comment
authorJoey Hess <joeyh@joeyh.name>
Thu, 18 Jan 2024 16:16:48 +0000 (12:16 -0400)
committerJoey Hess <joeyh@joeyh.name>
Thu, 18 Jan 2024 16:16:48 +0000 (12:16 -0400)
doc/submodules/comment_9_5e3a7d5c81fb37efbfe22cbbe8b8a247._comment [new file with mode: 0644]

diff --git a/doc/submodules/comment_9_5e3a7d5c81fb37efbfe22cbbe8b8a247._comment b/doc/submodules/comment_9_5e3a7d5c81fb37efbfe22cbbe8b8a247._comment
new file mode 100644 (file)
index 0000000..7cae800
--- /dev/null
@@ -0,0 +1,16 @@
+[[!comment format=mdwn
+ username="joey"
+ subject="""comment 9"""
+ date="2024-01-18T16:14:11Z"
+ content="""
+Datalad relies on git-annex's handling of submodules to work afaik so I
+don't see why using it would avoid this problem.
+
+If I understand correctly, the .git symlink in the submodule that git-annex
+sets up gets broken when the submodule is relocated. But I would have
+expected that moving from `a/b/` to `x/y/` would keep the symlink working,
+since it's at the same path depth.
+
+I'd need more information to fix this problem, so I recommend filing a
+[[bugs|bug]] report with details.
+"""]]