Skip to content

Commit 635452a

Browse files
committed
rn-62: add article about regression in Magit
1 parent d6d8af8 commit 635452a

File tree

1 file changed

+53
-2
lines changed

1 file changed

+53
-2
lines changed

‎rev_news/drafts/edition-62.md‎

Lines changed: 53 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -25,11 +25,62 @@ This edition covers what happened during the month of March 2020.
2525
### Reviews
2626
-->
2727

28-
<!---
2928
### Support
30-
-->
29+
30+
*[Regression in v2.26.0-rc0 and Magit](https://lore.kernel.org/git/3091652.KAqcNXvZJ4@cayenne/)
31+
32+
Jean-Noël Avila reported to the mailing list that git version
33+
2.26.0-rc0 segfaulted under Magit with auto-revert enabled.
34+
35+
[Magit](https://magit.vc/) is a popular Emacs interface to Git, and
36+
the [auto-revert mode](https://magit.vc/manual/magit/Automatic-Reverting-of-File_002dVisiting-Buffers.html)
37+
lets Emacs revert file that have changed on disk when a Git command
38+
has been run outside of Emacs.
39+
40+
Jean-Noël had bisected the issue to a commit that was improving the
41+
error message that Git gave when it dies because a path it is passed
42+
is outside the repository. This commit though didn't considered the
43+
case of a bare repo which triggered the segfault.
44+
45+
Jonathan Nieder replied that the bug was fixed by another commit by
46+
Emily Shaffer that had not yet made it to the master branch. He
47+
asked Junio Hamano, the Git maintainer, if the commit could be
48+
fast-tracked and Emily if she could add a test to the commit.
49+
50+
Junio replied that he agreed that a few tests should be added and
51+
that there a few days left before v2.26.0-rc2 to add them. The next
52+
day though Junio replied to himself with a patch adding the tests
53+
and asking for comments.
54+
55+
Jonathan Nieder reviewed Junio's tests adding his "Reviewed-by:",
56+
and said that Emily was out of office so they were well timed.
57+
58+
Junio replied to himself again discussing one test he wrote that
59+
tested that both `git log -- ..` and `git ls-files -- ..` fail when
60+
the current working directory is the `.git` directory.
61+
62+
He wondered why, if "." instead of ".." is used in the above
63+
commands, Git should behave as if the current working directory was
64+
the top-level of the working tree instead of `.git`, and why Magit
65+
is expecting `cd .git && git ls-files ..` to show the entire working
66+
tree.
67+
68+
Kyle Meyer replied to Junio that internally Magit's call that
69+
triggered the bug is running `git ls-files` from `.git` to ask
70+
whether the file used to edit a commit message is actually tracked,
71+
as it makes no distinction between files in .git and in the working
72+
tree. He said that he would propose a change in Magit to improve
73+
this.
74+
75+
Gábor Szeder also replied to Junio's patch suggesting a small
76+
improvement in the tests which Junio accepted sending an improved
77+
patch.
78+
79+
The fix with Emily's code and Junio's tests was then merged into
80+
v2.26.0-rc2.
3181

3282
## Developer Spotlight: Eric S. Raymond
83+
3384
* Who are you and what do you do?
3485

3586
I've been an open-source hacker since long before that term was coined.

0 commit comments

Comments
(0)