree object is uniquely determined by the
2506 set contents, and so two separate but identical trees will always
2507 share the exact same object. This is true at all levels, i.e. it's
2508 true for a "leaf" tree (which does not refer to any other trees, only
2509 blobs) as well as for a whole subdirectory.
2511 For that reason a "tree" object is just a pure data abstraction: it
2512 has no history, no signatures, no verification of validity, except
2513 that since the contents are again protected by the hash itself, we can
2514 trust that the tree is immutable and its contents never change.
2516 So you can trust the contents of a tree to be valid, the same way you
2517 can trust the contents of a blob, but you don't know where those
2518 contents 'came' from.
2520 Side note on trees: since a "tree" object is a sorted list of
2521 "filename+content", you can create a diff between two trees without
2522 actually having to unpack two trees. Just ignore all common parts,
2523 and your diff will look right. In other words, you can effectively
2524 (and efficiently) tell the difference between any two random trees by
2525 O(n) where "n" is the size of the difference, rather than the size of
2528 Side note 2 on trees: since the name of a "blob" depends entirely and
2529 exclusively on its contents (i.e. there are no names or permissions
2530 involved), you can see trivial renames or permission changes by
2531 noticing that the blob stayed the same. However, renames with data
2532 changes need a smarter "diff" implementation.
2534 A tree is created with gitlink:git-write-tree[1] and
2535 its data can be accessed by gitlink:git-ls-tree[1].
2536 Two trees can be compared with gitlink:git-diff-tree[1].
2542 The "commit" object is an object that introduces the notion of
2543 history into the picture. In contrast to the other objects, it
2544 doesn't just describe the physical state of a tree, it describes how
2545 we got there, and why.
2547 A "commit" is defined by the tree-object that it results in, the
2548 parent commits (zero, one or more) that led up to that point, and a
2549 comment on what happened. Again, a commit is not trusted per se:
2550 the contents are well-defined and "safe" due to the cryptographically
2551 strong signatures at all levels, but there is no reason to believe
2552 that the tree is "good" or that the merge information makes sense.
2553 The parents do not have to actually have any relationship with the
2554 result, for example.
2556 Note on commits: unlike some SCM's, commits do not contain
2557 rename information or file mode change information. All of that is
2558 implicit in the trees involved (the result tree, and the result trees
2559 of the parents), and describing that makes no sense in this idiotic
2562 A commit is created with gitlink:git-commit-tree[1] and
2563 its data can be accessed by gitlink:git-cat-file[1].
2569 An aside on the notion of "trust". Trust is really outside the scope
2570 of "git", but it's worth noting a few things. First off, since
2571 everything is hashed with SHA1, you 'can' trust that an object is
2572 intact and has not been messed with by external sources. So the name
2573 of an object uniquely identifies a known state - just not a state that
2574 you may want to trust.
2576 Furthermore, since the SHA1 signature of a commit refers to the
2577 SHA1 signatures of the tree it is associated with and the signatures
2578 of the parent, a single named commit specifies uniquely a whole set
2579 of history, with full contents. You can't later fake any step of the
2580 way once you have the name of a commit.
2582 So to introduce some real trust in the system, the only thing you need
2583 to do is to digitally sign just 'one' special note, which includes the
2584 name of a top-level commit. Your digital signature shows others
2585 that you trust that commit, and the immutability of the history of
2586 commits tells others that they can trust the whole history.
2588 In other words, you can easily validate a whole archive by just
2589 sending out a single email that tells the people the name (SHA1 hash)
2590 of the top commit, and digitally sign that email using something
2593 To assist in this, git also provides the tag object...
2599 Git provides the "tag" object to simplify creating, managing and
2600 exchanging symbolic and signed tokens. The "tag" object at its
2601 simplest simply symbolically identifies another object by containing
2602 the sha1, type and symbolic name.
2604 However it can optionally contain additional signature information
2605 (which git doesn't care about as long as there's less than 8k of
2606 it). This can then be verified externally to git.
2608 Note that despite the tag features, "git" itself only handles content
2609 integrity; the trust framework (and signature provision and
2610 verification) has to come from outside.
