4 Backward compatibility notes
5 ----------------------------
7 When "git push [$there]" does not say what to push, we have used the
8 traditional "matching" semantics so far (all your branches were sent
9 to the remote as long as there already are branches of the same name
10 over there). In Git 2.0, the default is now the "simple" semantics,
13 - only the current branch to the branch with the same name, and only
14 when the current branch is set to integrate with that remote
15 branch, if you are pushing to the same remote as you fetch from; or
17 - only the current branch to the branch with the same name, if you
18 are pushing to a remote that is not where you usually fetch from.
20 You can use the configuration variable "push.default" to change
21 this. If you are an old-timer who wants to keep using the
22 "matching" semantics, you can set the variable to "matching", for
23 example. Read the documentation for other possibilities.
25 When "git add -u" and "git add -A" are run inside a subdirectory
26 without specifying which paths to add on the command line, they
27 operate on the entire tree for consistency with "git commit -a" and
28 other commands (these commands used to operate only on the current
29 subdirectory). Say "git add -u ." or "git add -A ." if you want to
30 limit the operation to the current directory.
32 "git add <path>" is the same as "git add -A <path>" now, so that
33 "git add dir/" will notice paths you removed from the directory and
34 record the removal. In older versions of Git, "git add <path>" used
35 to ignore removals. You can say "git add --ignore-removal <path>" to
36 add only added or modified paths in <path>, if you really want to.
38 The "-q" option to "git diff-files", which does *NOT* mean "quiet",
39 has been removed (it told Git to ignore deletion, which you can do
40 with "git diff-files --diff-filter=d").
42 "git request-pull" lost a few "heuristics" that often led to mistakes.
45 Updates since v1.9 series
46 -------------------------
48 UI, Workflows & Features
50 * "git tag --list" output can be sorted using "version sort" with
51 "--sort=version:refname".
53 * Discard the accumulated "heuristics" to guess from which branch the
54 result wants to be pulled from and make sure what the end user
55 specified is not second-guessed by "git request-pull", to avoid
56 mistakes. When you pushed out your 'master' branch to your public
57 repository as 'for-linus', use the new "master:for-linus" syntax to
58 denote the branch to be pulled.
60 * "git grep" learned to behave in a way similar to native grep when
61 "-h" (no header) and "-c" (count) options are given.
63 * transport-helper, fast-import and fast-export have been updated to
64 allow the ref mapping and ref deletion in a way similar to the
65 natively supported transports.
67 * The "simple" mode is the default for "git push".
69 * "git add -u" and "git add -A", when run without any pathspec, is a
70 tree-wide operation even when run inside a subdirectory of a
73 * "git add <path> is the same as "git add -A <path>" now.
75 * "core.statinfo" configuration variable, which is a
76 never-advertised synonym to "core.checkstat", has been removed.
78 * The "-q" option to "git diff-files", which does *NOT* mean
79 "quiet", has been removed (it told Git to ignore deletion, which
80 you can do with "git diff-files --diff-filter=d").
82 * Server operators can loosen the "tips of refs only" restriction for
83 the remote archive service with the uploadarchive.allowUnreachable
86 * The progress indicators from various time-consuming commands have
87 been marked for i18n/l10n.
89 * "git notes -C <blob>" diagnoses an attempt to use an object that
90 is not a blob as an error.
92 * "git config" learned to read from the standard input when "-" is
93 given as the value to its "--file" parameter (attempting an
94 operation to update the configuration in the standard input of
97 * Trailing whitespaces in .gitignore files, unless they are quoted
98 for fnmatch(3), e.g. "path\ ", are warned and ignored. Strictly
99 speaking, this is a backward incompatible change, but very unlikely
100 to bite any sane user and adjusting should be obvious and easy.
102 * Many commands that create commits, e.g. "pull", "rebase",
103 learned to take the --gpg-sign option on the command line.
105 * "git commit" can be told to always GPG sign the resulting commit
106 by setting "commit.gpgsign" configuration variable to true (the
107 command line option --no-gpg-sign should override it).
109 * "git pull" can be told to only accept fast-forward by setting the
110 new "pull.ff" configuration.
112 * "git reset" learned "-N" option, which does not reset the index
113 fully for paths the index knows about but the tree-ish the command
114 resets to does not (these paths are kept as intend-to-add entries).
116 * Newly cloned submodule repositories by "git submodule update",
117 when the "checkout" update mode is used, will be on a local
118 branch instead of on a detached HEAD, just like submodules added
119 with "git submodule add".
122 Performance, Internal Implementation, etc.
124 * We started using wildmatch() in place of fnmatch(3) a few releases
125 ago; complete the process and stop using fnmatch(3).
