Documentation / RelNotes / 2.14.0.txton commit grep: recurse in-process using 'struct repository' (f9ee2fc)
   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   mechanism 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 * "git status" has long shown essentially the same message as "git
  98   commit"; the message it gives while preparing for the root commit,
  99   i.e. "Initial commit", was hard to understand for some new users.
 100   Now it says "No commits yet" to stress more on the current status
 101   (rather than the commit the user is preparing for, which is more in
 102   line with the focus of "git commit").
 103
 104 * "git send-email" learned to overcome some SMTP server limitation
 105   that does not allow many pieces of e-mails to be sent over a single
 106   session.
 107
 108 * An old message shown in the commit log template was removed, as it
 109   has outlived its usefulness.
 110
 111 * "git pull --rebase --recurse-submodules" learns to rebase the
 112   branch in the submodules to an updated base.
 113
 114
 115Performance, Internal Implementation, Development Support etc.
 116
 117 * The default packed-git limit value has been raised on larger
 118   platforms to save "git fetch" from a (recoverable) failure while
 119   "gc" is running in parallel.
 120
 121 * Code to update the cache-tree has been tightened so that we won't
 122   accidentally write out any 0{40} entry in the tree object.
 123
 124 * Attempt to allow us notice "fishy" situation where we fail to
 125   remove the temporary directory used during the test.
 126
 127 * Travis CI gained a task to format the documentation with both
 128   AsciiDoc and AsciiDoctor.
 129
 130 * Some platforms have ulong that is smaller than time_t, and our
 131   historical use of ulong for timestamp would mean they cannot
 132   represent some timestamp that the platform allows.  Invent a
 133   separate and dedicated timestamp_t (so that we can distingiuish
 134   timestamps and a vanilla ulongs, which along is already a good
 135   move), and then declare uintmax_t is the type to be used as the
 136   timestamp_t.
 137
 138 * We can trigger Windows auto-build tester (credits: Dscho &
 139   Microsoft) from our existing Travis CI tester now.
 140
 141 * Conversion from uchar[20] to struct object_id continues.
 142
 143 * Simplify parse_pathspec() codepath and stop it from looking at the
 144   default in-core index.
 145
 146 * Add perf-test for wildmatch.
 147
 148 * Code from "conversion using external process" codepath has been
 149   extracted to a separate sub-process.[ch] module.
 150
 151 * When "git checkout", "git merge", etc. manipulates the in-core
 152   index, various pieces of information in the index extensions are
 153   discarded from the original state, as it is usually not the case
 154   that they are kept up-to-date and in-sync with the operation on the
 155   main index.  The untracked cache extension is copied across these
 156   operations now, which would speed up "git status" (as long as the
 157   cache is properly invalidated).
 158
 159 * The internal implementation of "git grep" has seen some clean-up.
 160
 161 * Update the C style recommendation for notes for translators, as
 162   recent versions of gettext tools can work with our style of
 163   multi-line comments.
 164
 165 * The implementation of "ref" API around the "packed refs" have been
 166   cleaned up, in preparation for further changes.
 167
 168 * The internal logic used in "git blame" has been libified to make it
 169   easier to use by cgit.
 170
 171 * Our code often opens a path to an optional file, to work on its
 172   contents when we can successfully open it.  We can ignore a failure
 173   to open if such an optional file does not exist, but we do want to
 174   report a failure in opening for other reasons (e.g. we got an I/O
 175   error, or the file is there, but we lack the permission to open).
 176
 177   The exact errors we need to ignore are ENOENT (obviously) and
 178   ENOTDIR (less obvious).  Instead of repeating comparison of errno
 179   with these two constants, introduce a helper function to do so.
 180
 181 * We often try to open a file for reading whose existence is
 182   optional, and silently ignore errors from open/fopen; report such
 183   errors if they are not due to missing files.
 184
 185 * When an existing repository is used for t/perf testing, we first
 186   create bit-for-bit copy of it, which may grab a transient state of
 187   the repository and freeze it into the repository used for testing,
 188   which then may cause Git operations to fail.  Single out "the index
 189   being locked" case and forcibly drop the lock from the copy.
