summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorGravatar tomasz <tomasz@web>2017-09-15 14:33:31 +0000
committerGravatar admin <admin@branchable.com>2017-09-15 14:33:31 +0000
commit25dddc64e6554ae2dc91c7e9a7ce430e14acc1ef (patch)
tree497b6589ddff86e6d38519821f2cc4a3e091c50f
parent5255e4ca708442c729a93c816ec58d9007061aae (diff)
-rw-r--r--doc/forum/Syncing_with___39__include__39___rule_for_duplicate_files.mdwn17
1 files changed, 17 insertions, 0 deletions
diff --git a/doc/forum/Syncing_with___39__include__39___rule_for_duplicate_files.mdwn b/doc/forum/Syncing_with___39__include__39___rule_for_duplicate_files.mdwn
new file mode 100644
index 000000000..6121691b2
--- /dev/null
+++ b/doc/forum/Syncing_with___39__include__39___rule_for_duplicate_files.mdwn
@@ -0,0 +1,17 @@
+I have the following problem. I have file in git annex repo which is in two places in this repo.
+So there are two links to the file in the working tree.
+Let say (for clarity) that its path is: a/foo and b/foo.
+
+And when I do:
+
+ git annex wanted . "include=a/*"
+ git annex sync --content
+
+git-annex downloads the file and then drops it
+(i.e. it try to get a/foo and then drop b/foo).
+
+What should I do to avoid droping the file when include only
+one link to the file?
+
+The only one solution to the problem I've found is to use
+deprecated 'direct' mode, but let's say I want to do this the right way.