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