Documentation / RelNotes / 2.2.0.txton commit Merge branch 'sb/help-unknown-command-sort-fix' (5d7f49d)
   1Git v2.2 Release Notes
   2======================
   3
   4Updates since v2.1
   5------------------
   6
   7Ports
   8
   9 * Building on older MacOS X systems automatically sets
  10   the necessary NO_APPLE_COMMON_CRYPTO build-time option.
  11
  12
  13UI, Workflows & Features
  14
  15 * "git config --edit --global" starts from a skeletal per-user
  16   configuration file contents, instead of a total blank, when the
  17   user does not already have any.  This immediately reduces the
  18   need for a later "Have you forgotten setting core.user?" and we
  19   can add more to the template as we gain more experience.
  20
  21 * "git stash list -p" used to be almost always a no-op because each
  22   stash entry is represented as a merge commit.  It learned to show
  23   the difference between the base commit version and the working tree
  24   version, which is in line with what "git show" gives.
  25
  26 * Sometimes users want to report a bug they experience on their
  27   repository, but they are not at liberty to share the contents of
  28   the repository.  "fast-export" was taught an "--anonymize" option
  29   to replace blob contents, names of people and paths and log
  30   messages with bland and simple strings to help them.
  31
  32 * "log --date=iso" uses a slight variant of ISO 8601 format that is
  33   made more human readable.  A new "--date=iso-strict" option gives
  34   datetime output that is more strictly conformant.
  35
  36 * A broken reimplementation of Git could write an invalid index that
  37   records both stage #0 and higher stage entries for the same path.
  38   We now notice and reject such an index, as there is no sensible
  39   fallback (we do not know if the broken tool wanted to resolve and
  40   forgot to remove higher stage entries, or if it wanted to unresolve
  41   and forgot to remove the stage#0 entry).
  42
  43
  44Performance, Internal Implementation, etc.
  45
  46 * The API to manipulate the "refs" is currently undergoing a revamp
  47   to make it more transactional, with the eventual goal to allow
  48   all-or-none atomic updates and migrating the storage to something
  49   other than the traditional filesystem based one (e.g. databases).
  50
  51 * We no longer attempt to keep track of individual dependencies to
  52   the header files in the build procedure, relying on automated
  53   dependency generation support from modern compilers.
  54
  55 * In tests, we have been using NOT_{MINGW,CYGWIN} test prerequisites
  56   long before negated prerequisites e.g. !MINGW were invented.
  57   The former has been converted to the latter to avoid confusion.
  58
  59 * Looking up remotes configuration in a repository with very many
  60   remotes defined has been optimized.
  61
  62 * There are cases where you lock and open to write a file, close it
  63   to show the updated contents to external processes, and then have
  64   to update the file again while still holding the lock, but the
  65   lockfile API lacked support for such an access pattern.
  66
  67 * The API to allocate the structure to keep track of commit
  68   decoration has been updated to make it less cumbersome to use.
  69
  70 * An in-core caching layer to let us avoid reading the same
  71   configuration files number of times has been added.  A few commands
  72   have been converted to use this subsystem.
  73
  74 * Various code paths have been cleaned up and simplified by using
  75   "strbuf", "starts_with()", and "skip_prefix()" APIs more.
  76
  77 * A few codepaths that died when large blobs that would not fit in
  78   core are involved in their operation have been taught to punt
  79   instead, by e.g. marking too large a blob as not to be diffed.
  80
  81 * A few more code paths in "commit" and "checkout" have been taught
  82   to repopulate the cache-tree in the index, to help speed up later
  83   "write-tree" (used in "commit") and "diff-index --cached" (used in
  84   "status").
  85
  86 * A common programming mistake to assign the same short option name
  87   to two separate options is detected by parse_options() API to help
  88   developers.
  89
  90
  91Also contains various documentation updates and code clean-ups.
  92
  93
  94Fixes since v2.1
  95----------------
  96
  97Unless otherwise noted, all the fixes since v2.1 in the maintenance
  98track are contained in this release (see the maintenance releases'
  99notes for details).
 100
 101 * "git log --pretty/format=" with an empty format string did not
 102   mean the more obvious "No output whatsoever" but "Use default
 103   format", which was counterintuitive.
 104
 105 * Implementations of "tar" that do not understand an extended pax
 106   header would extract the contents of it in a regular file; make
 107   sure the permission bits of this file follows the same tar.umask
 108   configuration setting.
 109
 110 * "git -c section.var command" and "git -c section.var= command"
 111   should pass the configuration differently (the former should be a
 112   boolean true, the latter should be an empty string).
 113
 114 * Applying a patch not generated by Git in a subdirectory used to
 115   check the whitespace breakage using the attributes for incorrect
 116   paths. Also whitespace checks were performed even for paths
 117   excluded via "git apply --exclude=<path>" mechanism.
 118
 119 * "git bundle create" with date-range specification were meant to
 120   exclude tags outside the range, but it didn't.
 121
 122 * "git add x" where x that used to be a directory has become a
 123   symbolic link to a directory misbehaved.
 124
 125 * The prompt script checked $GIT_DIR/ref/stash file to see if there
 126   is a stash, which was a no-no.
 127
 128 * Pack-protocol documentation had a minor typo.
 129
 130 * "git checkout -m" did not switch to another branch while carrying
 131   the local changes forward when a path was deleted from the index.
 132
 133 * With sufficiently long refnames, "git fast-import" could have
 134   overflown an on-stack buffer.
 135
 136 * After "pack-refs --prune" packed refs at the top-level, it failed
 137   to prune them.
 138
 139 * Progress output from "git gc --auto" was visible in "git fetch -q".
 140
 141 * We used to pass -1000 to poll(2), expecting it to also mean "no
 142   timeout", which should be spelled as -1.
 143
 144 * "git rebase" documentation was unclear that it is required to
 145   specify on what <upstream> the rebase is to be done when telling it
 146   to first check out <branch>.
 147   (merge 95c6826 so/rebase-doc later to maint).
 148
 149 * "git push" over HTTP transport had an artificial limit on number of
 150   refs that can be pushed imposed by the command line length.
 151   (merge 26be19b jk/send-pack-many-refspecs later to maint).
 152
 153 * When receiving an invalid pack stream that records the same object
 154   twice, multiple threads got confused due to a race.
 155   (merge ab791dd jk/index-pack-threading-races later to maint).
 156
 157 * An attempt to remove the entire tree in the "git fast-import" input
 158   stream caused it to misbehave.
 159   (merge 2668d69 mb/fast-import-delete-root later to maint).
 160
 161 * Reachability check (used in "git prune" and friends) did not add a
 162   detached HEAD as a starting point to traverse objects still in use.
 163   (merge c40fdd0 mk/reachable-protect-detached-head later to maint).
 164
 165 * "git config --add section.var val" used to lose existing
 166   section.var whose value was an empty string.
 167   (merge c1063be ta/config-add-to-empty-or-true-fix later to maint).
 168
 169 * "git fsck" failed to report that it found corrupt objects via its
 170   exit status in some cases.
 171   (merge 30d1038 jk/fsck-exit-code-fix later to maint).