1Git 2.9 Release Notes 2===================== 3 4Backward compatibility notes 5---------------------------- 6 7The end-user facing Porcelain level commands in the "git diff" and 8"git log" family by default enable the rename detection; you can still 9use "diff.renames" configuration variable to disable this. 10 11Merging two branches that have no common ancestor with "git merge" is 12by default forbidden now to prevent creating such an unusual merge by 13mistake. 14 15The output formats of "git log" that indents the commit log message by 164 spaces now expands HT in the log message by default. You can use 17the "--no-expand-tabs" option to disable this. 18 19"git commit-tree" plumbing command required the user to always sign 20its result when the user sets the commit.gpgsign configuration 21variable, which was an ancient mistake, which this release corrects. 22A script that drives commit-tree, if it relies on this mistake, now 23needs to read commit.gpgsign and pass the -S option as necessary. 24 25 26Updates since v2.8 27------------------ 28 29UI, Workflows & Features 30 31 * Comes with git-multimail 1.3.1 (in contrib/). 32 33 * The end-user facing commands like "git diff" and "git log" 34 now enable the rename detection by default. 35 36 * The credential.helper configuration variable is cumulative and 37 there is no good way to override it from the command line. As 38 a special case, giving an empty string as its value now serves 39 as the signal to clear the values specified in various files. 40 41 * A new "interactive.diffFilter" configuration can be used to 42 customize the diff shown in "git add -i" sessions. 43 44 * "git p4" now allows P4 author names to be mapped to Git author 45 names. 46 47 * "git rebase -x" can be used without passing "-i" option. 48 49 * "git -c credential.<var>=<value> submodule" can now be used to 50 propagate configuration variables related to credential helper 51 down to the submodules. 52 53 * "git tag" can create an annotated tag without explicitly given an 54 "-a" (or "-s") option (i.e. when a tag message is given). A new 55 configuration variable, tag.forceSignAnnotated, can be used to tell 56 the command to create signed tag in such a situation. 57 58 * "git merge" used to allow merging two branches that have no common 59 base by default, which led to a brand new history of an existing 60 project created and then get pulled by an unsuspecting maintainer, 61 which allowed an unnecessary parallel history merged into the 62 existing project. The command has been taught not to allow this by 63 default, with an escape hatch "--allow-unrelated-histories" option 64 to be used in a rare event that merges histories of two projects 65 that started their lives independently. 66 67 * "git pull" has been taught to pass the "--allow-unrelated-histories" 68 option to underlying "git merge". 69 70 * "git apply -v" learned to report paths in the patch that were 71 skipped via --include/--exclude mechanism or being outside the 72 current working directory. 73 74 * Shell completion (in contrib/) updates. 75 76 * The commit object name reported when "rebase -i" stops has been 77 shortened. 78 79 * "git worktree add" can be given "--no-checkout" option to only 80 create an empty worktree without checking out the files. 81 82 * "git mergetools" learned to drive ExamDiff. 83 84 * "git pull --rebase" learned "--[no-]autostash" option, so that 85 the rebase.autostash configuration variable set to true can be 86 overridden from the command line. 87 88 * When "git log" shows the log message indented by 4-spaces, the 89 remainder of a line after a HT does not align in the way the author 90 originally intended. The command now expands tabs by default to help 91 such a case, and allows the users to override it with a new option, 92 "--no-expand-tabs". 93 94 * "git send-email" now uses a more readable timestamps when 95 formulating a message ID. 96 97 * "git rerere" can encounter two or more files with the same conflict 98 signature that have to be resolved in different ways, but there was 99 no way to record these separate resolutions. 100 101 * "git p4" learned to record P4 jobs in Git commit that imports from 102 the history in Perforce. 103 104 * "git describe --contains" often made a hard-to-justify choice of 105 tag to name a given commit, because it tried to come up 106 with a name with smallest number of hops from a tag, causing an old 107 commit whose close descendant that is recently tagged were not 108 described with respect to an old tag but with a newer tag. It did 109 not help that its computation of "hop" count was further tweaked to 110 penalize being on a side branch of a merge. The logic has been 111 updated to favor using the tag with the oldest tagger date, which 112 is a lot easier to explain to the end users: "We describe a commit 113 in terms of the (chronologically) oldest tag that contains the 114 commit." 