2612 A tag is created with gitlink:git-mktag[1],
2613 its data can be accessed by gitlink:git-cat-file[1],
2614 and the signature can be verified by
2615 gitlink:git-verify-tag[1].
2619 The "index" aka "Current Directory Cache"
2620 -----------------------------------------
2622 The index is a simple binary file, which contains an efficient
2623 representation of the contents of a virtual directory. It
2624 does so by a simple array that associates a set of names, dates,
2625 permissions and content (aka "blob") objects together. The cache is
2626 always kept ordered by name, and names are unique (with a few very
2627 specific rules) at any point in time, but the cache has no long-term
2628 meaning, and can be partially updated at any time.
2630 In particular, the index certainly does not need to be consistent with
2631 the current directory contents (in fact, most operations will depend on
2632 different ways to make the index 'not' be consistent with the directory
2633 hierarchy), but it has three very important attributes:
2635 '(a) it can re-generate the full state it caches (not just the
2636 directory structure: it contains pointers to the "blob" objects so
2637 that it can regenerate the data too)'
2639 As a special case, there is a clear and unambiguous one-way mapping
2640 from a current directory cache to a "tree object", which can be
2641 efficiently created from just the current directory cache without
2642 actually looking at any other data. So a directory cache at any one
2643 time uniquely specifies one and only one "tree" object (but has
2644 additional data to make it easy to match up that tree object with what
2645 has happened in the directory)
2647 '(b) it has efficient methods for finding inconsistencies between that
2648 cached state ("tree object waiting to be instantiated") and the
2651 '(c) it can additionally efficiently represent information about merge
2652 conflicts between different tree objects, allowing each pathname to be
2653 associated with sufficient information about the trees involved that
2654 you can create a three-way merge between them.'
2656 Those are the ONLY three things that the directory cache does. It's a
2657 cache, and the normal operation is to re-generate it completely from a
2658 known tree object, or update/compare it with a live tree that is being
2659 developed. If you blow the directory cache away entirely, you generally
2660 haven't lost any information as long as you have the name of the tree
2663 At the same time, the index is at the same time also the
2664 staging area for creating new trees, and creating a new tree always
2665 involves a controlled modification of the index file. In particular,
2666 the index file can have the representation of an intermediate tree that
2667 has not yet been instantiated. So the index can be thought of as a
2668 write-back cache, which can contain dirty information that has not yet
2669 been written back to the backing store.
2677 Generally, all "git" operations work on the index file. Some operations
2678 work *purely* on the index file (showing the current state of the
2679 index), but most operations move data to and from the index file. Either
2680 from the database or from the working directory. Thus there are four
2683 [[working-directory-to-index]]
2684 working directory -> index
2685 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2687 You update the index with information from the working directory with
2688 the gitlink:git-update-index[1] command. You
2689 generally update the index information by just specifying the filename
2690 you want to update, like so:
2692 -------------------------------------------------
2693 $ git-update-index filename
2694 -------------------------------------------------
2696 but to avoid common mistakes with filename globbing etc, the command
2697 will not normally add totally new entries or remove old entries,
2698 i.e. it will normally just update existing cache entries.
2700 To tell git that yes, you really do realize that certain files no
2701 longer exist, or that new files should be added, you
2702 should use the `--remove` and `--add` flags respectively.
2704 NOTE! A `--remove` flag does 'not' mean that subsequent filenames will
2705 necessarily be removed: if the files still exist in your directory
2706 structure, the index will be updated with their new status, not
2707 removed. The only thing `--remove` means is that update-cache will be
2708 considering a removed file to be a valid thing, and if the file really
2709 does not exist any more, it will update the index accordingly.
2711 As a special case, you can also do `git-update-index --refresh`, which
2712 will refresh the "stat" information of each index to match the current
2713 stat information. It will 'not' update the object status itself, and
2714 it will only update the fields that are used to quickly test whether
2715 an object still matches its old backing store object.
