1 NOTES ON COMMITTING TO EMACS'S REPOSITORY -*- outline -*-
5 This branch does not contain a copy of Emacs, but of the Emacs Lisp
6 package archive (elpa.gnu.org). See admin/notes/elpa for further
7 explanation, and the README file in the branch for usage
10 * Install changes only on one branch, let them get merged elsewhere if needed.
12 In particular, install bug-fixes only on the release branch (if there
13 is one) and let them get synced to the master; do not install them by
14 hand on the master as well. E.g. if there is an active "emacs-24" branch
15 and you have a bug-fix appropriate for the next emacs-24.x release,
16 install it only on the emacs-24 branch, not on the master as well.
18 Installing things manually into more than one branch makes merges more
21 http://lists.gnu.org/archive/html/emacs-devel/2010-03/msg01124.html
23 The exception is, if you know that the change will be difficult to
24 merge to the master (eg because the master code has changed a lot).
25 In that case, it's helpful if you can apply the change to both master
26 and branch yourself (when committing the branch change, indicate
27 in the commit log that it should not be merged to the master, by
28 including the phrase "Not to be merged to master", or any other phrase
29 that matches "merge").
31 * Installing changes from your personal branches.
33 If your branch has only a single commit, or many different real
34 commits, it is fine to do a merge. If your branch has only a very
35 small number of "real" commits, but several "merge from masters", it is
36 preferred that you take your branch's diff, apply it to the master, and
37 commit directly, not merge. This keeps the history cleaner.
39 In general, when working on some feature in a separate branch, it is
40 preferable not to merge from master until you are done with the
41 feature. Unless you really need some change that was done on the
42 master while you were developing on the branch, you don't really need
43 those merges; just merge once, when you are done with the feature, and
44 Bazaar will take care of the rest. Bazaar is much better in this than
45 CVS, so interim merges are unnecessary.
47 Or use shelves; or rebase; or do something else. See the thread for
48 yet another fun excursion into the exciting world of version control.
50 http://lists.gnu.org/archive/html/emacs-devel/2010-04/msg00086.html
52 * Installing changes from gnulib
54 Some of the files in Emacs are copied from gnulib. To synchronize
55 these files from the version of gnulib that you have checked out into
56 a sibling directory of your branch, type "admin/merge-gnulib"; this
57 will check out the latest version of gnulib if there is no sibling
58 directory already. It is a good idea to run "git status" afterwards,
59 so that if a gnulib module added a file, you can record the new file
60 using "git add". After synchronizing from gnulib, do a "make" in the
63 To change the set of gnulib modules, change the GNULIB_MODULES
64 variable in admin/merge-gnulib before running it.
66 If you remove a gnulib module, or if a gnulib module
67 removes a file, then remove the corresponding files by hand.
69 * How to merge changes from emacs-24 to master
71 [The section on git merge procedure has not yet been written.]
73 You may see conflicts in autoload md5sums in comments. Strictly
74 speaking, the right thing to do is merge everything else, resolve the
75 conflict by choosing either the master or branch version, then run
76 `make -C lisp autoloads' to update the md5sums to the correct master
77 value before committing.
79 * Re-adding a file that has been removed from the repository
81 Let's suppose you've done:
83 git rm file; git commit -a
85 You can just restore a copy of the file and then re-add it;
86 git does not have per-file history so this will not harm
89 Alternatively, you can do
93 where XXXXX is the hash of the commit in which file was removed.
94 This backs out the entire changeset the deletion was part of,
95 which is often more appropriate.
97 * Undoing a commit (uncommitting)
99 If you have not pushed the commit, you may be able to use `git reset
100 --hard' with a hash argument to revert the your local repo copy to the
103 If you have pushed commit, resetting will be ineffective because it
104 will only vanish the commit in your local copy. Instead, use `git
105 revert', giving it the commit ID as argument. This will create a
106 new commit that backs out the change. Then push that.
108 Note that git will generate a log message for the revert that includes
109 a git hash. Please edit this to refer to the commit by the first line
110 of its log comment, or by committer and date, or by something else
111 that is not the hash. As noted previously, it is best to avoid hashes
112 in comments in case we someday have to change version-control systems
117 This is a semi-automated way to find the revision that introduced a bug.
118 Browse `git help bisect' for technical instructions.
120 * Maintaining ChangeLog history
122 Older ChangeLog entries are kept in history files named ChangeLog.1,
123 ChangeLog.2, etc., and can be edited just as any other source files
124 can. Newer ChangeLog entries are stored in the repository as commit
125 messages, which cannot be edited directly.
127 'make ChangeLog' copies newer ChangeLog entries into a file
128 'ChangeLog' that is intended to be put into the distribution tarball.
129 This ChangeLog file is not put into the repository.
131 'make change-history' copies all newer ChangeLog entries into the
132 start of the newest ChangeLog history file. These ChangeLog entries
133 are thereafter considered to be old, so later uses of 'make ChangeLog'
134 and/or 'make change-history' will no longer copy the entries. To
135 alter ChangeLog history, run 'make change-history', then edit
136 the ChangeLog history files manually and commit your changes.