1Git 2.14 Release Notes 2====================== 3 4Backward compatibility notes. 5 6 * Use of an empty string as a pathspec element that is used for 7 'everything matches' is still warned and Git asks users to use a 8 more explicit '.' for that instead. The hope is that existing 9 users will not mind this change, and eventually the warning can be 10 turned into a hard error, upgrading the deprecation into removal of 11 this (mis)feature. That is not scheduled to happen in the upcoming 12 release (yet). 13 14 * Git now avoids blindly falling back to ".git" when the setup 15 sequence said we are _not_ in Git repository. A corner case that 16 happens to work right now may be broken by a call to die("BUG"). 17 We've tried hard to locate such cases and fixed them, but there 18 might still be cases that need to be addressed--bug reports are 19 greatly appreciated. 20 21 * The experiment to improve the hunk-boundary selection of textual 22 diff output has finished, and the "indent heuristics" has now 23 become the default. 24 25 26Updates since v2.13 27------------------- 28 29UI, Workflows & Features 30 31 * The colors in which "git status --short --branch" showed the names 32 of the current branch and its remote-tracking branch are now 33 configurable. 34 35 * "git clone" learned the "--no-tags" option not to fetch all tags 36 initially, and also set up the tagopt not to follow any tags in 37 subsequent fetches. 38 39 * "git archive --format=zip" learned to use zip64 extension when 40 necessary to go beyond the 4GB limit. 41 42 * "git reset" learned "--recurse-submodules" option. 43 44 * "git diff --submodule=diff" now recurses into nested submodules. 45 46 * "git repack" learned to accept the --threads=<n> option and pass it 47 to pack-objects. 48 49 * "git send-email" learned to run sendemail-validate hook to inspect 50 and reject a message before sending it out. 51 52 * There is no good reason why "git fetch $there $sha1" should fail 53 when the $sha1 names an object at the tip of an advertised ref, 54 even when the other side hasn't enabled allowTipSHA1InWant. 55 56 * The recently introduced "[includeIf "gitdir:$dir"] path=..." 57 mechansim has further been taught to take symlinks into account. 58 The directory "$dir" specified in "gitdir:$dir" may be a symlink to 59 a real location, not something that $(getcwd) may return. In such 60 a case, a realpath of "$dir" is compared with the real path of the 61 current repository to determine if the contents from the named path 62 should be included. 63 64 * Make the "indent" heuristics the default in "diff" and diff.indentHeuristics 65 configuration variable an escape hatch for those who do no want it. 66 67 * Many commands learned to pay attention to submodule.recurse 68 configuration. 69 70 * The convention for a command line is to follow "git cmdname 71 --options" with revisions followed by an optional "--" 72 disambiguator and then finally pathspecs. When "--" is not there, 73 we make sure early ones are all interpretable as revs (and do not 74 look like paths) and later ones are the other way around. A 75 pathspec with "magic" (e.g. ":/p/a/t/h" that matches p/a/t/h from 76 the top-level of the working tree, no matter what subdirectory you 77 are working from) are conservatively judged as "not a path", which 78 required disambiguation more often. The command line parser 79 learned to say "it's a pathspec" a bit more often when the syntax 80 looks like so. 81 82 * Update "perl-compatible regular expression" support to enable JIT 83 and also allow linking with the newer PCRE v2 library. 84 85 * "filter-branch" learned a pseudo filter "--setup" that can be used 86 to define common functions/variables that can be used by other 87 filters. 88 89 * Using "git add d/i/r" when d/i/r is the top of the working tree of 90 a separate repository would create a gitlink in the index, which 91 would appear as a not-quite-initialized submodule to others. We 92 learned to give warnings when this happens. 93 94 * "git status" learned to optionally give how many stash entries the 95 user has in its output. 96 97 98Performance, Internal Implementation, Development Support etc. 99 100 * The default packed-git limit value has been raised on larger 101 platforms to save "git fetch" from a (recoverable) failure while 102 "gc" is running in parallel. 103 104 * Code to update the cache-tree has been tightened so that we won't 105 accidentally write out any 0{40} entry in the tree object. 