2717 [[index-to-object-database]]
2718 index -> object database
2719 ~~~~~~~~~~~~~~~~~~~~~~~~
2721 You write your current index file to a "tree" object with the program
2723 -------------------------------------------------
2725 -------------------------------------------------
2727 that doesn't come with any options - it will just write out the
2728 current index into the set of tree objects that describe that state,
2729 and it will return the name of the resulting top-level tree. You can
2730 use that tree to re-generate the index at any time by going in the
2733 [[object-database-to-index]]
2734 object database -> index
2735 ~~~~~~~~~~~~~~~~~~~~~~~~
2737 You read a "tree" file from the object database, and use that to
2738 populate (and overwrite - don't do this if your index contains any
2739 unsaved state that you might want to restore later!) your current
2740 index. Normal operation is just
2742 -------------------------------------------------
2743 $ git-read-tree <sha1 of tree>
2744 -------------------------------------------------
2746 and your index file will now be equivalent to the tree that you saved
2747 earlier. However, that is only your 'index' file: your working
2748 directory contents have not been modified.
2750 [[index-to-working-directory]]
2751 index -> working directory
2752 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2754 You update your working directory from the index by "checking out"
2755 files. This is not a very common operation, since normally you'd just
2756 keep your files updated, and rather than write to your working
2757 directory, you'd tell the index files about the changes in your
2758 working directory (i.e. `git-update-index`).
2760 However, if you decide to jump to a new version, or check out somebody
2761 else's version, or just restore a previous tree, you'd populate your
2762 index file with read-tree, and then you need to check out the result
2765 -------------------------------------------------
2766 $ git-checkout-index filename
2767 -------------------------------------------------
2769 or, if you want to check out all of the index, use `-a`.
2771 NOTE! git-checkout-index normally refuses to overwrite old files, so
2772 if you have an old version of the tree already checked out, you will
2773 need to use the "-f" flag ('before' the "-a" flag or the filename) to
2774 'force' the checkout.
2777 Finally, there are a few odds and ends which are not purely moving
2778 from one representation to the other:
2780 [[tying-it-all-together]]
2781 Tying it all together
2782 ~~~~~~~~~~~~~~~~~~~~~
2784 To commit a tree you have instantiated with "git-write-tree", you'd
2785 create a "commit" object that refers to that tree and the history
2786 behind it - most notably the "parent" commits that preceded it in
2789 Normally a "commit" has one parent: the previous state of the tree
2790 before a certain change was made. However, sometimes it can have two
2791 or more parent commits, in which case we call it a "merge", due to the
2792 fact that such a commit brings together ("merges") two or more
2793 previous states represented by other commits.
2795 In other words, while a "tree" represents a particular directory state
2796 of a working directory, a "commit" represents that state in "time",
2797 and explains how we got there.
2799 You create a commit object by giving it the tree that describes the
2800 state at the time of the commit, and a list of parents:
2802 -------------------------------------------------
2803 $ git-commit-tree <tree> -p <parent> [-p <parent2> ..]
2804 -------------------------------------------------
2806 and then giving the reason for the commit on stdin (either through
2807 redirection from a pipe or file, or by just typing it at the tty).
2809 git-commit-tree will return the name of the object that represents
2810 that commit, and you should save it away for later use. Normally,
2811 you'd commit a new `HEAD` state, and while git doesn't care where you
2812 save the note about that state, in practice we tend to just write the
2813 result to the file pointed at by `.git/HEAD`, so that we can always see
2814 what the last committed state was.
2816 Here is an ASCII art by Jon Loeliger that illustrates how
2817 various pieces fit together.
2845 checkout-index -u | | checkout-index
2856 [[examining-the-data]]
2860 You can examine the data represented in the object database and the
2861 index with various helper tools. For every object, you can use
2862 gitlink:git-cat-file[1] to examine details about the
2865 -------------------------------------------------
2866 $ git-cat-file -t <objectname>
2867 -------------------------------------------------
2869 shows the type of the object, and once you have the type (which is
2870 usually implicit in where you find the object), you can use
2872 -------------------------------------------------
2873 $ git-cat-file blob|tree|commit|tag <objectname>
2874 -------------------------------------------------
2876 to show its contents. NOTE! Trees have binary content, and as a result
2877 there is a special helper for showing that content, called
2878 `git-ls-tree`, which turns the binary content into a more easily
2881 It's especially instructive to look at "commit" objects, since those
2882 tend to be small and fairly self-explanatory. In particular, if you
2883 follow the convention of having the top commit name in `.git/HEAD`,
2886 -------------------------------------------------
2887 $ git-cat-file commit HEAD
2888 -------------------------------------------------
2890 to see what the top commit was.
