1Git v1.8.2 Release Notes 2======================== 3 4Backward compatibility notes 5---------------------------- 6 7In the upcoming major release (tentatively called 1.8.2), we will 8change the behavior of the "git push" command. 9 10When "git push [$there]" does not say what to push, we have used the 11traditional "matching" semantics so far (all your branches were sent 12to the remote as long as there already are branches of the same name 13over there). We will use the "simple" semantics that pushes the 14current branch to the branch with the same name, only when the current 15branch is set to integrate with that remote branch. There is a user 16preference configuration variable "push.default" to change this. 17 18"git push $there tag v1.2.3" used to allow replacing a tag v1.2.3 19that already exists in the repository $there, if the rewritten tag 20you are pushing points at a commit that is a decendant of a commit 21that the old tag v1.2.3 points at. This was found to be error prone 22and starting with this release, any attempt to update an existing 23ref under refs/tags/ hierarchy will fail, without "--force". 24 25 26Updates since v1.8.1 27-------------------- 28 29UI, Workflows & Features 30 31 * Initial ports to QNX and z/OS UNIX System Services have started. 32 33 * Output from the tests is coloured using "green is okay, yellow is 34 questionable, red is bad and blue is informative" scheme. 35 36 * In bare repositories, "git shortlog" and other commands now read 37 mailmap files from the tip of the history, to help running these 38 tools in server settings. 39 40 * Color specifiers, e.g. "%C(blue)Hello%C(reset)", used in the 41 "--format=" option of "git log" and friends can be disabled when 42 the output is not sent to a terminal by prefixing them with 43 "auto,", e.g. "%C(auto,blue)Hello%C(auto,reset)". 44 45 * Scripts can ask Git that wildcard patterns in pathspecs they give do 46 not have any significance, i.e. take them as literal strings. 47 48 * The patterns in .gitignore and .gitattributes files can have **/, 49 as a pattern that matches 0 or more levels of subdirectory. 50 E.g. "foo/**/bar" matches "bar" in "foo" itself or in a 51 subdirectory of "foo". 52 53 * "git blame" (and "git diff") learned the "--no-follow" option. 54 55 * "git check-ignore" command to help debugging .gitignore files has 56 been added. 57 58 * "git cherry-pick" can be used to replay a root commit to an unborn 59 branch. 60 61 * "git commit" can be told to use --cleanup=whitespace by setting the 62 configuration variable commit.cleanup to 'whitespace'. 63 64 * "git fetch --mirror" and fetch that uses other forms of refspec 65 with wildcard used to attempt to update a symbolic ref that match 66 the wildcard on the receiving end, which made little sense (the 67 real ref that is pointed at by the symbolic ref would be updated 68 anyway). Symbolic refs no longer are affected by such a fetch. 69 70 * "git format-patch" now detects more cases in which a whole branch 71 is being exported, and uses the description for the branch, when 72 asked to write a cover letter for the series. 73 74 * "git format-patch" learned "-v $count" option, and prepends a 75 string "v$count-" to the names of its output files, and also 76 automatically sets the subject prefix to "PATCH v$count". This 77 allows patches from rerolled series to be stored under different 78 names and makes it easier to reuse cover letter messsages. 79 80 * "git log" and friends can be told with --use-mailmap option to 81 rewrite the names and email addresses of people using the mailmap 82 mechanism. 83 84 * "git push" now requires "-f" to update a tag, even if it is a 85 fast-forward, as tags are meant to be fixed points. 86 87 * "git push" will stop without doing anything if the new "pre-push" 88 hook exists and exits with a failure. 89 90 * When "git rebase" fails to generate patches to be applied (e.g. due 91 to oom), it failed to detect the failure and instead behaved as if 92 there were nothing to do. A workaround to use a temporary file has 93 been applied, but we probably would want to revisit this later, as 94 it hurts the common case of not failing at all. 95 96 * Input and preconditions to "git reset" has been loosened where 97 appropriate. "git reset $fromtree Makefile" requires $fromtree to 98 be any tree (it used to require it to be a commit), for example. 99 "git reset" (without options or parameters) used to error out when 100 you do not have any commits in your history, but it now gives you 101 an empty index (to match non-existent commit you are not even on). 