 190
 191 * Three instances of the same helper function have been consolidated
 192   to one.
 193
 194 * "fast-import" uses a default pack chain depth that is consistent
 195   with other parts of the system.
 196
 197 * A new test to show the interaction between the pattern [^a-z]
 198   (which matches '/') and a slash in a path has been added.  The
 199   pattern should not match the slash with "pathmatch", but should
 200   with "wildmatch".
 201
 202 * The 'diff-highlight' program (in contrib/) has been restructured
 203   for easier reuse by an external project 'diff-so-fancy'.
 204
 205 * A common pattern to free a piece of memory and assign NULL to the
 206   pointer that used to point at it has been replaced with a new
 207   FREE_AND_NULL() macro.
 208
 209 * Traditionally, the default die() routine had a code to prevent it
 210   from getting called multiple times, which interacted badly when a
 211   threaded program used it (one downside is that the real error may
 212   be hidden and instead the only error message given to the user may
 213   end up being "die recursion detected", which is not very useful).
 214
 215 * Introduce a "repository" object to eventually make it easier to
 216   work in multiple repositories (the primary focus is to work with
 217   the superproject and its submodules) in a single process.
 218
 219 * Optimize "what are the object names already taken in an alternate
 220   object database?" query that is used to derive the length of prefix
 221   an object name is uniquely abbreviated to.
 222
 223 * The hashmap API has been updated so that data to customize the
 224   behaviour of the comparison function can be specified at the time a
 225   hashmap is initialized.
 226
 227
 228Also contains various documentation updates and code clean-ups.
 229
 230
 231Fixes since v2.13
 232-----------------
 233
 234Unless otherwise noted, all the fixes since v2.13 in the maintenance
 235track are contained in this release (see the maintenance releases'
 236notes for details).
 237
 238 * "git gc" did not interact well with "git worktree"-managed
 239   per-worktree refs.
 240
 241 * "git cherry-pick" and other uses of the sequencer machinery
 242   mishandled a trailer block whose last line is an incomplete line.
 243   This has been fixed so that an additional sign-off etc. are added
 244   after completing the existing incomplete line.
 245
 246 * The codepath in "git am" that is used when running "git rebase"
 247   leaked memory held for the log message of the commits being rebased.
 248
 249 * "git clone --config var=val" is a way to populate the
 250   per-repository configuration file of the new repository, but it did
 251   not work well when val is an empty string.  This has been fixed.
 252
 253 * Setting "log.decorate=false" in the configuration file did not take
 254   effect in v2.13, which has been corrected.
 255
 256 * A few codepaths in "checkout" and "am" working on an unborn branch
 257   tried to access an uninitialized piece of memory.
 258
 259 * The Web interface to gmane news archive is long gone, even though
 260   the articles are still accessible via NTTP.  Replace the links with
 261   ones to public-inbox.org.  Because their message identification is
 262   based on the actual message-id, it is likely that it will be easier
 263   to migrate away from it if/when necessary.
 264
 265 * The receive-pack program now makes sure that the push certificate
 266   records the same set of push options used for pushing.
 267
 268 * Tests have been updated to pass under GETTEXT_POISON (a mechanism
 269   to ensure that output strings that should not be translated are
 270   not translated by mistake), and TravisCI is told to run them.
 271
 272 * "git checkout --recurse-submodules" did not quite work with a
 273   submodule that itself has submodules.
 274
 275 * "pack-objects" can stream a slice of an existing packfile out when
 276   the pack bitmap can tell that the reachable objects are all needed
 277   in the output, without inspecting individual objects.  This
 278   strategy however would not work well when "--local" and other
 279   options are in use, and need to be disabled.
 280
 281 * Fix memory leaks pointed out by Coverity (and people).
 282
 283 * "git read-tree -m" (no tree-ish) gave a nonsense suggestion "use
 284   --empty if you want to clear the index".  With "-m", such a request
 285   will still fail anyway, as you'd need to name at least one tree-ish
 286   to be merged.