2892 [[merging-multiple-trees]]
2893 Merging multiple trees
2894 ----------------------
2896 Git helps you do a three-way merge, which you can expand to n-way by
2897 repeating the merge procedure arbitrary times until you finally
2898 "commit" the state. The normal situation is that you'd only do one
2899 three-way merge (two parents), and commit it, but if you like to, you
2900 can do multiple parents in one go.
2902 To do a three-way merge, you need the two sets of "commit" objects
2903 that you want to merge, use those to find the closest common parent (a
2904 third "commit" object), and then use those commit objects to find the
2905 state of the directory ("tree" object) at these points.
2907 To get the "base" for the merge, you first look up the common parent
2910 -------------------------------------------------
2911 $ git-merge-base <commit1> <commit2>
2912 -------------------------------------------------
2914 which will return you the commit they are both based on. You should
2915 now look up the "tree" objects of those commits, which you can easily
2916 do with (for example)
2918 -------------------------------------------------
2919 $ git-cat-file commit <commitname> | head -1
2920 -------------------------------------------------
2922 since the tree object information is always the first line in a commit
2925 Once you know the three trees you are going to merge (the one "original"
2926 tree, aka the common tree, and the two "result" trees, aka the branches
2927 you want to merge), you do a "merge" read into the index. This will
2928 complain if it has to throw away your old index contents, so you should
2929 make sure that you've committed those - in fact you would normally
2930 always do a merge against your last commit (which should thus match what
2931 you have in your current index anyway).
2935 -------------------------------------------------
2936 $ git-read-tree -m -u <origtree> <yourtree> <targettree>
2937 -------------------------------------------------
2939 which will do all trivial merge operations for you directly in the
2940 index file, and you can just write the result out with
2944 [[merging-multiple-trees-2]]
2945 Merging multiple trees, continued
2946 ---------------------------------
2948 Sadly, many merges aren't trivial. If there are files that have
2949 been added.moved or removed, or if both branches have modified the
2950 same file, you will be left with an index tree that contains "merge
2951 entries" in it. Such an index tree can 'NOT' be written out to a tree
2952 object, and you will have to resolve any such merge clashes using
2953 other tools before you can write out the result.
2955 You can examine such index state with `git-ls-files --unmerged`
2956 command. An example:
2958 ------------------------------------------------
2959 $ git-read-tree -m $orig HEAD $target
2960 $ git-ls-files --unmerged
2961 100644 263414f423d0e4d70dae8fe53fa34614ff3e2860 1 hello.c
2962 100644 06fa6a24256dc7e560efa5687fa84b51f0263c3a 2 hello.c
2963 100644 cc44c73eb783565da5831b4d820c962954019b69 3 hello.c
2964 ------------------------------------------------
2966 Each line of the `git-ls-files --unmerged` output begins with
2967 the blob mode bits, blob SHA1, 'stage number', and the
2968 filename. The 'stage number' is git's way to say which tree it
2969 came from: stage 1 corresponds to `$orig` tree, stage 2 `HEAD`
2970 tree, and stage3 `$target` tree.
2972 Earlier we said that trivial merges are done inside
2973 `git-read-tree -m`. For example, if the file did not change
2974 from `$orig` to `HEAD` nor `$target`, or if the file changed
2975 from `$orig` to `HEAD` and `$orig` to `$target` the same way,
2976 obviously the final outcome is what is in `HEAD`. What the
2977 above example shows is that file `hello.c` was changed from
2978 `$orig` to `HEAD` and `$orig` to `$target` in a different way.
