1Git 2.19 Release Notes 2====================== 3 4Updates since v2.18 5------------------- 6 7UI, Workflows & Features 8 9 * "git diff" compares the index and the working tree. For paths 10 added with intent-to-add bit, the command shows the full contents 11 of them as added, but the paths themselves were not marked as new 12 files. They are now shown as new by default. 13 14 "git apply" learned the "--intent-to-add" option so that an 15 otherwise working-tree-only application of a patch will add new 16 paths to the index marked with the "intent-to-add" bit. 17 18 * "git grep" learned the "--column" option that gives not just the 19 line number but the column number of the hit. 20 21 * The "-l" option in "git branch -l" is an unfortunate short-hand for 22 "--create-reflog", but many users, both old and new, somehow expect 23 it to be something else, perhaps "--list". This step warns when "-l" 24 is used as a short-hand for "--create-reflog" and warns about the 25 future repurposing of the it when it is used. 26 27 * The userdiff pattern for .php has been updated. 28 29 * The content-transfer-encoding of the message "git send-email" sends 30 out by default was 8bit, which can cause trouble when there is an 31 overlong line to bust RFC 5322/2822 limit. A new option 'auto' to 32 automatically switch to quoted-printable when there is such a line 33 in the payload has been introduced and is made the default. 34 35 * "git checkout" and "git worktree add" learned to honor 36 checkout.defaultRemote when auto-vivifying a local branch out of a 37 remote tracking branch in a repository with multiple remotes that 38 have tracking branches that share the same names. 39 (merge 8d7b558bae ab/checkout-default-remote later to maint). 40 41 * "git grep" learned the "--only-matching" option. 42 43 * "git rebase --rebase-merges" mode now handles octopus merges as 44 well. 45 46 * Add a server-side knob to skip commits in exponential/fibbonacci 47 stride in an attempt to cover wider swath of history with a smaller 48 number of iterations, potentially accepting a larger packfile 49 transfer, instead of going back one commit a time during common 50 ancestor discovery during the "git fetch" transaction. 51 (merge 42cc7485a2 jt/fetch-negotiator-skipping later to maint). 52 53 * A new configuration variable core.usereplacerefs has been added, 54 primarily to help server installations that want to ignore the 55 replace mechanism altogether. 56 57 * Teach "git tag -s" etc. a few configuration variables (gpg.format 58 that can be set to "openpgp" or "x509", and gpg.<format>.program 59 that is used to specify what program to use to deal with the format) 60 to allow x.509 certs with CMS via "gpgsm" to be used instead of 61 openpgp via "gnupg". 62 63 * Many more strings are prepared for l10n. 64 65 * "git p4 submit" learns to ask its own pre-submit hook if it should 66 continue with submitting. 67 68 69Performance, Internal Implementation, Development Support etc. 70 71 * The bulk of "git submodule foreach" has been rewritten in C. 72 73 * The in-core "commit" object had an all-purpose "void *util" field, 74 which was tricky to use especially in library-ish part of the 75 code. All of the existing uses of the field has been migrated to a 76 more dedicated "commit-slab" mechanism and the field is eliminated. 77 78 * A less often used command "git show-index" has been modernized. 79 (merge fb3010c31f jk/show-index later to maint). 80 81 * The conversion to pass "the_repository" and then "a_repository" 82 throughout the object access API continues. 83 84 * Continuing with the idea to programatically enumerate various 85 pieces of data required for command line completion, teach the 86 codebase to report the list of configuration variables 87 subcommands care about to help complete them. 88 89 * Separate "rebase -p" codepath out of "rebase -i" implementation to 90 slim down the latter and make it easier to manage. 91 92 * Make refspec parsing codepath more robust. 93 94 * Some flaky tests have been fixed. 95 96 * Continuing with the idea to programmatically enumerate various 97 pieces of data required for command line completion, the codebase 98 has been taught to enumerate options prefixed with "--no-" to 99 negate them. 100 101 * Build and test procedure for netrc credential helper (in contrib/) 102 has been updated. 103 104 * The conversion to pass "the_repository" and then "a_repository" 105 throughout the object access API continues. 106 107 * Remove unused function definitions and declarations from ewah 108 bitmap subsystem. 109 110 * Code preparation to make "git p4" closer to be usable with Python 3. 111 112 * Tighten the API to make it harder to misuse in-tree .gitmodules 113 file, even though it shares the same syntax with configuration 114 files, to read random configuration items from it. 115 116 * "git fast-import" has been updated to avoid attempting to create 117 delta against a zero-byte-long string, which is pointless. 