 287
 288 * Make sure our tests would pass when the sources are checked out
 289   with "platform native" line ending convention by default on
 290   Windows.  Some "text" files out tests use and the test scripts
 291   themselves that are meant to be run with /bin/sh, ought to be
 292   checked out with eol=LF even on Windows.
 293
 294 * Introduce the BUG() macro to improve die("BUG: ...").
 295
 296 * Clarify documentation for include.path and includeIf.<condition>.path
 297   configuration variables.
 298
 299 * Git sometimes gives an advice in a rhetorical question that does
 300   not require an answer, which can confuse new users and non native
 301   speakers.  Attempt to rephrase them.
 302
 303 * A few http:// links that are redirected to https:// in the
 304   documentation have been updated to https:// links.
 305
 306 * "git for-each-ref --format=..." with %(HEAD) in the format used to
 307   resolve the HEAD symref as many times as it had processed refs,
 308   which was wasteful, and "git branch" shared the same problem.
 309
 310 * Regression fix to topic recently merged to 'master'.
 311
 312 * The shell completion script (in contrib/) learned "git stash" has
 313   a new "push" subcommand.
 314
 315 * "git interpret-trailers", when used as GIT_EDITOR for "git commit
 316   -v", looked for and appended to a trailer block at the very end,
 317   i.e. at the end of the "diff" output.  The command has been
 318   corrected to pay attention to the cut-mark line "commit -v" adds to
 319   the buffer---the real trailer block should appear just before it.
 320
 321 * A test allowed both "git push" and "git receive-pack" on the other
 322   end write their traces into the same file.  This is OK on platforms
 323   that allows atomically appending to a file opened with O_APPEND,
 324   but on other platforms led to a mangled output, causing
 325   intermittent test failures.  This has been fixed by disabling
 326   traces from "receive-pack" in the test.
 327
 328 * Tag objects, which are not reachable from any ref, that point at
 329   missing objects were mishandled by "git gc" and friends (they
 330   should silently be ignored instead)
 331
 332 * "git describe --contains" penalized light-weight tags so much that
 333   they were almost never considered.  Instead, give them about the
 334   same chance to be considered as an annotated tag that is the same
 335   age as the underlying commit would.
 336
 337 * The "run-command" API implementation has been made more robust
 338   against dead-locking in a threaded environment.
 339
 340 * A recent update to t5545-push-options.sh started skipping all the
 341   tests in the script when a web server testing is disabled or
 342   unavailable, not just the ones that require a web server.  Non HTTP
 343   tests have been salvaged to always run in this script.
 344
 345 * "git send-email" now uses Net::SMTP::SSL, which is obsolete, only
 346   when needed.  Recent versions of Net::SMTP can do TLS natively.
 347
 348 * "foo\bar\baz" in "git fetch foo\bar\baz", even though there is no
 349   slashes in it, cannot be a nickname for a remote on Windows, as
 350   that is likely to be a pathname on a local filesystem.
 351
 352 * "git clean -d" used to clean directories that has ignored files,
 353   even though the command should not lose ignored ones without "-x".
 354   "git status --ignored"  did not list ignored and untracked files
 355   without "-uall".  These have been corrected.
 356
 357 * The result from "git diff" that compares two blobs, e.g. "git diff
 358   $commit1:$path $commit2:$path", used to be shown with the full
 359   object name as given on the command line, but it is more natural to
 360   use the $path in the output and use it to look up .gitattributes.
 361
 362 * The "collision detecting" SHA-1 implementation shipped with 2.13
 363   was quite broken on some big-endian platforms and/or platforms that
 364   do not like unaligned fetches.  Update to the upstream code which
 365   has already fixed these issues.
 366
 367 * "git am -h" triggered a BUG().
 368
 369 * The interaction of "url.*.insteadOf" and custom URL scheme's
 370   whitelisting is now documented better.
 371
 372 * The timestamp of the index file is now taken after the file is
 373   closed, to help Windows, on which a stale timestamp is reported by
 374   fstat() on a file that is opened for writing and data was written
 375   but not yet closed.
 376
 377 * "git pull --rebase --autostash" didn't auto-stash when the local history
 378   fast-forwards to the upstream.
