1Git v2.0.1 Release Notes 2======================== 3 4 * We used to unconditionally disable the pager in the pager process 5 we spawn to feed out output, but that prevented people who want to 6 run "less" within "less" from doing so. 7 8 * Tools that read diagnostic output in our standard error stream do 9 not want to see terminal control sequence (e.g. erase-to-eol). 10 Detect them by checking if the standard error stream is connected 11 to a tty. 12 * Reworded the error message given upon a failure to open an existing 13 loose object file due to e.g. permission issues; it was reported as 14 the object being corrupt, but that is not quite true. 15 16 * "git log -2master" is a common typo that shows two commits starting 17 from whichever random branch that is not 'master' that happens to 18 be checked out currently. 19 20 * The "%<(10,trunc)%s" pretty format specifier in the log family of 21 commands is used to truncate the string to a given length (e.g. 10 22 in the example) with padding to column-align the output, but did 23 not take into account that number of bytes and number of display 24 columns are different. 25 26 * The "mailmap.file" configuration option did not support the tilde 27 expansion (i.e. ~user/path and ~/path). 28 29 * The completion scripts (in contrib/) did not know about quite a few 30 options that are common between "git merge" and "git pull", and a 31 couple of options unique to "git merge". 32 33 * "--ignore-space-change" option of "git apply" ignored the spaces 34 at the beginning of line too aggressively, which is inconsistent 35 with the option of the same name "diff" and "git diff" have. 36 37 * "git blame" miscounted number of columns needed to show localized 38 timestamps, resulting in jaggy left-side-edge of the source code 39 lines in its output. 40 41 * "git blame" assigned the blame to the copy in the working-tree if 42 the repository is set to core.autocrlf=input and the file used CRLF 43 line endings. 44 45 * "git commit --allow-empty-message -C $commit" did not work when the 46 commit did not have any log message. 47 48 * "git diff --find-copies-harder" sometimes pretended as if the mode 49 bits have changed for paths that are marked with assume-unchanged 50 bit. 51 52 * "git format-patch" did not enforce the rule that the "--follow" 53 option from the log/diff family of commands must be used with 54 exactly one pathspec. 55 56 * "git gc --auto" was recently changed to run in the background to 57 give control back early to the end-user sitting in front of the 58 terminal, but it forgot that housekeeping involving reflogs should 59 be done without other processes competing for accesses to the refs. 60 61 * "git grep -O" to show the lines that hit in the pager did not work 62 well with case insensitive search. We now spawn "less" with its 63 "-I" option when it is used as the pager (which is the default). 64 65 * We used to disable threaded "git index-pack" on platforms without 66 thread-safe pread(); use a different workaround for such 67 platforms to allow threaded "git index-pack". 68 69 * The error reporting from "git index-pack" has been improved to 70 distinguish missing objects from type errors. 71 72 * "git mailinfo" used to read beyond the end of header string while 73 parsing an incoming e-mail message to extract the patch. 74 75 * On a case insensitive filesystem, merge-recursive incorrectly 76 deleted the file that is to be renamed to a name that is the same 77 except for case differences. 78 79 * "git pack-objects" unnecessarily copied the previous contents when 80 extending the hashtable, even though it will populate the table 81 from scratch anyway. 82 83 * "git rerere forget" did not work well when merge.conflictstyle 84 was set to a non-default value. 85 86 * "git remote rm" and "git remote prune" can involve removing many 87 refs at once, which is not a very efficient thing to do when very 88 many refs exist in the packed-refs file. 89 90 * "git log --exclude=<glob> --all | git shortlog" worked as expected, 91 but "git shortlog --exclude=<glob> --all", which is supposed to be 92 identical to the above pipeline, was not accepted at the command 93 line argument parser level. 94 95 * The autostash mode of "git rebase -i" did not restore the dirty 96 working tree state if the user aborted the interactive rebase by 97 emptying the insn sheet. 98 99 * "git show -s" (i.e. show log message only) used to incorrectly emit 100 an extra blank line after a merge commit. 101 102 * "git status", even though it is a read-only operation, tries to 103 update the index with refreshed lstat(2) info to optimize future 104 accesses to the working tree opportunistically, but this could 105 race with a "read-write" operation that modify the index while it 106 is running. Detect such a race and avoid overwriting the index. 107 108 * "git status" (and "git commit") behaved as if changes in a modified 109 submodule are not there if submodule.*.ignore configuration is set, 110 which was misleading. The configuration is only to unclutter diff 111 output during the course of development, and should not to hide 112 changes in the "status" output to cause the users forget to commit 113 them. 114 115 * The mode to run tests with HTTP server tests disabled was broken.