118 119 * The codebase has been updated to compile cleanly with -pedantic 120 option. 121 (merge 2b647a05d7 bb/pedantic later to maint). 122 123 * The character display width table has been updated to match the 124 latest Unicode standard. 125 (merge 570951eea2 bb/unicode-11-width later to maint). 126 127 * test-lint now looks for broken use of "VAR=VAL shell_func" in test 128 scripts. 129 130 * Conversion from uchar[40] to struct object_id continues. 131 132 * Recent "security fix" to pay attention to contents of ".gitmodules" 133 while accepting "git push" was a bit overly strict than necessary, 134 which has been adjusted. 135 136 * "git fsck" learns to make sure the optional commit-graph file is in 137 a sane state. 138 139 * "git diff --color-moved" feature has further been tweaked. 140 141 * Code restructuring and a small fix to transport protocol v2 during 142 fetching. 143 144 * Parsing of -L[<N>][,[<M>]] parameters "git blame" and "git log" 145 take has been tweaked. 146 147 * lookup_commit_reference() and friends have been updated to find 148 in-core object for a specific in-core repository instance. 149 150 * Various glitches in the heuristics of merge-recursive strategy have 151 been documented in new tests. 152 153 * "git fetch" learned a new option "--negotiation-tip" to limit the 154 set of commits it tells the other end as "have", to reduce wasted 155 bandwidth and cycles, which would be helpful when the receiving 156 repository has a lot of refs that have little to do with the 157 history at the remote it is fetching from. 158 159 * For a large tree, the index needs to hold many cache entries 160 allocated on heap. These cache entries are now allocated out of a 161 dedicated memory pool to amortize malloc(3) overhead. 162 163 * Tests to cover various conflicting cases have been added for 164 merge-recursive. 165 166 * Tests to cover conflict cases that involve submodules have been 167 added for merge-recursive. 168 169 * Look for broken "&&" chains that are hidden in subshell, many of 170 which have been found and corrected. 171 172 * The singleton commit-graph in-core instance is made per in-core 173 repository instance. 174 175 * "make DEVELOPER=1 DEVOPTS=pedantic" allows developers to compile 176 with -pedantic option, which may catch more problematic program 177 constructs and potential bugs. 178 179 * Preparatory code to later add json output for telemetry data has 180 been added. 181 182 * Update the way we use Coccinelle to find out-of-style code that 183 need to be modernised. 184 185 * It is too easy to misuse system API functions such as strcat(); 186 these selected functions are now forbidden in this codebase and 187 will cause a compilation failure. 188 189 * Add a script (in contrib/) to help users of VSCode work better with 190 our codebase. 191 192 * The Travis CI scripts were taught to ship back the test data from 193 failed tests. 194 (merge aea8879a6a sg/travis-retrieve-trash-upon-failure later to maint). 195 196 197Fixes since v2.18 198----------------- 199 200 * "git remote update" can take both a single remote nickname and a 201 nickname for remote groups, and the completion script (in contrib/) 202 has been taught about it. 203 (merge 9cd4382ad5 ls/complete-remote-update-names later to maint). 204 205 * "git fetch --shallow-since=<cutoff>" that specifies the cut-off 206 point that is newer than the existing history used to end up 207 grabbing the entire history. Such a request now errors out. 208 (merge e34de73c56 nd/reject-empty-shallow-request later to maint). 209 210 * Fix for 2.17-era regression around `core.safecrlf`. 211 (merge 6cb09125be as/safecrlf-quiet-fix later to maint). 212 213 * The recent addition of "partial clone" experimental feature kicked 214 in when it shouldn't, namely, when there is no partial-clone filter 215 defined even if extensions.partialclone is set. 216 (merge cac1137dc4 jh/partial-clone later to maint). 217 218 * "git send-pack --signed" (hence "git push --signed" over the http 219 transport) did not read user ident from the config mechanism to 220 determine whom to sign the push certificate as, which has been 221 corrected. 222 (merge d067d98887 ms/send-pack-honor-config later to maint). 223 224 * "git fetch-pack --all" used to unnecessarily fail upon seeing an 225 annotated tag that points at an object other than a commit. 226 (merge c12c9df527 jk/fetch-all-peeled-fix later to maint). 227 228 * When user edits the patch in "git add -p" and the user's editor is 229 set to strip trailing whitespaces indiscriminately, an empty line 230 that is unchanged in the patch would become completely empty 231 (instead of a line with a sole SP on it). The code introduced in 232 Git 2.17 timeframe failed to parse such a patch, but now it learned 233 to notice the situation and cope with it. 234 (merge f4d35a6b49 pw/add-p-recount later to maint). 