1Git v2.3.2 Release Notes 2======================== 3 4Fixes since v2.3.1 5------------------ 6 7 * "update-index --refresh" used to leak when an entry cannot be 8 refreshed for whatever reason. 9 10 * "git fast-import" used to crash when it could not close and 11 conclude the resulting packfile cleanly. 12 13 * "git blame" died, trying to free an uninitialized piece of memory. 14 15 * "git merge-file" did not work correctly in a subdirectory. 16 17 * "git submodule add" failed to squash "path/to/././submodule" to 18 "path/to/submodule". 19 20 * In v2.2.0, we broke "git prune" that runs in a repository that 21 borrows from an alternate object store. 22 23 * Certain older vintages of cURL give irregular output from 24 "curl-config --vernum", which confused our build system. 25 26 * An earlier workaround to squelch unhelpful deprecation warnings 27 from the complier on Mac OSX unnecessarily set minimum required 28 version of the OS, which the user might want to raise (or lower) 29 for other reasons. 30 31 * Longstanding configuration variable naming rules has been added to 32 the documentation. 33 34 * The credential helper for Windows (in contrib/) used to mishandle 35 a user name with an at-sign in it. 36 37 * Older GnuPG implementations may not correctly import the keyring 38 material we prepare for the tests to use. 39 40 * Clarify in the documentation that "remote.<nick>.pushURL" and 41 "remote.<nick>.URL" are there to name the same repository accessed 42 via different transports, not two separate repositories. 43 44 * The pack bitmap support did not build with older versions of GCC. 45 46 * Reading configuration from a blob object, when it ends with a lone 47 CR, use to confuse the configuration parser. 48 49 * We didn't format an integer that wouldn't fit in "int" but in 50 "uintmax_t" correctly. 51 52 * "git push --signed" gave an incorrectly worded error message when 53 the other side did not support the capability. 54 55 * "git fetch" over a remote-helper that cannot respond to "list" 56 command could not fetch from a symbolic reference e.g. HEAD. 57 58 * The insn sheet "git rebase -i" creates did not fully honor 59 core.abbrev settings. 60 61 * The tests that wanted to see that file becomes unreadable after 62 running "chmod a-r file", and the tests that wanted to make sure it 63 is not run as root, we used "can we write into the / directory?" as 64 a cheap substitute, but on some platforms that is not a good 65 heuristics. The tests and their prerequisites have been updated to 66 check what they really require. 67 68Also contains typofixes, documentation updates and trivial code clean-ups.