2979 You could resolve this by running your favorite 3-way merge
2980 program, e.g. `diff3`, `merge`, or git's own merge-file, on
2981 the blob objects from these three stages yourself, like this:
2983 ------------------------------------------------
2984 $ git-cat-file blob 263414f... >hello.c~1
2985 $ git-cat-file blob 06fa6a2... >hello.c~2
2986 $ git-cat-file blob cc44c73... >hello.c~3
2987 $ git merge-file hello.c~2 hello.c~1 hello.c~3
2988 ------------------------------------------------
2990 This would leave the merge result in `hello.c~2` file, along
2991 with conflict markers if there are conflicts. After verifying
2992 the merge result makes sense, you can tell git what the final
2993 merge result for this file is by:
2995 -------------------------------------------------
2996 $ mv -f hello.c~2 hello.c
2997 $ git-update-index hello.c
2998 -------------------------------------------------
3000 When a path is in unmerged state, running `git-update-index` for
3001 that path tells git to mark the path resolved.
3003 The above is the description of a git merge at the lowest level,
3004 to help you understand what conceptually happens under the hood.
3005 In practice, nobody, not even git itself, uses three `git-cat-file`
3006 for this. There is `git-merge-index` program that extracts the
3007 stages to temporary files and calls a "merge" script on it:
3009 -------------------------------------------------
3010 $ git-merge-index git-merge-one-file hello.c
3011 -------------------------------------------------
3013 and that is what higher level `git merge -s resolve` is implemented with.
3016 How git stores objects efficiently: pack files
3017 ----------------------------------------------
3019 We've seen how git stores each object in a file named after the
3022 Unfortunately this system becomes inefficient once a project has a
3023 lot of objects. Try this on an old project:
3025 ------------------------------------------------
3027 6930 objects, 47620 kilobytes
3028 ------------------------------------------------
3030 The first number is the number of objects which are kept in
3031 individual files. The second is the amount of space taken up by
3032 those "loose" objects.
3034 You can save space and make git faster by moving these loose objects in
3035 to a "pack file", which stores a group of objects in an efficient
3036 compressed format; the details of how pack files are formatted can be
3037 found in link:technical/pack-format.txt[technical/pack-format.txt].
3039 To put the loose objects into a pack, just run git repack:
3041 ------------------------------------------------
3044 Done counting 6020 objects.
3045 Deltifying 6020 objects.
3046 100% (6020/6020) done
3047 Writing 6020 objects.
3048 100% (6020/6020) done
3049 Total 6020, written 6020 (delta 4070), reused 0 (delta 0)
3050 Pack pack-3e54ad29d5b2e05838c75df582c65257b8d08e1c created.
3051 ------------------------------------------------
3055 ------------------------------------------------
3057 ------------------------------------------------
3059 to remove any of the "loose" objects that are now contained in the
3060 pack. This will also remove any unreferenced objects (which may be
3061 created when, for example, you use "git reset" to remove a commit).
3062 You can verify that the loose objects are gone by looking at the
3063 .git/objects directory or by running
3065 ------------------------------------------------
3067 0 objects, 0 kilobytes
3068 ------------------------------------------------
3070 Although the object files are gone, any commands that refer to those
3071 objects will work exactly as they did before.
3073 The gitlink:git-gc[1] command performs packing, pruning, and more for
3074 you, so is normally the only high-level command you need.
3076 [[dangling-objects]]
3080 The gitlink:git-fsck[1] command will sometimes complain about dangling
3081 objects. They are not a problem.
3083 The most common cause of dangling objects is that you've rebased a
3084 branch, or you have pulled from somebody else who rebased a branch--see
3085 <<cleaning-up-history>>. In that case, the old head of the original
3086 branch still exists, as does everything it pointed to. The branch
3087 pointer itself just doesn't, since you replaced it with another one.
3089 There are also other situations that cause dangling objects. For
3090 example, a "dangling blob" may arise because you did a "git add" of a
3091 file, but then, before you actually committed it and made it part of the
3092 bigger picture, you changed something else in that file and committed
3093 that *updated* thing - the old state that you added originally ends up
3094 not being pointed to by any commit or tree, so it's now a dangling blob
3097 Similarly, when the "recursive" merge strategy runs, and finds that
3098 there are criss-cross merges and thus more than one merge base (which is
3099 fairly unusual, but it does happen), it will generate one temporary
3100 midway tree (or possibly even more, if you had lots of criss-crossing
3101 merges and more than two merge bases) as a temporary internal merge
3102 base, and again, those are real objects, but the end result will not end
3103 up pointing to them, so they end up "dangling" in your repository.
