| blob | a5e3ce8f5de41ff7eb1e17686be80bf7f9c11775 |
1 Git v2.2 Release Notes
2 ======================
4 Updates since v2.1
5 ------------------
7 Ports
9 * Building on older MacOS X systems automatically sets
10 the necessary NO_APPLE_COMMON_CRYPTO build-time option.
13 UI, Workflows & Features
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.
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.
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.
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.
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).
43 * The "pre-receive" and "post-receive" hooks are no longer required
44 to consume their input fully (not following this requirement used
45 to result in intermittent errors in "git push").
48 Performance, Internal Implementation, etc.
50 * The API to manipulate the "refs" is currently undergoing a revamp
51 to make it more transactional, with the eventual goal to allow
52 all-or-none atomic updates and migrating the storage to something
53 other than the traditional filesystem based one (e.g. databases).
55 * We no longer attempt to keep track of individual dependencies to
56 the header files in the build procedure, relying on automated
57 dependency generation support from modern compilers.
59 * In tests, we have been using NOT_{MINGW,CYGWIN} test prerequisites
60 long before negated prerequisites e.g. !MINGW were invented.
61 The former has been converted to the latter to avoid confusion.
63 * Looking up remotes configuration in a repository with very many
64 remotes defined has been optimized.
66 * There are cases where you lock and open to write a file, close it
67 to show the updated contents to external processes, and then have
68 to update the file again while still holding the lock, but the
69 lockfile API lacked support for such an access pattern.
71 * The API to allocate the structure to keep track of commit
72 decoration has been updated to make it less cumbersome to use.
74 * An in-core caching layer to let us avoid reading the same
75 configuration files number of times has been added. A few commands
76 have been converted to use this subsystem.
78 * Various code paths have been cleaned up and simplified by using
79 "strbuf", "starts_with()", and "skip_prefix()" APIs more.
81 * A few codepaths that died when large blobs that would not fit in
82 core are involved in their operation have been taught to punt
83 instead, by e.g. marking too large a blob as not to be diffed.
85 * A few more code paths in "commit" and "checkout" have been taught
86 to repopulate the cache-tree in the index, to help speed up later
87 "write-tree" (used in "commit") and "diff-index --cached" (used in
88 "status").
90 * A common programming mistake to assign the same short option name
91 to two separate options is detected by parse_options() API to help
92 developers.
94 * The code path to write out the packed-refs file has been optimized,
95 which especially matters in a repository with a large number of
96 refs.
98 * The check to see if a ref $F can be created by making sure no
99 existing ref has $F/ as its prefix has been optimized, which
100 especially matters in a repository with a large number of existing
101 refs.
103 * "git fsck" was taught to check contents of tag objects a bit more.
105 * "git hash-object" was taught a "--literally" option to help
106 debugging.
109 Also contains various documentation updates and code clean-ups.
112 Fixes since v2.1
113 ----------------
115 Unless otherwise noted, all the fixes since v2.1 in the maintenance
116 track are contained in this release (see the maintenance releases'
117 notes for details).
119 * "git log --pretty/format=" with an empty format string did not
120 mean the more obvious "No output whatsoever" but "Use default
121 format", which was counterintuitive.
123 * Implementations of "tar" that do not understand an extended pax
124 header would extract the contents of it in a regular file; make
125 sure the permission bits of this file follows the same tar.umask
126 configuration setting.
128 * "git -c section.var command" and "git -c section.var= command"
129 should pass the configuration differently (the former should be a
130 boolean true, the latter should be an empty string).
132 * Applying a patch not generated by Git in a subdirectory used to
133 check the whitespace breakage using the attributes for incorrect
134 paths. Also whitespace checks were performed even for paths
135 excluded via "git apply --exclude=<path>" mechanism.
137 * "git bundle create" with date-range specification were meant to
138 exclude tags outside the range, but it didn't.
140 * "git add x" where x that used to be a directory has become a
141 symbolic link to a directory misbehaved.
143 * The prompt script checked $GIT_DIR/ref/stash file to see if there
144 is a stash, which was a no-no.
146 * Pack-protocol documentation had a minor typo.
148 * "git checkout -m" did not switch to another branch while carrying
149 the local changes forward when a path was deleted from the index.
151 * With sufficiently long refnames, "git fast-import" could have
152 overflown an on-stack buffer.
154 * After "pack-refs --prune" packed refs at the top-level, it failed
155 to prune them.
157 * Progress output from "git gc --auto" was visible in "git fetch -q".
159 * We used to pass -1000 to poll(2), expecting it to also mean "no
160 timeout", which should be spelled as -1.
162 * "git rebase" documentation was unclear that it is required to
163 specify on what <upstream> the rebase is to be done when telling it
164 to first check out <branch>.
165 (merge 95c6826 so/rebase-doc later to maint).
167 * "git push" over HTTP transport had an artificial limit on number of
168 refs that can be pushed imposed by the command line length.
169 (merge 26be19b jk/send-pack-many-refspecs later to maint).
171 * When receiving an invalid pack stream that records the same object
172 twice, multiple threads got confused due to a race.
173 (merge ab791dd jk/index-pack-threading-races later to maint).
175 * An attempt to remove the entire tree in the "git fast-import" input
176 stream caused it to misbehave.
177 (merge 2668d69 mb/fast-import-delete-root later to maint).
179 * Reachability check (used in "git prune" and friends) did not add a
180 detached HEAD as a starting point to traverse objects still in use.
181 (merge c40fdd0 mk/reachable-protect-detached-head later to maint).
183 * "git config --add section.var val" used to lose existing
184 section.var whose value was an empty string.
185 (merge c1063be ta/config-add-to-empty-or-true-fix later to maint).
187 * "git fsck" failed to report that it found corrupt objects via its
188 exit status in some cases.
189 (merge 30d1038 jk/fsck-exit-code-fix later to maint).
191 * Use of "--verbose" option used to break "git branch --merged".
192 (merge 12994dd jk/maint-branch-verbose-merged later to maint).