235 236 * The code to try seeing if a fetch is necessary in a submodule 237 during a fetch with --recurse-submodules got confused when the path 238 to the submodule was changed in the range of commits in the 239 superproject, sometimes showing "(null)". This has been corrected. 240 241 * "git submodule" did not correctly adjust core.worktree setting that 242 indicates whether/where a submodule repository has its associated 243 working tree across various state transitions, which has been 244 corrected. 245 (merge 984cd77ddb sb/submodule-core-worktree later to maint). 246 247 * Bugfix for "rebase -i" corner case regression. 248 (merge a9279c6785 pw/rebase-i-keep-reword-after-conflict later to maint). 249 250 * Recently added "--base" option to "git format-patch" command did 251 not correctly generate prereq patch ids. 252 (merge 15b76c1fb3 xy/format-patch-prereq-patch-id-fix later to maint). 253 254 * POSIX portability fix in Makefile to fix a glitch introduced a few 255 releases ago. 256 (merge 6600054e9b dj/runtime-prefix later to maint). 257 258 * "git filter-branch" when used with the "--state-branch" option 259 still attempted to rewrite the commits whose filtered result is 260 known from the previous attempt (which is recorded on the state 261 branch); the command has been corrected not to waste cycles doing 262 so. 263 (merge 709cfe848a mb/filter-branch-optim later to maint). 264 265 * Clarify that setting core.ignoreCase to deviate from reality would 266 not turn a case-incapable filesystem into a case-capable one. 267 (merge 48294b512a ms/core-icase-doc later to maint). 268 269 * "fsck.skipList" did not prevent a blob object listed there from 270 being inspected for is contents (e.g. we recently started to 271 inspect the contents of ".gitmodules" for certain malicious 272 patterns), which has been corrected. 273 (merge fb16287719 rj/submodule-fsck-skip later to maint). 274 275 * "git checkout --recurse-submodules another-branch" did not report 276 in which submodule it failed to update the working tree, which 277 resulted in an unhelpful error message. 278 (merge ba95d4e4bd sb/submodule-move-head-error-msg later to maint). 279 280 * "git rebase" behaved slightly differently depending on which one of 281 the three backends gets used; this has been documented and an 282 effort to make them more uniform has begun. 283 (merge b00bf1c9a8 en/rebase-consistency later to maint). 284 285 * The "--ignore-case" option of "git for-each-ref" (and its friends) 286 did not work correctly, which has been fixed. 287 (merge e674eb2528 jk/for-each-ref-icase later to maint). 288 289 * "git fetch" failed to correctly validate the set of objects it 290 received when making a shallow history deeper, which has been 291 corrected. 292 (merge cf1e7c0770 jt/connectivity-check-after-unshallow later to maint). 293 294 * Partial clone support of "git clone" has been updated to correctly 295 validate the objects it receives from the other side. The server 296 side has been corrected to send objects that are directly 297 requested, even if they may match the filtering criteria (e.g. when 298 doing a "lazy blob" partial clone). 299 (merge a7e67c11b8 jt/partial-clone-fsck-connectivity later to maint). 300 301 * Handling of an empty range by "git cherry-pick" was inconsistent 302 depending on how the range ended up to be empty, which has been 303 corrected. 304 (merge c5e358d073 jk/empty-pick-fix later to maint). 305 306 * "git reset --merge" (hence "git merge ---abort") and "git reset --hard" 307 had trouble working correctly in a sparsely checked out working 308 tree after a conflict, which has been corrected. 309 (merge b33fdfc34c mk/merge-in-sparse-checkout later to maint). 310 311 * Correct a broken use of "VAR=VAL shell_func" in a test. 312 (merge 650161a277 jc/t3404-one-shot-export-fix later to maint). 313 314 * "git rev-parse ':/substring'" did not consider the history leading 315 only to HEAD when looking for a commit with the given substring, 316 when the HEAD is detached. This has been fixed. 317 (merge 6b3351e799 wc/find-commit-with-pattern-on-detached-head later to maint). 318 319 * Build doc update for Windows. 320 (merge ede8d89bb1 nd/command-list later to maint). 321 322 * core.commentchar is now honored when preparing the list of commits 323 to replay in "rebase -i". 324 325 * "git pull --rebase" on a corrupt HEAD caused a segfault. In 326 general we substitute an empty tree object when running the in-core 327 equivalent of the diff-index command, and the codepath has been 328 corrected to do so as well to fix this issue. 329 (merge 3506dc9445 jk/has-uncommitted-changes-fix later to maint). 330 331 * httpd tests saw occasional breakage due to the way its access log 332 gets inspected by the tests, which has been updated to make them 333 less flaky. 