3105 Generally, dangling objects aren't anything to worry about. They can
3106 even be very useful: if you screw something up, the dangling objects can
3107 be how you recover your old tree (say, you did a rebase, and realized
3108 that you really didn't want to - you can look at what dangling objects
3109 you have, and decide to reset your head to some old dangling state).
3111 For commits, you can just use:
3113 ------------------------------------------------
3114 $ gitk <dangling-commit-sha-goes-here> --not --all
3115 ------------------------------------------------
3117 This asks for all the history reachable from the given commit but not
3118 from any branch, tag, or other reference. If you decide it's something
3119 you want, you can always create a new reference to it, e.g.,
3121 ------------------------------------------------
3122 $ git branch recovered-branch <dangling-commit-sha-goes-here>
3123 ------------------------------------------------
3125 For blobs and trees, you can't do the same, but you can still examine
3126 them. You can just do
3128 ------------------------------------------------
3129 $ git show <dangling-blob/tree-sha-goes-here>
3130 ------------------------------------------------
3132 to show what the contents of the blob were (or, for a tree, basically
3133 what the "ls" for that directory was), and that may give you some idea
3134 of what the operation was that left that dangling object.
3136 Usually, dangling blobs and trees aren't very interesting. They're
3137 almost always the result of either being a half-way mergebase (the blob
3138 will often even have the conflict markers from a merge in it, if you
3139 have had conflicting merges that you fixed up by hand), or simply
3140 because you interrupted a "git fetch" with ^C or something like that,
3141 leaving _some_ of the new objects in the object database, but just
3142 dangling and useless.
3144 Anyway, once you are sure that you're not interested in any dangling
3145 state, you can just prune all unreachable objects:
3147 ------------------------------------------------
3149 ------------------------------------------------
3151 and they'll be gone. But you should only run "git prune" on a quiescent
3152 repository - it's kind of like doing a filesystem fsck recovery: you
3153 don't want to do that while the filesystem is mounted.
3155 (The same is true of "git-fsck" itself, btw - but since
3156 git-fsck never actually *changes* the repository, it just reports
3157 on what it found, git-fsck itself is never "dangerous" to run.
3158 Running it while somebody is actually changing the repository can cause
3159 confusing and scary messages, but it won't actually do anything bad. In
3160 contrast, running "git prune" while somebody is actively changing the
3161 repository is a *BAD* idea).
3164 include::glossary.txt[]
3167 Notes and todo list for this manual
3168 ===================================
3170 This is a work in progress.
3172 The basic requirements:
3173 - It must be readable in order, from beginning to end, by
3174 someone intelligent with a basic grasp of the unix
3175 commandline, but without any special knowledge of git. If
3176 necessary, any other prerequisites should be specifically
3177 mentioned as they arise.
3178 - Whenever possible, section headings should clearly describe
3179 the task they explain how to do, in language that requires
3180 no more knowledge than necessary: for example, "importing
3181 patches into a project" rather than "the git-am command"
3183 Think about how to create a clear chapter dependency graph that will
3184 allow people to get to important topics without necessarily reading
3185 everything in between.
3187 Say something about .gitignore.
3189 Scan Documentation/ for other stuff left out; in particular:
3193 list of commands in gitlink:git[1]
3195 Scan email archives for other stuff left out
3197 Scan man pages to see if any assume more background than this manual
3200 Simplify beginning by suggesting disconnected head instead of
3201 temporary branch creation?
3203 Add more good examples. Entire sections of just cookbook examples
3204 might be a good idea; maybe make an "advanced examples" section a
3205 standard end-of-chapter section?
3207 Include cross-references to the glossary, where appropriate.
3209 Document shallow clones? See draft 1.5.0 release notes for some
3212 Add a section on working with other version control systems, including
3213 CVS, Subversion, and just imports of series of release tarballs.
3215 More details on gitweb?
3217 Write a chapter on using plumbing and writing scripts.