106 107 * Attempt to allow us notice "fishy" situation where we fail to 108 remove the temporary directory used during the test. 109 110 * Travis CI gained a task to format the documentation with both 111 AsciiDoc and AsciiDoctor. 112 113 * Some platforms have ulong that is smaller than time_t, and our 114 historical use of ulong for timestamp would mean they cannot 115 represent some timestamp that the platform allows. Invent a 116 separate and dedicated timestamp_t (so that we can distingiuish 117 timestamps and a vanilla ulongs, which along is already a good 118 move), and then declare uintmax_t is the type to be used as the 119 timestamp_t. 120 121 * We can trigger Windows auto-build tester (credits: Dscho & 122 Microsoft) from our existing Travis CI tester now. 123 124 * Conversion from uchar[20] to struct object_id continues. 125 126 * Simplify parse_pathspec() codepath and stop it from looking at the 127 default in-core index. 128 129 * Add perf-test for wildmatch. 130 131 * Code from "conversion using external process" codepath has been 132 extracted to a separate sub-process.[ch] module. 133 134 * When "git checkout", "git merge", etc. manipulates the in-core 135 index, various pieces of information in the index extensions are 136 discarded from the original state, as it is usually not the case 137 that they are kept up-to-date and in-sync with the operation on the 138 main index. The untracked cache extension is copied across these 139 operations now, which would speed up "git status" (as long as the 140 cache is properly invalidated). 141 142 * The internal implementation of "git grep" has seen some clean-up. 143 144 * Update the C style recommendation for notes for translators, as 145 recent versions of gettext tools can work with our style of 146 multi-line comments. 147 148 * The implementation of "ref" API around the "packed refs" have been 149 cleaned up, in preparation for further changes. 150 151 * The internal logic used in "git blame" has been libified to make it 152 easier to use by cgit. 153 154 * Our code often opens a path to an optional file, to work on its 155 contents when we can successfully open it. We can ignore a failure 156 to open if such an optional file does not exist, but we do want to 157 report a failure in opening for other reasons (e.g. we got an I/O 158 error, or the file is there, but we lack the permission to open). 159 160 The exact errors we need to ignore are ENOENT (obviously) and 161 ENOTDIR (less obvious). Instead of repeating comparison of errno 162 with these two constants, introduce a helper function to do so. 163 164 * We often try to open a file for reading whose existence is 165 optional, and silently ignore errors from open/fopen; report such 166 errors if they are not due to missing files. 167 168 * When an existing repository is used for t/perf testing, we first 169 create bit-for-bit copy of it, which may grab a transient state of 170 the repository and freeze it into the repository used for testing, 171 which then may cause Git operations to fail. Single out "the index 172 being locked" case and forcibly drop the lock from the copy. 173 174 * Three instances of the same helper function have been consolidated 175 to one. 176 177 * "fast-import" uses a default pack chain depth that is consistent 178 with other parts of the system. 179 180 * A new test to show the interaction between the pattern [^a-z] 181 (which matches '/') and a slash in a path has been added. The 182 pattern should not match the slash with "pathmatch", but should 183 with "wildmatch". 184 185 * The 'diff-highlight' program (in contrib/) has been restructured 186 for easier reuse by an external project 'diff-so-fancy'. 187 (merge 0c977dbc81 jk/diff-highlight-module later to maint). 188 189 * A common pattern to free a piece of memory and assign NULL to the 190 pointer that used to point at it has been replaced with a new 191 FREE_AND_NULL() macro. 192 193 194Also contains various documentation updates and code clean-ups. 195 196 197Fixes since v2.13 198----------------- 199 200Unless otherwise noted, all the fixes since v2.13 in the maintenance 201track are contained in this release (see the maintenance releases' 202notes for details). 203 204 * "git gc" did not interact well with "git worktree"-managed 205 per-worktree refs. 206 207 * "git cherry-pick" and other uses of the sequencer machinery 208 mishandled a trailer block whose last line is an incomplete line. 209 This has been fixed so that an additional sign-off etc. are added 210 after completing the existing incomplete line. 211 212 * The codepath in "git am" that is used when running "git rebase" 213 leaked memory held for the log message of the commits being rebased. 