aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorGravatar zardoz <zardoz@web>2014-05-13 11:59:56 +0000
committerGravatar admin <admin@branchable.com>2014-05-13 11:59:56 +0000
commit0a539df674800c8e24a3f235b2b2c23014f459c3 (patch)
treeec150103ae33bdea0e2f38f7b4ae1d9e2c85d6c9
parent7c0fc5d1baea8b0c00df9b106785b5076d53e010 (diff)
-rw-r--r--doc/bugs/git_annex_list__47__whereis_and_uncommited_local_changes.mdwn28
1 files changed, 28 insertions, 0 deletions
diff --git a/doc/bugs/git_annex_list__47__whereis_and_uncommited_local_changes.mdwn b/doc/bugs/git_annex_list__47__whereis_and_uncommited_local_changes.mdwn
new file mode 100644
index 000000000..1af31415a
--- /dev/null
+++ b/doc/bugs/git_annex_list__47__whereis_and_uncommited_local_changes.mdwn
@@ -0,0 +1,28 @@
+### Please describe the problem.
+
+I’m not sure whether this is a bug or a technical limitation, but here
+you go: In direct mode, if a file is locally changed but unsynced,
+«git annex whereis» and «git annex list» behave as if the file were
+still unchanged. Any changes would only show after «git annex sync» or
+«git annex add» (altough I’m not sure what the latter does if a file
+is already annexed).
+
+Syncing prior to «whereis» is not viable since I might not want to
+proliferate the changes to other remotes. Would it be reasonable making
+sure that «whereis» considers uncommitted local changes?
+
+If doing so is too expensive to do implicitly, it would be useful
+having an explicit means of seeing local changes in direct mode
+repositories. Is «git annex add» the intended way of doing so?
+
+### What steps will reproduce the problem?
+
+1. Clone A to B.
+2. Get file $f from A to B.
+3. Switch B to direct mode.
+4. Edit file $f.
+5. «git annex list» reports that $f is on both A and B.
+
+### What version of git-annex are you using? On what operating system?
+git-annex 5.20140421
+Linux 3.14.3