334 (merge e8b3b2e275 sg/httpd-test-unflake later to maint). 335 336 * Tests to cover more D/F conflict cases have been added for 337 merge-recursive. 338 339 * "git gc --auto" opens file descriptors for the packfiles before 340 spawning "git repack/prune", which would upset Windows that does 341 not want a process to work on a file that is open by another 342 process. The issue has been worked around. 343 (merge 12e73a3ce4 kg/gc-auto-windows-workaround later to maint). 344 345 * The recursive merge strategy did not properly ensure there was no 346 change between HEAD and the index before performing its operation, 347 which has been corrected. 348 (merge 55f39cf755 en/dirty-merge-fixes later to maint). 349 350 * "git rebase" started exporting GIT_DIR environment variable and 351 exposing it to hook scripts when part of it got rewritten in C. 352 Instead of matching the old scripted Porcelains' behaviour, 353 compensate by also exporting GIT_WORK_TREE environment as well to 354 lessen the damage. This can harm existing hooks that want to 355 operate on different repository, but the current behaviour is 356 already broken for them anyway. 357 (merge ab5e67d751 bc/sequencer-export-work-tree-as-well later to maint). 358 359 * "git send-email" when using in a batched mode that limits the 360 number of messages sent in a single SMTP session lost the contents 361 of the variable used to choose between tls/ssl, unable to send the 362 second and later batches, which has been fixed. 363 (merge 636f3d7ac5 jm/send-email-tls-auth-on-batch later to maint). 364 365 * The lazy clone support had a few places where missing but promised 366 objects were not correctly tolerated, which have been fixed. 367 368 * One of the "diff --color-moved" mode "dimmed_zebra" that was named 369 in an unusual way has been deprecated and replaced by 370 "dimmed-zebra". 371 (merge e3f2f5f9cd es/diff-color-moved-fix later to maint). 372 373 * The wire-protocol v2 relies on the client to send "ref prefixes" to 374 limit the bandwidth spent on the initial ref advertisement. "git 375 clone" when learned to speak v2 forgot to do so, which has been 376 corrected. 377 (merge 402c47d939 bw/clone-ref-prefixes later to maint). 378 379 * "git diff --histogram" had a bad memory usage pattern, which has 380 been rearranged to reduce the peak usage. 381 (merge 79cb2ebb92 sb/histogram-less-memory later to maint). 382 383 * Code clean-up to use size_t/ssize_t when they are the right type. 384 (merge 7726d360b5 jk/size-t later to maint). 385 386 * The wire-protocol v2 relies on the client to send "ref prefixes" to 387 limit the bandwidth spent on the initial ref advertisement. "git 388 fetch $remote branch:branch" that asks tags that point into the 389 history leading to the "branch" automatically followed sent to 390 narrow prefix and broke the tag following, which has been fixed. 391 (merge 2b554353a5 jt/tag-following-with-proto-v2-fix later to maint). 392 393 * When the sparse checkout feature is in use, "git cherry-pick" and 394 other mergy operations lost the skip_worktree bit when a path that 395 is excluded from checkout requires content level merge, which is 396 resolved as the same as the HEAD version, without materializing the 397 merge result in the working tree, which made the path appear as 398 deleted. This has been corrected by preserving the skip_worktree 399 bit (and not materializing the file in the working tree). 400 (merge 2b75fb601c en/merge-recursive-skip-fix later to maint). 401 402 * Code cleanup, docfix, build fix, etc. 403 (merge aee9be2ebe sg/update-ref-stdin-cleanup later to maint). 404 (merge 037714252f jc/clean-after-sanity-tests later to maint). 405 (merge 5b26c3c941 en/merge-recursive-cleanup later to maint). 406 (merge 0dcbc0392e bw/config-refer-to-gitsubmodules-doc later to maint). 407 (merge bb4d000e87 bw/protocol-v2 later to maint). 408 (merge 928f0ab4ba vs/typofixes later to maint). 409 (merge d7f590be84 en/rebase-i-microfixes later to maint). 410 (merge 81d395cc85 js/rebase-recreate-merge later to maint). 411 (merge 51d1863168 tz/exclude-doc-smallfixes later to maint). 412 (merge a9aa3c0927 ds/commit-graph later to maint). 413 (merge 5cf8e06474 js/enhanced-version-info later to maint). 414 (merge 6aaded5509 tb/config-default later to maint). 415 (merge 022d2ac1f3 sb/blame-color later to maint). 416 (merge 5a06a20e0c bp/test-drop-caches-for-windows later to maint). 417 (merge dd61cc1c2e jk/ui-color-always-to-auto later to maint). 418 (merge 1e83b9bfdd sb/trailers-docfix later to maint). 419 (merge ab29f1b329 sg/fast-import-dump-refs-on-checkpoint-fix later to maint). 420 (merge 6a8ad880f0 jn/subtree-test-fixes later to maint). 421 (merge ffbd51cc60 nd/pack-objects-threading-doc later to maint). 422 (merge e9dac7be60 es/mw-to-git-chain-fix later to maint). 423 (merge fe583c6c7a rs/remote-mv-leakfix later to maint).