115 (merge 7550424 js/name-rev-use-oldest-ref later to maint). 116 117 * "git clone" learned the "--shallow-submodules" option. 118 119 * HTTP transport clients learned to throw extra HTTP headers at the 120 server, specified via http.extraHeader configuration variable. 121 122 * Patch output from "git diff" and friends has been tweaked to be 123 more readable by using a blank line as a strong hint that the 124 contents before and after it belong to logically separate units. 125 126 * A new configuration variable core.hooksPath allows customizing 127 where the hook directory is. 128 129 * An earlier addition of "sanitize_submodule_env" with 14111fc4 (git: 130 submodule honor -c credential.* from command line, 2016-02-29) 131 turned out to be a convoluted no-op; implement what it wanted to do 132 correctly, and stop filtering settings given via "git -c var=val". 133 134 * "git commit --dry-run" reported "No, no, you cannot commit." in one 135 case where "git commit" would have allowed you to commit, and this 136 improves it a little bit ("git commit --dry-run --short" still does 137 not give you the correct answer, for example). This is a stop-gap 138 measure in that "commit --short --dry-run" still gives an incorrect 139 result. 140 141 * The experimental "multiple worktree" feature gains more safety to 142 forbid operations on a branch that is checked out or being actively 143 worked on elsewhere, by noticing that e.g. it is being rebased. 144 145 * "git format-patch" learned a new "--base" option to record what 146 (public, well-known) commit the original series was built on in 147 its output. 148 149 * "git commit" learned to pay attention to the "commit.verbose" 150 configuration variable and act as if the "--verbose" option 151 was given from the command line. 152 153 * Updated documentation gives hints to GMail users with two-factor 154 auth enabled that they need app-specific-password when using 155 "git send-email". 156 157 * The manpage output of our documentation did not render well in 158 terminal; typeset literals in bold by default to make them stand 159 out more. 160 161 * The mark-up in the top-level README.md file has been updated to 162 typeset CLI command names differently from the body text. 163 164 165Performance, Internal Implementation, Development Support etc. 166 167 * The embedded args argv-array in the child process is used to build 168 the command line to run pack-objects instead of using a separate 169 array of strings. 170 171 * A test for tags has been restructured so that more parts of it can 172 easily be run on a platform without a working GnuPG. 173 174 * The startup_info data, which records if we are working inside a 175 repository (among other things), are now uniformly available to Git 176 subcommand implementations, and Git avoids attempting to touch 177 references when we are not in a repository. 178 179 * The command line argument parser for "receive-pack" has been 180 rewritten to use parse-options. 181 182 * A major part of "git submodule update" has been ported to C to take 183 advantage of the recently added framework to run download tasks in 184 parallel. Other updates to "git submodule" that move pieces of 185 logic to C continues. 186 187 * Rename bunch of tests on "git clone" for better organization. 188 189 * The tests that involve running httpd leaked the system-wide 190 configuration in /etc/gitconfig to the tested environment. 191 192 * Build updates for MSVC. 193 194 * The repository set-up sequence has been streamlined (the biggest 195 change is that there is no longer git_config_early()), so that we 196 do not attempt to look into refs/* when we know we do not have a 197 Git repository. 198 199 * Code restructuring around the "refs" API to prepare for pluggable 200 refs backends. 201 202 * Sources to many test helper binaries and the generated helpers 203 have been moved to t/helper/ subdirectory to reduce clutter at the 204 top level of the tree. 205 206 * Unify internal logic between "git tag -v" and "git verify-tag" 207 commands by making one directly call into the other. 208 209 * "merge-recursive" strategy incorrectly checked if a path that is 210 involved in its internal merge exists in the working tree. 211 212 * The test scripts for "git p4" (but not "git p4" implementation 213 itself) has been updated so that they would work even on a system 214 where the installed version of Python is python 3. 215 216 * As nobody maintains our in-tree git.spec.in and distros use their 217 own spec file, we stopped pretending that we support "make rpm". 