127 * Uses of curl's "multi" interface and "easy" interface do not mix
128 well when we attempt to reuse outgoing connections. Teach the RPC
129 over http code, used in the smart HTTP transport, not to use the
132 * The bitmap-index feature from JGit has been ported, which should
133 significantly improve performance when serving objects form a
134 repository that uses it.
136 * The way "git log --cc" shows a combined diff against multiple
137 parents have been optimized.
139 * The prefixcmp() and suffixcmp() functions are gone. Use
140 starts_with() and ends_with(), and also consider if skip_prefix()
141 suits your needs better when using the former.
144 Also contains various documentation updates and code clean-ups. Many
145 of them came from flurry of activities as GSoC candidate microproject
149 Fixes since v1.9 series
150 -----------------------
152 Unless otherwise noted, all the fixes since v1.9 in the maintenance
153 track are contained in this release (see the maintenance releases'
156 * Serving objects from a shallow repository needs to write a
157 new file to hold the temporary shallow boundaries but it was not
158 cleaned when we exit due to die() or a signal.
159 (merge 7839632 jk/shallow-update-fix later to maint).
161 * When "git stash pop" stops after failing to apply the stash
162 (e.g. due to conflicting changes), the stash is not dropped. State
163 that explicitly in the output to let the users know.
164 (merge 2d4c993 jc/stash-pop-not-popped later to maint).
166 * The labels in "git status" output that describe the nature of
167 conflicts (e.g. "both deleted") were limited to 20 bytes, which was
168 too short for some l10n (e.g. fr).
169 (merge c7cb333 jn/wt-status later to maint).
171 * "git clean -d pathspec" did not use the given pathspec correctly
172 and ended up cleaning too much.
173 (merge 1f2e108 jk/clean-d-pathspec later to maint).
175 * "git difftool" misbehaved when the repository is bound to the
176 working tree with the ".git file" mechanism, where a textual file
177 ".git" tells us where it is.
178 (merge fcfec8b da/difftool-git-files later to maint).
180 * "git push" did not pay attention to branch.*.pushremote if it is
181 defined earlier than remote.pushdefault; the order of these two
182 variables in the configuration file should not matter, but it did
184 (merge 98b406f jk/remote-pushremote-config-reading later to maint).
186 * Codepaths that parse timestamps in commit objects have been
188 (merge 3f419d4 jk/commit-dates-parsing-fix later to maint).
190 * "git diff --external-diff" incorrectly fed the submodule directory
191 in the working tree to the external diff driver when it knew it is
192 the same as one of the versions being compared.
193 (merge aba4727 tr/diff-submodule-no-reuse-worktree later to maint).
195 * "git reset" needs to refresh the index when working in a working
196 tree (it can also be used to match the index to the HEAD in an
197 otherwise bare repository), but it failed to set up the working
198 tree properly, causing GIT_WORK_TREE to be ignored.
199 (merge b7756d4 nd/reset-setup-worktree later to maint).
201 * "git check-attr" when working on a repository with a working tree
202 did not work well when the working tree was specified via the
203 --work-tree (and obviously with --git-dir) option.
204 (merge cdbf623 jc/check-attr-honor-working-tree later to maint).
206 * "merge-recursive" was broken in 1.7.7 era and stopped working in
207 an empty (temporary) working tree, when there are renames
208 involved. This has been corrected.
209 (merge 6e2068a bk/refresh-missing-ok-in-merge-recursive later to maint.)
211 * "git rev-parse" was loose in rejecting command line arguments
212 that do not make sense, e.g. "--default" without the required
213 value for that option.
214 (merge a43219f ds/rev-parse-required-args later to maint.)
216 * include.path variable (or any variable that expects a path that
217 can use ~username expansion) in the configuration file is not a
218 boolean, but the code failed to check it.
219 (merge 67beb60 jk/config-path-include-fix later to maint.)
221 * Commands that take pathspecs on the command line misbehaved when
222 the pathspec is given as an absolute pathname (which is a
223 practice not particularly encouraged) that points at a symbolic
224 link in the working tree.
225 (merge later 655ee9e mw/symlinks to maint.)
227 * "git diff --quiet -- pathspec1 pathspec2" sometimes did not return
228 correct status value.
229 (merge f34b205 nd/diff-quiet-stat-dirty later to maint.)
231 * Attempting to deepen a shallow repository by fetching over smart
232 HTTP transport failed in the protocol exchange, when no-done
233 extension was used. The fetching side waited for the list of
234 shallow boundary commits after the sending end stopped talking to
236 (merge 0232852 nd/http-fetch-shallow-fix later to maint.)
238 * Allow "git cmd path/", when the 'path' is where a submodule is
239 bound to the top-level working tree, to match 'path', despite the
240 extra and unnecessary trailing slash (such a slash is often
241 given by command line completion).
242 (merge 2e70c01 nd/submodule-pathspec-ending-with-slash later to maint.)