214 215 * "git clone --config var=val" is a way to populate the 216 per-repository configuration file of the new repository, but it did 217 not work well when val is an empty string. This has been fixed. 218 219 * Setting "log.decorate=false" in the configuration file did not take 220 effect in v2.13, which has been corrected. 221 222 * A few codepaths in "checkout" and "am" working on an unborn branch 223 tried to access an uninitialized piece of memory. 224 225 * The Web interface to gmane news archive is long gone, even though 226 the articles are still accessible via NTTP. Replace the links with 227 ones to public-inbox.org. Because their message identification is 228 based on the actual message-id, it is likely that it will be easier 229 to migrate away from it if/when necessary. 230 231 * The receive-pack program now makes sure that the push certificate 232 records the same set of push options used for pushing. 233 234 * Tests have been updated to pass under GETTEXT_POISON (a mechanism 235 to ensure that output strings that should not be translated are 236 not translated by mistake), and TravisCI is told to run them. 237 238 * "git checkout --recurse-submodules" did not quite work with a 239 submodule that itself has submodules. 240 241 * "pack-objects" can stream a slice of an existing packfile out when 242 the pack bitmap can tell that the reachable objects are all needed 243 in the output, without inspecting individual objects. This 244 strategy however would not work well when "--local" and other 245 options are in use, and need to be disabled. 246 247 * Fix memory leaks pointed out by Coverity (and people). 248 249 * "git read-tree -m" (no tree-ish) gave a nonsense suggestion "use 250 --empty if you want to clear the index". With "-m", such a request 251 will still fail anyway, as you'd need to name at least one tree-ish 252 to be merged. 253 254 * Make sure our tests would pass when the sources are checked out 255 with "platform native" line ending convention by default on 256 Windows. Some "text" files out tests use and the test scripts 257 themselves that are meant to be run with /bin/sh, ought to be 258 checked out with eol=LF even on Windows. 259 260 * Introduce the BUG() macro to improve die("BUG: ..."). 261 262 * Clarify documentation for include.path and includeIf.<condition>.path 263 configuration variables. 264 265 * Git sometimes gives an advice in a rhetorical question that does 266 not require an answer, which can confuse new users and non native 267 speakers. Attempt to rephrase them. 268 269 * A few http:// links that are redirected to https:// in the 270 documentation have been updated to https:// links. 271 272 * "git for-each-ref --format=..." with %(HEAD) in the format used to 273 resolve the HEAD symref as many times as it had processed refs, 274 which was wasteful, and "git branch" shared the same problem. 275 276 * Regression fix to topic recently merged to 'master'. 277 278 * The shell completion script (in contrib/) learned "git stash" has 279 a new "push" subcommand. 280 281 * "git interpret-trailers", when used as GIT_EDITOR for "git commit 282 -v", looked for and appended to a trailer block at the very end, 283 i.e. at the end of the "diff" output. The command has been 284 corrected to pay attention to the cut-mark line "commit -v" adds to 285 the buffer---the real trailer block should appear just before it. 286 287 * A test allowed both "git push" and "git receive-pack" on the other 288 end write their traces into the same file. This is OK on platforms 289 that allows atomically appending to a file opened with O_APPEND, 290 but on other platforms led to a mangled output, causing 291 intermittent test failures. This has been fixed by disabling 292 traces from "receive-pack" in the test. 293 294 * Tag objects, which are not reachable from any ref, that point at 295 missing objects were mishandled by "git gc" and friends (they 296 should silently be ignored instead) 297 298 * "git describe --contains" penalized light-weight tags so much that 299 they were almost never considered. Instead, give them about the 300 same chance to be considered as an annotated tag that is the same 301 age as the underlying commit would. 302 303 * The "run-command" API implementation has been made more robust 304 against dead-locking in a threaded environment. 