102 103 * "git submodule" started learning a new mode to integrate with the 104 tip of the remote branch (as opposed to integrating with the commit 105 recorded in the superproject's gitlink). 106 107 108Foreign Interface 109 110 * "git fast-export" has been updated for its use in the context of 111 the remote helper interface. 112 113 * A new remote helper to interact with bzr has been added to contrib/. 114 115 * "git p4" got various bugfixes around its branch handling. 116 117 * The remote helper to interact with Hg in contrib/ has seen a few 118 fixes. 119 120 121Performance, Internal Implementation, etc. 122 123 * "git fsck" has been taught to be pickier about entries in tree 124 objects that should not be there, e.g. ".", ".git", and "..". 125 126 * Matching paths with common forms of pathspecs that contain wildcard 127 characters has been optimized further. 128 129 * "git reset" internals has been reworked and should be faster in 130 general. We tried to be careful not to break any behaviour but 131 there could be corner cases, especially when running the command 132 from a conflicted state, that we may have missed. 133 134 * The implementation of "imap-send" has been updated to reuse xml 135 quoting code from http-push codepath, and lost a lot of unused 136 code. 137 138 * There is a simple-minded checker for the test scripts in t/ 139 directory to catch most common mistakes (it is not enabled by 140 default). 141 142 * You can build with USE_WILDMATCH=YesPlease to use a replacement 143 implementation of pattern matching logic used for pathname-like 144 things, e.g. refnames and paths in the repository. This new 145 implementation is not expected change the existing behaviour of Git 146 in this release, except for "git for-each-ref" where you can now 147 say "refs/**/master" and match with both refs/heads/master and 148 refs/remotes/origin/master. We plan to use this new implementation 149 in wider places (e.g. "git ls-files '**/Makefile' may find Makefile 150 at the top-level, and "git log '**/t*.sh'" may find commits that 151 touch a shell script whose name begins with "t" at any level) in 152 future versions of Git, but we are not there yet. By building with 153 USE_WILDMATCH, using the resulting Git daily and reporting when you 154 find breakages, you can help us get closer to that goal. 155 156 157Also contains minor documentation updates and code clean-ups. 158 159 160Fixes since v1.8.1 161------------------ 162 163Unless otherwise noted, all the fixes since v1.8.1 in the maintenance 164track are contained in this release (see release notes to them for 165details). 166 167 * An element on GIT_CEILING_DIRECTORIES list that does not name the 168 real path to a directory (i.e. a symbolic link) could have caused 169 the GIT_DIR discovery logic to escape the ceiling. 170 171 * When attempting to read the XDG-style $HOME/.config/git/config and 172 finding that $HOME/.config/git is a file, we gave a wrong error 173 message, instead of treating the case as "a custom config file does 174 not exist there" and moving on. 175 176 * The behaviour visible to the end users was confusing, when they 177 attempt to kill a process spawned in the editor that was in turn 178 launched by Git with SIGINT (or SIGQUIT), as Git would catch that 179 signal and die. We ignore these signals now. 180 (merge 1250857 pf/editor-ignore-sigint later to maint). 181 182 * A child process that was killed by a signal (e.g. SIGINT) was 183 reported in an inconsistent way depending on how the process was 184 spawned by us, with or without a shell in between. 185 186 * After failing to create a temporary file using mkstemp(), failing 187 pathname was not reported correctly on some platforms. 188 189 * The attribute mechanism didn't allow limiting attributes to be 190 applied to only a single directory itself with "path/" like the 191 exclude mechanism does. The initial implementation of this that 192 was merged to 'maint' and 1.8.1.2 was with a severe performance 193 degradations and needs to merge a fix-up topic. 194 (merge 9db9eec nd/fix-directory-attrs-off-by-one later to maint). 195 196 * "git am" did not parse datestamp correctly from Hg generated patch, 197 when it is run in a locale outside C (or en). 198 (merge 5185b97 dl/am-hg-locale later to maint). 199 200 * "git apply" misbehaved when fixing whitespace breakages by removing 201 excess trailing blank lines. 202 203 * A tar archive created by "git archive" recorded a directory in a 204 way that made NetBSD's implementation of "tar" sometimes unhappy. 