|
| 1 | +Git v1.8.4.1 Release Notes |
| 2 | +======================== |
| 3 | + |
| 4 | +Fixes since v1.8.4 |
| 5 | +------------------ |
| 6 | + |
| 7 | + * Some people still use rather old versions of bash, which cannot |
| 8 | + grok some constructs like 'printf -v varname' the prompt and |
| 9 | + completion code started to use recently. The completion and |
| 10 | + prompt scripts have been adjusted to work better with these old |
| 11 | + versions of bash. |
| 12 | + |
| 13 | + * "git rebase -i" had a minor bug (the same could be in other |
| 14 | + programs, as the root cause is pretty generic) where the code |
| 15 | + feeds a random, data dependeant string to 'echo' and expects it |
| 16 | + to come out literally. |
| 17 | + |
| 18 | + * "submodule.<name>.path" variable mistakenly set to the empty |
| 19 | + "true" caused the configuration parser to segfault. |
| 20 | + |
| 21 | + * Output from "git log --full-diff -- <pathspec>" looked strange, |
| 22 | + because comparison was done with the previous ancestor that |
| 23 | + touched the specified <pathspec>, causing the patches for paths |
| 24 | + outside the pathspec to show more than the single commit has |
| 25 | + changed. |
| 26 | + |
| 27 | + * The auto-tag-following code in "git fetch" tries to reuse the |
| 28 | + same transport twice when the serving end does not cooperate and |
| 29 | + does not give tags that point to commits that are asked for as |
| 30 | + part of the primary transfer. Unfortunately, Git-aware transport |
| 31 | + helper interface is not designed to be used more than once, hence |
| 32 | + this did not work over smart-http transfer. Fixed. |
| 33 | + |
| 34 | + * Send a large request to read(2)/write(2) as a smaller but still |
| 35 | + reasonably large chunks, which would improve the latency when the |
| 36 | + operation needs to be killed and incidentally works around broken |
| 37 | + 64-bit systems that cannot take a 2GB write or read in one go. |
| 38 | + |
| 39 | + * A ".mailmap" file that ends with an incomplete line, when read |
| 40 | + from a blob, was not handled properly. |
| 41 | + |
| 42 | + * The recent "short-cut clone connectivity check" topic broke a |
| 43 | + shallow repository when a fetch operation tries to auto-follow |
| 44 | + tags. |
| 45 | + |
| 46 | + * On platforms with fgetc() and friends defined as macros, |
| 47 | + the configuration parser did not compile. |
| 48 | + |
| 49 | +Also contains a handful of trivial code clean-ups, documentation |
| 50 | +updates, updates to the test suite, etc. |
0 commit comments