| 1 | GIT v1.5.6.3 Release Notes |
| 2 | ========================== |
| 3 | |
| 4 | Fixes since v1.5.6.2 |
| 5 | -------------------- |
| 6 | |
| 7 | * Setting core.sharedrepository to traditional "true" value was supposed to make |
| 8 | the repository group writable but should not affect permission for others. |
| 9 | However, since 1.5.6, it was broken to drop permission for others when umask is |
| 10 | 022, making the repository unreadable by others. |
| 11 | |
| 12 | * Setting GIT_TRACE will report spawning of external process via run_command(). |
| 13 | |
| 14 | * Using an object with very deep delta chain pinned memory needed for extracting |
| 15 | intermediate base objects unnecessarily long, leading to excess memory usage. |
| 16 | |
| 17 | * Bash completion script did not notice '--' marker on the command |
| 18 | line and tried the relatively slow "ref completion" even when |
| 19 | completing arguments after one. |
| 20 | |
| 21 | * Registering a non-empty blob racily and then truncating the working |
| 22 | tree file for it confused "racy-git avoidance" logic into thinking |
| 23 | that the path is now unchanged. |
| 24 | |
| 25 | * The section that describes attributes related to git-archive were placed |
| 26 | in a wrong place in the gitattributes(5) manual page. |
| 27 | |
| 28 | * "git am" was not helpful to the users when it detected that the committer |
| 29 | information is not set up properly yet. |
| 30 | |
| 31 | * "git clone" had a leftover debugging fprintf(). |
| 32 | |
| 33 | * "git clone -q" was not quiet enough as it used to and gave object count |
| 34 | and progress reports. |
| 35 | |
| 36 | * "git clone" marked downloaded packfile with .keep; this could be a |
| 37 | good thing if the remote side is well packed but otherwise not, |
| 38 | especially for a project that is not really big. |
| 39 | |
| 40 | * "git daemon" used to call syslog() from a signal handler, which |
| 41 | could raise signals of its own but generally is not reentrant. This |
| 42 | was fixed by restructuring the code to report syslog() after the handler |
| 43 | returns. |
| 44 | |
| 45 | * When "git push" tries to remove a remote ref, and corresponding |
| 46 | tracking ref is missing, we used to report error (i.e. failure to |
| 47 | remove something that does not exist). |
| 48 | |
| 49 | * "git mailinfo" (hence "git am") did not handle commit log messages in a |
| 50 | MIME multipart mail correctly. |
| 51 | |
| 52 | Contains other various documentation fixes. |