1Git v1.8.4.2 Release Notes 2========================== 3 4Fixes since v1.8.4.1 5-------------------- 6 7 * "git clone" gave some progress messages to the standard output, not 8 to the standard error, and did not allow suppressing them with the 9 "--no-progress" option. 10 11 * "format-patch --from=<whom>" forgot to omit unnecessary in-body 12 from line, i.e. when <whom> is the same as the real author. 13 14 * "git shortlog" used to choke and die when there is a malformed 15 commit (e.g. missing authors); it now simply ignore such a commit 16 and keeps going. 17 18 * "git merge-recursive" did not parse its "--diff-algorithm=" command 19 line option correctly. 20 21 * "git branch --track" had a minor regression in v1.8.3.2 and later 22 that made it impossible to base your local work on anything but a 23 local branch of the upstream repository you are tracking from. 24 25 * "git ls-files -k" needs to crawl only the part of the working tree 26 that may overlap the paths in the index to find killed files, but 27 shared code with the logic to find all the untracked files, which 28 made it unnecessarily inefficient. 29 30 * When there is no sufficient overlap between old and new history 31 during a "git fetch" into a shallow repository, objects that the 32 sending side knows the receiving end has were unnecessarily sent. 33 34 * When running "fetch -q", a long silence while the sender side 35 computes the set of objects to send can be mistaken by proxies as 36 dropped connection. The server side has been taught to send a 37 small empty messages to keep the connection alive. 38 39 * When the webserver responds with "405 Method Not Allowed", "git 40 http-backend" should tell the client what methods are allowed with 41 the "Allow" header. 42 43 * "git cvsserver" computed the permission mode bits incorrectly for 44 executable files. 45 46 * The implementation of "add -i" has a crippling code to work around 47 ActiveState Perl limitation but it by mistake also triggered on Git 48 for Windows where MSYS perl is used. 49 50 * We made sure that we notice the user-supplied GIT_DIR is actually a 51 gitfile, but did not do the same when the default ".git" is a 52 gitfile. 53 54 * When an object is not found after checking the packfiles and then 55 loose object directory, read_sha1_file() re-checks the packfiles to 56 prevent racing with a concurrent repacker; teach the same logic to 57 has_sha1_file(). 58 59 * "git commit --author=$name", when $name is not in the canonical 60 "A. U. Thor <au.thor@example.xz>" format, looks for a matching name 61 from existing history, but did not consult mailmap to grab the 62 preferred author name. 63 64 * The commit object names in the insn sheet that was prepared at the 65 beginning of "rebase -i" session can become ambiguous as the 66 rebasing progresses and the repository gains more commits. Make 67 sure the internal record is kept with full 40-hex object names. 68 69 * "git rebase --preserve-merges" internally used the merge machinery 70 and as a side effect, left merge summary message in the log, but 71 when rebasing, there should not be a need for merge summary. 72 73 * "git rebase -i" forgot that the comment character can be 74 configurable while reading its insn sheet. 75 76Also contains a handful of trivial code clean-ups, documentation 77updates, updates to the test suite, etc.