1Git 2.11 Release Notes 2====================== 3 4Updates since v2.10 5------------------- 6 7UI, Workflows & Features 8 9 * "git format-patch --cover-letter HEAD^" to format a single patch 10 with a separate cover letter now numbers the output as [PATCH 0/1] 11 and [PATCH 1/1] by default. 12 13 * An incoming "git push" that attempts to push too many bytes can now 14 be rejected by setting a new configuration variable at the receiving 15 end. 16 17 * "git nosuchcommand --help" said "No manual entry for gitnosuchcommand", 18 which was not intuitive, given that "git nosuchcommand" said "git: 19 'nosuchcommand' is not a git command". 20 21 * "git clone --resurse-submodules --reference $path $URL" is a way to 22 reduce network transfer cost by borrowing objects in an existing 23 $path repository when cloning the superproject from $URL; it 24 learned to also peek into $path for presense of corresponding 25 repositories of submodules and borrow objects from there when able. 26 27 * The "git diff --submodule={short,log}" mechanism has been enhanced 28 to allow "--submodule=diff" to show the patch between the submodule 29 commits bound to the superproject. 30 31 * Even though "git hash-objects", which is a tool to take an 32 on-filesystem data stream and put it into the Git object store, 33 allowed to perform the "outside-world-to-Git" conversions (e.g. 34 end-of-line conversions and application of the clean-filter), and 35 it had the feature on by default from very early days, its reverse 36 operation "git cat-file", which takes an object from the Git object 37 store and externalize for the consumption by the outside world, 38 lacked an equivalent mechanism to run the "Git-to-outside-world" 39 conversion. The command learned the "--filters" option to do so. 40 41 42Performance, Internal Implementation, Development Support etc. 43 44 * The delta-base-cache mechanism has been a key to the performance in 45 a repository with a tightly packed packfile, but it did not scale 46 well even with a larger value of core.deltaBaseCacheLimit. 47 48 * Enhance "git status --porcelain" output by collecting more data on 49 the state of the index and the working tree files, which may 50 further be used to teach git-prompt (in contrib/) to make fewer 51 calls to git. 52 53 * Extract a small helper out of the function that reads the authors 54 script file "git am" internally uses. 55 (merge a77598e jc/am-read-author-file later to maint). 56 57 * Lifts calls to exit(2) and die() higher in the callchain in 58 sequencer.c files so that more helper functions in it can be used 59 by callers that want to handle error conditions themselves. 60 61 * "git am" has been taught to make an internal call to "git apply"'s 62 innards without spawning the latter as a separate process. 63 64 * The ref-store abstraction was introduced to the refs API so that we 65 can plug in different backends to store references. 66 67 * The "unsigned char sha1[20]" to "struct object_id" conversion 68 continues. Notable changes in this round includes that ce->sha1, 69 i.e. the object name recorded in the cache_entry, turns into an 70 object_id. 71 72 * JGit can show a fake ref "capabilities^{}" to "git fetch" when it 73 does not advertise any refs, but "git fetch" was not prepared to 74 see such an advertisement. When the other side disconnects without 75 giving any ref advertisement, we used to say "there may not be a 76 repository at that URL", but we may have seen other advertisement 77 like "shallow" and ".have" in which case we definitely know that a 78 repository is there. The code to detect this case has also been 79 updated. 80 81 * Some codepaths in "git pack-objects" were not ready to use an 82 existing pack bitmap; now they are and as the result they have 83 become faster. 84 85 86Also contains various documentation updates and code clean-ups. 87 88 89Fixes since v2.10 90----------------- 91 92Unless otherwise noted, all the fixes since v2.9 in the maintenance 93track are contained in this release (see the maintenance releases' 94notes for details). 95 96 * Clarify various ways to specify the "revision ranges" in the 97 documentation. 98 99 * "diff-highlight" script (in contrib/) learned to work better with 100 "git log -p --graph" output. 101 102 * The test framework left the number of tests and success/failure 103 count in the t/test-results directory, keyed by the name of the 104 test script plus the process ID. The latter however turned out not 105 to serve any useful purpose. The process ID part of the filename 106 has been removed. 107 108 * Having a submodule whose ".git" repository is somehow corrupt 109 caused a few commands that recurse into submodules loop forever. 