305 306 * A recent update to t5545-push-options.sh started skipping all the 307 tests in the script when a web server testing is disabled or 308 unavailable, not just the ones that require a web server. Non HTTP 309 tests have been salvaged to always run in this script. 310 311 * "git send-email" now uses Net::SMTP::SSL, which is obsolete, only 312 when needed. Recent versions of Net::SMTP can do TLS natively. 313 314 * "foo\bar\baz" in "git fetch foo\bar\baz", even though there is no 315 slashes in it, cannot be a nickname for a remote on Windows, as 316 that is likely to be a pathname on a local filesystem. 317 318 * "git clean -d" used to clean directories that has ignored files, 319 even though the command should not lose ignored ones without "-x". 320 "git status --ignored" did not list ignored and untracked files 321 without "-uall". These have been corrected. 322 323 * The result from "git diff" that compares two blobs, e.g. "git diff 324 $commit1:$path $commit2:$path", used to be shown with the full 325 object name as given on the command line, but it is more natural to 326 use the $path in the output and use it to look up .gitattributes. 327 328 * The "collision detecting" SHA-1 implementation shipped with 2.13 329 was quite broken on some big-endian platforms and/or platforms that 330 do not like unaligned fetches. Update to the upstream code which 331 has already fixed these issues. 332 333 * "git am -h" triggered a BUG(). 334 335 * The interaction of "url.*.insteadOf" and custom URL scheme's 336 whitelisting is now documented better. 337 338 * The timestamp of the index file is now taken after the file is 339 closed, to help Windows, on which a stale timestamp is reported by 340 fstat() on a file that is opened for writing and data was written 341 but not yet closed. 342 343 * "git pull --rebase --autostash" didn't auto-stash when the local history 344 fast-forwards to the upstream. 345 346 * A flaky test has been corrected. 347 348 * "git $cmd -h" for builtin commands calls the implementation of the 349 command (i.e. cmd_$cmd() function) without doing any repository 350 set-up, and the commands that expect RUN_SETUP is done by the Git 351 potty needs to be prepared to show the help text without barfing. 352 (merge d691551192 jk/consistent-h later to maint). 353 354 * Help contributors that visit us at GitHub. 355 356 * "git stash push <pathspec>" did not work from a subdirectory at all. 357 Bugfix for a topic in v2.13 358 359 * As there is no portable way to pass timezone information to 360 strftime, some output format from "git log" and friends are 361 impossible to produce. Teach our own strbuf_addftime to replace %z 362 and %Z with caller-supplied values to help working around this. 363 (merge 6eced3ec5e rs/strbuf-addftime-zZ later to maint). 364 365 * "git mergetool" learned to work around a wrapper MacOS X adds 366 around underlying meld. 367 (merge 0af85f84bd da/mergetools-meld-output-opt-on-macos later to maint). 368 369 * An example in documentation that does not work in multi worktree 370 configuration has been corrected. 371 (merge 773a88914f ah/doc-gitattributes-empty-index later to maint). 372 373 * The pretty-format specifiers like '%h', '%t', etc. had an 374 optimization that no longer works correctly. In preparation/hope 375 of getting it correctly implemented, first discard the optimization 376 that is broken. 377 (merge fe9e2aefd4 rs/pretty-add-again later to maint). 378 379 * The code to pick up and execute command alias definition from the 380 configuration used to switch to the top of the working tree and 381 then come back when the expanded alias was executed, which was 382 unnecessarilyl complex. Attempt to simplify the logic by using the 383 early-config mechanism that does not chdir around. 384 (merge a9bcf6586d js/alias-early-config later to maint). 385 386 * Fix configuration codepath to pay proper attention to commondir 387 that is used in multi-worktree situation, and isolate config API 388 into its own header file. 389 (merge dc8441fdb4 bw/config-h later to maint). 390 391 * "git add -p" were updated in 2.12 timeframe to cope with custom 392 core.commentchar but the implementation was buggy and a 393 metacharacter like $ and * did not work. 394 (merge d85d7ecb80 jk/add-p-commentchar-fix later to maint). 395 396 * Other minor doc, test and build updates and code cleanups. 397 (merge 68241cb9dd sb/t4005-modernize later to maint). 398 (merge 4fced24712 ks/t7508-indent-fix later to maint). 399 (merge 968b1fe263 mb/reword-autocomplete-message later to maint).