 379
 380 * A flaky test has been corrected.
 381
 382 * "git $cmd -h" for builtin commands calls the implementation of the
 383   command (i.e. cmd_$cmd() function) without doing any repository
 384   set-up, and the commands that expect RUN_SETUP is done by the Git
 385   potty needs to be prepared to show the help text without barfing.
 386   (merge d691551192 jk/consistent-h later to maint).
 387
 388 * Help contributors that visit us at GitHub.
 389
 390 * "git stash push <pathspec>" did not work from a subdirectory at all.
 391   Bugfix for a topic in v2.13
 392
 393 * As there is no portable way to pass timezone information to
 394   strftime, some output format from "git log" and friends are
 395   impossible to produce.  Teach our own strbuf_addftime to replace %z
 396   and %Z with caller-supplied values to help working around this.
 397   (merge 6eced3ec5e rs/strbuf-addftime-zZ later to maint).
 398
 399 * "git mergetool" learned to work around a wrapper MacOS X adds
 400   around underlying meld.
 401
 402 * An example in documentation that does not work in multi worktree
 403   configuration has been corrected.
 404
 405 * The pretty-format specifiers like '%h', '%t', etc. had an
 406   optimization that no longer works correctly.  In preparation/hope
 407   of getting it correctly implemented, first discard the optimization
 408   that is broken.
 409
 410 * The code to pick up and execute command alias definition from the
 411   configuration used to switch to the top of the working tree and
 412   then come back when the expanded alias was executed, which was
 413   unnecessarilyl complex.  Attempt to simplify the logic by using the
 414   early-config mechanism that does not chdir around.
 415
 416 * Fix configuration codepath to pay proper attention to commondir
 417   that is used in multi-worktree situation, and isolate config API
 418   into its own header file.
 419   (merge dc8441fdb4 bw/config-h later to maint).
 420
 421 * "git add -p" were updated in 2.12 timeframe to cope with custom
 422   core.commentchar but the implementation was buggy and a
 423   metacharacter like $ and * did not work.
 424
 425 * A recent regression in "git rebase -i" has been fixed and tests
 426   that would have caught it and others have been added.
 427
 428 * An unaligned 32-bit access in pack-bitmap code ahs been corrected.
 429
 430 * Tighten error checks for invalid "git apply" input.
 431
 432 * The split index code did not honor core.sharedrepository setting
 433   correctly.
 434
 435 * The Makefile rule in contrib/subtree for building documentation
 436   learned to honour USE_ASCIIDOCTOR just like the main documentation
 437   set does.
 438
 439 * Update the sha1dc again to fix portability glitches.
 440
 441 * Code clean-up to fix possible buffer over-reading.
 442   (merge 2d105451c0 rs/apply-avoid-over-reading later to maint).
 443
 444 * A few tests that tried to verify the contents of push certificates
 445   did not use 'git rev-parse' to formulate the line to look for in
 446   the certificate correctly.
 447
 448 * Update the character width tables.
 449   (merge 7560aacd7c bb/unicode-10.0 later to maint).
 450
 451 * After "git branch --move" of the currently checked out branch, the
 452   code to walk the reflog of HEAD via "log -g" and friends
 453   incorrectly stopped at the reflog entry that records the renaming
 454   of the branch.
 455
 456 * The rewrite of "git branch --list" using for-each-ref's internals
 457   that happened in v2.13 regressed its handling of color.branch.local;
 458   this has been fixed.
 459
 460 * The build procedure has been improved to allow building and testing
 461   Git with address sanitizer more easily.
 462   (merge 566cf0b3bd jk/build-with-asan later to maint).
 463
 464 * Other minor doc, test and build updates and code cleanups.
 465   (merge 3f9c637ec7 pw/unquote-path-in-git-pm later to maint).
 466   (merge 5053313562 rs/urlmatch-cleanup later to maint).
 467   (merge 42c78a216e rs/use-div-round-up later to maint).
 468   (merge 5e8d2729ae rs/wt-status-cleanup later to maint).
 469   (merge 01826066b0 ks/fix-rebase-doc-picture later to maint).