218 219 * Move from "unsigned char[20]" to "struct object_id" continues. 220 221 * The code for warning_errno/die_errno has been refactored and a new 222 error_errno() reporting helper is introduced. 223 (merge 1da045f nd/error-errno later to maint). 224 225 * Running tests with '-x' option to trace the individual command 226 executions is a useful way to debug test scripts, but some tests 227 that capture the standard error stream and check what the command 228 said can be broken with the trace output mixed in. When running 229 our tests under "bash", however, we can redirect the trace output 230 to another file descriptor to keep the standard error of programs 231 being tested intact. 232 (merge d88785e jk/test-send-sh-x-trace-elsewhere later to maint). 233 234 * t0040 had too many unnecessary repetitions in its test data. Teach 235 test-parse-options program so that a caller can tell what it 236 expects in its output, so that these repetitions can be cleaned up. 237 238 * Add perf test for "rebase -i". 239 240 * Common mistakes when writing gitlink: in our documentation are 241 found by "make check-docs". 242 243 * t9xxx series has been updated primarily for readability, while 244 fixing small bugs in it. A few scripted Porcelain commands have 245 also been updated to fix possible bugs around their use of 246 "test -z" and "test -n". 247 248 * CI test was taught to run git-svn tests. 249 250 * "git cat-file --batch-all" has been sped up, by taking advantage 251 of the fact that it does not have to read a list of objects, in two 252 ways. 253 254 * test updates to make it more readable and maintainable. 255 (merge e6273f4 es/t1500-modernize later to maint). 256 257 * "make DEVELOPER=1" worked as expected; setting DEVELOPER=1 in 258 config.mak didn't. 259 (merge 51dd3e8 mm/makefile-developer-can-be-in-config-mak later to maint). 260 261 262Also contains various documentation updates and code clean-ups. 263 264 265Fixes since v2.8 266---------------- 267 268Unless otherwise noted, all the fixes since v2.8 in the maintenance 269track are contained in this release (see the maintenance releases' 270notes for details). 271 272 * "git config --get-urlmatch", unlike other variants of the "git 273 config --get" family, did not signal error with its exit status 274 when there was no matching configuration. 275 276 * The "--local-env-vars" and "--resolve-git-dir" options of "git 277 rev-parse" failed to work outside a repository when the command's 278 option parsing was rewritten in 1.8.5 era. 279 280 * "git index-pack --keep[=<msg>] pack-$name.pack" simply did not work. 281 282 * Fetching of history by naming a commit object name directly didn't 283 work across remote-curl transport. 284 285 * A small memory leak in an error codepath has been plugged in xdiff 286 code. 287 288 * strbuf_getwholeline() did not NUL-terminate the buffer on certain 289 corner cases in its error codepath. 290 291 * "git mergetool" did not work well with conflicts that both sides 292 deleted. 293 294 * "git send-email" had trouble parsing alias file in mailrc format 295 when lines in it had trailing whitespaces on them. 296 297 * When "git merge --squash" stopped due to conflict, the concluding 298 "git commit" failed to read in the SQUASH_MSG that shows the log 299 messages from all the squashed commits. 300 301 * "git merge FETCH_HEAD" dereferenced NULL pointer when merging 302 nothing into an unborn history (which is arguably unusual usage, 303 which perhaps was the reason why nobody noticed it). 304 305 * When "git worktree" feature is in use, "git branch -d" allowed 306 deletion of a branch that is checked out in another worktree, 307 which was wrong. 308 309 * When "git worktree" feature is in use, "git branch -m" renamed a 310 branch that is checked out in another worktree without adjusting 311 the HEAD symbolic ref for the worktree. 312 313 * "git diff -M" used to work better when two originally identical 314 files A and B got renamed to X/A and X/B by pairing A to X/A and B 315 to X/B, but this was broken in the 2.0 timeframe. 316 317 * "git send-pack --all <there>" was broken when its command line 318 option parsing was written in the 2.6 timeframe. 319 320 * "git format-patch --help" showed `-s` and `--no-patch` as if these 321 are valid options to the command. We already hide `--patch` option 322 from the documentation, because format-patch is about showing the 323 diff, and the documentation now hides these options as well. 324 325 * When running "git blame $path" with unnormalized data in the index 326 for the path, the data in the working tree was blamed, even though 327 "git add" would not have changed what is already in the index, due 328 to "safe crlf" that disables the line-end conversion. It has been 329 corrected. 