1Git 2.15 Release Notes 2====================== 3 4Backward compatibility notes and other notable changes. 5 6 * Use of an empty string as a pathspec element that is used for 7 'everything matches' is still warned and Git asks users to use a 8 more explicit '.' for that instead. The hope is that existing 9 users will not mind this change, and eventually the warning can be 10 turned into a hard error, upgrading the deprecation into removal of 11 this (mis)feature. That is now scheduled to happen in the upcoming 12 release. 13 14 * Git now avoids blindly falling back to ".git" when the setup 15 sequence said we are _not_ in Git repository. A corner case that 16 happens to work right now may be broken by a call to die("BUG"). 17 We've tried hard to locate such cases and fixed them, but there 18 might still be cases that need to be addressed--bug reports are 19 greatly appreciated. 20 21 22Updates since v2.14 23------------------- 24 25UI, Workflows & Features 26 27 * An example that is now obsolete has been removed from a sample hook, 28 and an old example in it that added a sign-off manually has been 29 improved to use the interpret-trailers command. 30 31 * The advice message given when "git rebase" stops for conflicting 32 changes has been improved. 33 34 * The "rerere-train" script (in contrib/) learned the "--overwrite" 35 option to allow overwriting existing recorded resolutions. 36 (merge ad53bf79aa rg/rerere-train-overwrite later to maint). 37 38 * "git contacts" (in contrib/) now lists the address on the 39 "Reported-by:" trailer to its output, in addition to those on 40 S-o-b: and other trailers, to make it easier to notify (and thank) 41 the original bug reporter. 42 (merge 09ac673788 eb/contacts-reported-by later to maint). 43 44 45Performance, Internal Implementation, Development Support etc. 46 47 * Conversion from uchar[20] to struct object_id continues. 48 49 * Start using selected c99 constructs in small, stable and 50 essentialpart of the system to catch people who care about 51 older compilers that do not grok them. 52 53 * The filter-process interface learned to allow a process with long 54 latency give a "delayed" response. 55 56 * Many uses of comparision callback function the hashmap API uses 57 cast the callback function type when registering it to 58 hashmap_init(), which defeats the compile time type checking when 59 the callback interface changes (e.g. gaining more parameters). 60 The callback implementations have been updated to take "void *" 61 pointers and cast them to the type they expect instead. 62 63 * Because recent Git for Windows do come with a real msgfmt, the 64 build procedure for git-gui has been updated to use it instead of a 65 hand-rolled substitute. 66 (merge 90dbf226ba js/git-gui-msgfmt-on-windows later to maint). 67 68 69Also contains various documentation updates and code clean-ups. 70 71 72Fixes since v2.14 73----------------- 74 75 * "%C(color name)" in the pretty print format always produced ANSI 76 color escape codes, which was an early design mistake. They now 77 honor the configuration (e.g. "color.ui = never") and also tty-ness 78 of the output medium. 79 (merge 11b087adfd jk/ref-filter-colors later to maint). 80 81 * The http.{sslkey,sslCert} configuration variables are to be 82 interpreted as a pathname that honors "~[username]/" prefix, but 83 weren't, which has been fixed. 84 (merge 8d1549643e jc/http-sslkey-and-ssl-cert-are-paths later to maint). 85 86 * Numerous bugs in walking of reflogs via "log -g" and friends have 87 been fixed. 88 (merge de239446b6 jk/reflog-walk later to maint). 89 90 * "git commit" when seeing an totally empty message said "you did not 91 edit the message", which is clearly wrong. The message has been 92 corrected. 93 (merge bc17f35f8c ks/commit-abort-on-empty-message-fix later to maint). 94 95 * When a directory is not readable, "gitweb" fails to build the 96 project list. Work this around by skipping such a directory. 97 (merge 46a13857fc hb/gitweb-project-list later to maint). 98 99 * Some versions of GnuPG fails to kill gpg-agent it auto-spawned 100 and such a left-over agent can interfere with a test. Work it 101 around by attempting to kill one before starting a new test. 102 (merge 29ff1f8f74 st/lib-gpg-kill-stray-agent later to maint). 103 104 * A recently added test for the "credential-cache" helper revealed 105 that EOF detection done around the time the connection to the cache 106 daemon is torn down were flaky. This was fixed by reacting to 107 ECONNRESET and behaving as if we got an EOF. 108 (merge 1f180e5eb9 dl/credential-cache-socket-in-xdg-cache later to maint). 109 110 * "git log --tag=no-such-tag" showed log starting from HEAD, which 111 has been fixed---it now shows nothing. 112 (merge 5d34d1ac06 jk/rev-list-empty-input later to maint). 113 114 * The "tag.pager" configuration variable was useless for those who 115 actually create tag objects, as it interfered with the use of an 116 editor. A new mechanism has been introduced for commands to enable 117 pager depending on what operation is being carried out to fix this, 118 and then "git tag -l" is made to run pager by default. 119 (merge 595d59e2b5 ma/pager-per-subcommand-action later to maint). 120 121 * Other minor doc, test and build updates and code cleanups. 122 (merge 5b114f3bb0 rs/bswap-ubsan-fix later to maint). 123 (merge 168e63554c rs/move-array later to maint). 124 (merge 268ba20110 rs/stat-data-unaligned-reads-fix later to maint). 125 (merge 78e7b98f45 jt/fsck-code-cleanup later to maint). 126 (merge c7b0780545 rs/pack-objects-pbase-cleanup later to maint). 127 (merge c1e860f1dc js/run-process-parallel-api-fix later to maint). 128 (merge 7a40a95eb4 cc/ref-is-hidden-microcleanup later to maint). 129 (merge c0bb6d9cef ah/doc-wserrorhighlight later to maint). 130 (merge edd64ef4f7 dc/fmt-merge-msg-microcleanup later to maint). 131 (merge fa64a2fdbe jt/subprocess-handshake later to maint). 132 (merge 0ba9c9a0fb jb/t8008-cleanup later to maint). 133 (merge a7c28a2161 jt/t1450-fsck-corrupt-packfile later to maint).