]> git.ipfire.org Git - thirdparty/git.git/blob - Documentation/RelNotes/2.28.0.txt
b54c6a00d98594c86eb74a8f08512b0ff97d6fd9
[thirdparty/git.git] / Documentation / RelNotes / 2.28.0.txt
1 Git 2.28 Release Notes
2 ======================
3
4 Updates since v2.27
5 -------------------
6
7 Backward compatibility notes
8
9 * "feature.experimental" configuration variable is to let volunteers
10 easily opt into a set of newer features, which use of the v2
11 transport protocol is now a part of.
12
13
14 UI, Workflows & Features
15
16 * The commands in the "diff" family learned to honor "diff.relative"
17 configuration variable.
18
19 * The check in "git fsck" to ensure that the tree objects are sorted
20 still had corner cases it missed unsorted entries.
21
22 * The interface to redact sensitive information in the trace output
23 has been simplified.
24
25 * The command line completion (in contrib/) learned to complete
26 options that the "git switch" command takes.
27
28 * "git diff" used to take arguments in random and nonsense range
29 notation, e.g. "git diff A..B C", "git diff A..B C...D", etc.,
30 which has been cleaned up.
31
32
33 Performance, Internal Implementation, Development Support etc.
34
35 * Code optimization for a common case.
36 (merge 8777616e4d an/merge-single-strategy-optim later to maint).
37
38 * We've adopted a convention that any on-stack structure can be
39 initialized to have zero values in all fields with "= { 0 }",
40 even when the first field happens to be a pointer, but sparse
41 complained that a null pointer should be spelled NULL for a long
42 time. Start using -Wno-universal-initializer option to squelch
43 it (the latest sparse has it on by default).
44
45 * "git log -L..." now takes advantage of the "which paths are touched
46 by this commit?" info stored in the commit-graph system.
47
48 * As FreeBSD is not the only platform whose regexp library reports
49 a REG_ILLSEQ error when fed invalid UTF-8, add logic to detect that
50 automatically and skip the affected tests.
51
52 * "git bugreport" learns to report what shell is in use.
53
54 * Support for GIT_CURL_VERBOSE has been rewritten in terms of
55 GIT_TRACE_CURL.
56
57 * Preliminary clean-ups around refs API, plus file format
58 specification documentation for the reftable backend.
59
60 * Workaround breakage in MSVC build, where "curl-config --cflags"
61 gives settings appropriate for GCC build.
62
63 * Code clean-up of "git clean" resulted in a fix of recent
64 performance regression.
65
66 * Code clean-up in the codepath that serves "git fetch" continues.
67
68 * "git merge-base --is-ancestor" is taught to take advantage of the
69 commit graph.
70
71 * Rewrite of parts of the scripted "git submodule" Porcelain command
72 continues; this time it is "git submodule set-branch" subcommand's
73 turn.
74
75 * The "fetch/clone" protocol has been updated to allow the server to
76 instruct the clients to grab pre-packaged packfile(s) in addition
77 to the packed object data coming over the wire.
78
79
80 Fixes since v2.27
81 -----------------
82
83 * The "--prepare-p4-only" option of "git p4" is supposed to stop
84 after replaying one changeset, but kept going (by mistake?)
85
86 * The error message from "git checkout -b foo -t bar baz" was
87 confusing.
88
89 * Some repositories in the wild have commits that record nonsense
90 committer timezone (e.g. rails.git); "git fast-import" learned an
91 option to pass these nonsense timestamps intact to allow recreating
92 existing repositories as-is.
93 (merge d42a2fb72f en/fast-import-looser-date later to maint).
94
95 * The command line completion script (in contrib/) tried to complete
96 "git stash -p" as if it were "git stash push -p", but it was too
97 aggressive and also affected "git stash show -p", which has been
98 corrected.
99 (merge fffd0cf520 vs/complete-stash-show-p-fix later to maint).
100
101 * On-the-wire protocol v2 easily falls into a deadlock between the
102 remote-curl helper and the fetch-pack process when the server side
103 prematurely throws an error and disconnects. The communication has
104 been updated to make it more robust.
105
106 * "git checkout -p" did not handle a newly added path at all.
107 (merge 2c8bd8471a js/checkout-p-new-file later to maint).
108
109 * The code to parse "git bisect start" command line was lax in
110 validating the arguments.
111 (merge 4d9005ff5d cb/bisect-helper-parser-fix later to maint).
112
113 * Reduce memory usage during "diff --quiet" in a worktree with too
114 many stat-unmatched paths.
115 (merge d2d7fbe129 jk/diff-memuse-optim-with-stat-unmatch later to maint).
116
117 * The reflog entries for "git clone" and "git fetch" did not
118 anonymize the URL they operated on.
119 (merge 46da295a77 js/reflog-anonymize-for-clone-and-fetch later to maint).
120
121 * The behaviour of "sparse-checkout" in the state "git clone
122 --no-checkout" left was changed accidentally in 2.27, which has
123 been corrected.
124
125 * Use of negative pathspec, while collecting paths including
126 untracked ones in the working tree, was broken.
127
128 * The same worktree directory must be registered only once, but
129 "git worktree move" allowed this invariant to be violated, which
130 has been corrected.
131 (merge 810382ed37 es/worktree-duplicate-paths later to maint).
132
133 * The effect of sparse checkout settings on submodules is documented.
134 (merge e7d7c73249 en/sparse-with-submodule-doc later to maint).
135
136 * Code clean-up around "git branch" with a minor bugfix.
137 (merge dc44639904 dl/branch-cleanup later to maint).
138
139 * A branch name used in a test has been clarified to match what is
140 going on.
141 (merge 08dc26061f pb/t4014-unslave later to maint).
142
143 * Other code cleanup, docfix, build fix, etc.
144 (merge 2c31a7aa44 jx/pkt-line-doc-count-fix later to maint).
145 (merge d63ae31962 cb/t5608-cleanup later to maint).
146 (merge 788db145c7 dl/t-readme-spell-git-correctly later to maint).
147 (merge 45a87a83bb dl/python-2.7-is-the-floor-version later to maint).
148 (merge b75a219904 es/advertise-contribution-doc later to maint).