330 331 * A change back in version 2.7 to "git branch" broke display of a 332 symbolic ref in a non-standard place in the refs/ hierarchy (we 333 expect symbolic refs to appear in refs/remotes/*/HEAD to point at 334 the primary branch the remote has, and as .git/HEAD to point at the 335 branch we locally checked out). 336 337 * A partial rewrite of "git submodule" in the 2.7 timeframe changed 338 the way the gitdir: pointer in the submodules point at the real 339 repository location to use absolute paths by accident. This has 340 been corrected. 341 342 * "git commit" misbehaved in a few minor ways when an empty message 343 is given via -m '', all of which has been corrected. 344 345 * Support for CRAM-MD5 authentication method in "git imap-send" did 346 not work well. 347 348 * Upcoming OpenSSL 1.1.0 will break compilation by updating a few API 349 elements we use in imap-send, which has been adjusted for the change. 350 351 * The socks5:// proxy support added back in 2.6.4 days was not aware 352 that socks5h:// proxies behave differently from socks5:// proxies. 353 354 * "git config" had a codepath that tried to pass a NULL to 355 printf("%s"), which nobody seems to have noticed. 356 357 * On Cygwin, object creation uses the "create a temporary and then 358 rename it to the final name" pattern, not "create a temporary, 359 hardlink it to the final name and then unlink the temporary" 360 pattern. 361 362 This is necessary to use Git on Windows shared directories, and is 363 already enabled for the MinGW and plain Windows builds. It also 364 has been used in Cygwin packaged versions of Git for quite a while. 365 See http://thread.gmane.org/gmane.comp.version-control.git/291853 366 367 * "merge-octopus" strategy did not ensure that the index is clean 368 when merge begins. 369 370 * When "git merge" notices that the merge can be resolved purely at 371 the tree level (without having to merge blobs) and the resulting 372 tree happens to already exist in the object store, it forgot to 373 update the index, which left an inconsistent state that would 374 break later operations. 375 376 * "git submodule" reports the paths of submodules the command 377 recurses into, but these paths were incorrectly reported when 378 the command was not run from the root level of the superproject. 379 380 * The "user.useConfigOnly" configuration variable makes it an error 381 if users do not explicitly set user.name and user.email. However, 382 its check was not done early enough and allowed another error to 383 trigger, reporting that the default value we guessed from the 384 system setting was unusable. This was a suboptimal end-user 385 experience as we want the users to set user.name/user.email without 386 relying on the auto-detection at all. 387 388 * "git mv old new" did not adjust the path for a submodule that lives 389 as a subdirectory inside old/ directory correctly. 390 391 * "git replace -e" did not honour "core.editor" configuration. 392 393 * "git push" from a corrupt repository that attempts to push a large 394 number of refs deadlocked; the thread to relay rejection notices 395 for these ref updates blocked on writing them to the main thread, 396 after the main thread at the receiving end notices that the push 397 failed and decides not to read these notices and return a failure. 398 399 * mmap emulation on Windows has been optimized and work better without 400 consuming paging store when not needed. 401 402 * A question by "git send-email" to ask the identity of the sender 403 has been updated. 404 405 * UI consistency improvements for "git mergetool". 406 407 * "git rebase -m" could be asked to rebase an entire branch starting 408 from the root, but failed by assuming that there always is a parent 409 commit to the first commit on the branch. 410 411 * Fix a broken "p4 lfs" test. 412 413 * Recent update to Git LFS broke "git p4" by changing the output from 414 its "lfs pointer" subcommand. 415 416 * "git fetch" test t5510 was flaky while running a (forced) automagic 417 garbage collection. 418 419 * Documentation updates to help contributors setting up Travis CI 420 test for their patches. 421 422 * Some multi-byte encoding can have a backslash byte as a later part 423 of one letter, which would confuse "highlight" filter used in 424 gitweb. 425 426 * "git commit-tree" plumbing command required the user to always sign 427 its result when the user sets the commit.gpgsign configuration 428 variable, which was an ancient mistake. Rework "git rebase" that 429 relied on this mistake so that it reads commit.gpgsign and pass (or 430 not pass) the -S option to "git commit-tree" to keep the end-user 431 expectation the same, while teaching "git commit-tree" to ignore 432 the configuration variable. This will stop requiring the users to 433 sign commit objects used internally as an implementation detail of 434 "git stash". 