110 111 * "git symbolic-ref -d HEAD" happily removes the symbolic ref, but 112 the resulting repository becomes an invalid one. Teach the command 113 to forbid removal of HEAD. 114 115 * A test spawned a short-lived background process, which sometimes 116 prevented the test directory from getting removed at the end of the 117 script on some platforms. 118 119 * Update a few tests that used to use GIT_CURL_VERBOSE to use the 120 newer GIT_TRACE_CURL. 121 122 * "git pack-objects --include-tag" was taught that when we know that 123 we are sending an object C, we want a tag B that directly points at 124 C but also a tag A that points at the tag B. We used to miss the 125 intermediate tag B in some cases. 126 (merge b773dde jk/pack-tag-of-tag later to maint). 127 128 * Update Japanese translation for "git-gui". 129 (merge 02748bc sy/git-gui-i18n-ja later to maint). 130 131 * "git fetch http::/site/path" did not die correctly and segfaulted 132 instead. 133 (merge d63ed6e jk/fix-remote-curl-url-wo-proto later to maint). 134 135 * "git commit-tree" stopped reading commit.gpgsign configuration 136 variable that was meant for Porcelain "git commit" in Git 2.9; we 137 forgot to update "git gui" to look at the configuration to match 138 this change. 139 (merge f14a310 js/git-gui-commit-gpgsign later to maint). 140 141 * "git add --chmod=+x" added recently lacked documentation, which has 142 been corrected. 143 (merge 7ef7903 et/add-chmod-x later to maint). 144 145 * "git log --cherry-pick" used to include merge commits as candidates 146 to be matched up with other commits, resulting a lot of wasted time. 147 The patch-id generation logic has been updated to ignore merges to 148 avoid the wastage. 149 (merge 7c81040 jk/patch-ids-no-merges later to maint). 150 151 * The http transport (with curl-multi option, which is the default 152 these days) failed to remove curl-easy handle from a curlm session, 153 which led to unnecessary API failures. 154 (merge 2abc848 ew/http-do-not-forget-to-call-curl-multi-remove-handle later to maint). 155 156 * There were numerous corner cases in which the configuration files 157 are read and used or not read at all depending on the directory a 158 Git command was run, leading to inconsistent behaviour. The code 159 to set-up repository access at the beginning of a Git process has 160 been updated to fix them. 161 (merge 4d0efa1 jk/setup-sequence-update later to maint). 162 163 * "git diff -W" output needs to extend the context backward to 164 include the header line of the current function and also forward to 165 include the body of the entire current function up to the header 166 line of the next one. This process may have to merge to adjacent 167 hunks, but the code forgot to do so in some cases. 168 (merge 45d2f75 rs/xdiff-merge-overlapping-hunks-for-W-context later to maint). 169 170 * Performance tests done via "t/perf" did not use the same set of 171 build configuration if the user relied on autoconf generated 172 configuration. 173 (merge cd5c281 ks/perf-build-with-autoconf later to maint). 174 175 * "git format-patch --base=..." feature that was recently added 176 showed the base commit information after "-- " e-mail signature 177 line, which turned out to be inconvenient. The base information 178 has been moved above the signature line. 179 (merge 480871e jt/format-patch-base-info-above-sig later to maint). 180 181 * More i18n. 182 (merge 43073f8 va/i18n later to maint). 183 184 * Even when "git pull --rebase=preserve" (and the underlying "git 185 rebase --preserve") can complete without creating any new commit 186 (i.e. fast-forwards), it still insisted on having a usable ident 187 information (read: user.email is set correctly), which was less 188 than nice. As the underlying commands used inside "git rebase" 189 would fail with a more meaningful error message and advice text 190 when the bogus ident matters, this extra check was removed. 191 (merge 1e461c4 jk/rebase-i-drop-ident-check later to maint). 192 193 * "git gc --aggressive" used to limit the delta-chain length to 250, 194 which is way too deep for gaining additional space savings and is 195 detrimental for runtime performance. The limit has been reduced to 196 50. 197 (merge 07e7dbf jk/reduce-gc-aggressive-depth later to maint). 198 199 * Other minor doc, test and build updates and code cleanups. 200 (merge e78d57e bw/pathspec-remove-unused-extern-decl later to maint). 201 (merge ce25e4c rs/checkout-some-states-are-const later to maint). 202 (merge a8342a4 rs/strbuf-remove-fix later to maint). 203 (merge b56aa5b rs/unpack-trees-reduce-file-scope-global later to maint). 204 (merge 5efc60c mr/vcs-svn-printf-ulong later to maint).