205 206 * "git archive" did not record uncompressed size in the header when 207 streaming a zip archive, which confused some implementations of unzip. 208 209 * "git clean" showed what it was going to do, but sometimes end up 210 finding that it was not allowed to do so, which resulted in a 211 confusing output (e.g. after saying that it will remove an 212 untracked directory, it found an embedded git repository there 213 which it is not allowed to remove). It now performs the actions 214 and then reports the outcome more faithfully. 215 (merge f538a91 zk/clean-report-failure later to maint). 216 217 * When "git clone --separate-git-dir=$over_there" is interrupted, it 218 failed to remove the real location of the $GIT_DIR it created. 219 This was most visible when interrupting a submodule update. 220 221 * The way "git svn" asked for password using SSH_ASKPASS and 222 GIT_ASKPASS was not in line with the rest of the system. 223 224 * The --graph code fell into infinite loop when asked to do what the 225 code did not expect. 226 227 * http transport was wrong to ask for the username when the 228 authentication is done by certificate identity. 229 230 * "git pack-refs" that ran in parallel to another process that 231 created new refs had a nasty race. 232 233 * After "git add -N" and then writing a tree object out of the 234 index, the cache-tree data structure got corrupted. 235 236 * "git clone" used to allow --bare and --separate-git-dir=$there 237 options at the same time, which was nonsensical. 238 (merge 95b63f1 nd/clone-no-separate-git-dir-with-bare later to maint). 239 240 * "git rebase --preserve-merges" lost empty merges in recent versions 241 of Git. 242 (merge 9869778 ph/rebase-preserve-all-merges later to maint). 243 244 * "git merge --no-edit" computed who were involved in the work done 245 on the side branch, even though that information is to be discarded 246 without getting seen in the editor. 247 248 * "git merge" started calling prepare-commit-msg hook like "git 249 commit" does some time ago, but forgot to pay attention to the exit 250 status of the hook. 251 252 * When users spell "cc:" in lowercase in the fake "header" in the 253 trailer part, "git send-email" failed to pick up the addresses from 254 there. As e-mail headers field names are case insensitive, this 255 script should follow suit and treat "cc:" and "Cc:" the same way. 256 257 * Output from "git status --ignored" showed an unexpected interaction 258 with "--untracked". 259 260 * "gitweb", when sorting by age to show repositories with new 261 activities first, used to sort repositories with absolutely 262 nothing in it early, which was not very useful. 263 264 * "gitweb"'s code to sanitize control characters before passing it to 265 "highlight" filter lost known-to-be-safe control characters by 266 mistake. 267 268 * When a line to be wrapped has a solid run of non space characters 269 whose length exactly is the wrap width, "git shortlog -w" failed 270 to add a newline after such a line. 271 272 * Command line completion leaked an unnecessary error message while 273 looking for possible matches with paths in <tree-ish>. 274 275 * Command line completion for "tcsh" emitted an unwanted space 276 after completing a single directory name. 277 278 * Command line completion code was inadvertently made incompatible with 279 older versions of bash by using a newer array notation. 280 (merge 50c5885 bc/fix-array-syntax-for-3.0-in-completion-bash later to maint). 281 282 * Some shells do not behave correctly when IFS is unset; work it 283 around by explicitly setting it to the default value. 284 285 * Some scripted programs written in Python did not get updated when 286 PYTHON_PATH changed. 287 (cherry-pick 96a4647fca54031974cd6ad1 later to maint). 288 289 * When autoconf is used, any build on a different commit always ran 290 "config.status --recheck" even when unnecessary. 291 292 * We have been carrying a translated and long-unmaintained copy of an 293 old version of the tutorial; removed. 294 295 * t0050 had tests expecting failures from a bug that was fixed some 296 time ago. 297 (merge 336e2e2 tb/t0050-maint later to maint). 298 299 * t4014, t9502 and t0200 tests had various portability issues that 300 broke on OpenBSD. 301 302 * t9020 and t3600 tests had various portability issues. 303 304 * t9200 runs "cvs init" on a directory that already exists, but a 305 platform can configure this fail for the current user (e.g. you 306 need to be in the cvsadmin group on NetBSD 6.0). 307 308 * t9020 and t9810 had a few non-portable shell script construct.