435 436 * "http.cookieFile" configuration variable clearly wants a pathname, 437 but we forgot to treat it as such by e.g. applying tilde expansion. 438 (merge e5a39ad bn/http-cookiefile-config later to maint). 439 440 * Consolidate description of tilde-expansion that is done to 441 configuration variables that take pathname to a single place. 442 (merge dca83ab jc/config-pathname-type later to maint). 443 444 * Correct faulty recommendation to use "git submodule deinit ." when 445 de-initialising all submodules, which would result in a strange 446 error message in a pathological corner case. 447 (merge f6a5279 sb/submodule-deinit-all later to maint). 448 449 * Many 'linkgit:<git documentation page>' references were broken, 450 which are all fixed with this. 451 (merge 1cca17d jc/linkgit-fix later to maint). 452 453 * "git rerere" can get confused by conflict markers deliberately left 454 by the inner merge step, because they are indistinguishable from 455 the real conflict markers left by the outermost merge which are 456 what the end user and "rerere" need to look at. This was fixed by 457 making the conflict markers left by the inner merges a bit longer. 458 (merge 0f9fd5c jc/ll-merge-internal later to maint). 459 460 * CI test was taught to build documentation pages. 461 (merge b98712b ls/travis-build-doc later to maint). 462 463 * "git fsck" learned to catch NUL byte in a commit object as 464 potential error and warn. 465 (merge 6d2d780 jc/fsck-nul-in-commit later to maint). 466 467 * Portability enhancement for "rebase -i" to help platforms whose 468 shell does not like "for i in <empty>" (which is not POSIX-kosher). 469 (merge 8e98b35 jk/rebase-interactive-eval-fix later to maint). 470 471 * On Windows, .git and optionally any files whose name starts with a 472 dot are now marked as hidden, with a core.hideDotFiles knob to 473 customize this behaviour. 474 (merge ebf31e7 js/windows-dotgit later to maint). 475 476 * Documentation for "git merge --verify-signatures" has been updated 477 to clarify that the signature of only the commit at the tip is 478 verified. Also the phrasing used for signature and key validity is 479 adjusted to align with that used by OpenPGP. 480 (merge 05a5869 kf/gpg-sig-verification-doc later to maint). 481 482 * A couple of bugs around core.autocrlf have been fixed. 483 (merge caa47ad tb/core-eol-fix later to maint). 484 485 * Many commands normalize command line arguments from NFD to NFC 486 variant of UTF-8 on OSX, but commands in the "diff" family did 487 not, causing "git diff $path" to complain that no such path is 488 known to Git. They have been taught to do the normalization. 489 (merge 90a78b8 ar/diff-args-osx-precompose later to maint). 490 491 * "git difftool" learned to handle unmerged paths correctly in 492 dir-diff mode. 493 (merge 366f9ce da/difftool later to maint). 494 495 * The "are we talking with TTY, doing an interactive session?" 496 detection has been updated to work better for "Git for Windows". 497 (merge f7f90e0 kb/msys2-tty later to maint). 498 499 * We forgot to add "git log --decorate=auto" to documentation when we 500 added the feature back in v2.1.0 timeframe. 501 (merge 462cbb4 rj/log-decorate-auto later to maint). 502 503 * "git fast-import --export-marks" would overwrite the existing marks 504 file even when it makes a dump from its custom die routine. 505 Prevent it from doing so when we have an import-marks file but 506 haven't finished reading it. 507 (merge f4beed6 fc/fast-import-broken-marks-file later to maint). 508 509 * "git rebase -i", after it fails to auto-resolve the conflict, had 510 an unnecessary call to "git rerere" from its very early days, which 511 was spotted recently; the call has been removed. 512 (merge 7063693 js/rebase-i-dedup-call-to-rerere later to maint). 513 514 * Other minor clean-ups and documentation updates 515 (merge 832c0e5 lp/typofixes later to maint). 516 (merge f5ee54a sb/z-is-gnutar-ism later to maint). 517 (merge 2e3926b va/i18n-misc-updates later to maint). 518 (merge f212dcc bn/config-doc-tt-varnames later to maint). 519 (merge f54bea4 nd/remote-plural-ours-plus-theirs later to maint). 520 (merge 2bb0518 ak/t4151-ls-files-could-be-empty later to maint). 521 (merge 4df4313 jc/test-seq later to maint). 522 (merge a75a308 tb/t5601-sed-fix later to maint). 523 (merge 6c1fbe1 va/i18n-remote-comment-to-align later to maint). 524 (merge dee2303 va/mailinfo-doc-typofix later to maint). 525 (merge cd82b7a pa/cherry-pick-doc-typo later to maint). 526 (merge 2bb73ae rs/patch-id-use-skip-prefix later to maint). 527 (merge aa20cbc rs/apply-name-terminate later to maint).