summaryrefslogtreecommitdiff
path: root/doc/todo/does_not_preserve_timestamps
diff options
context:
space:
mode:
authorGravatar Joey Hess <joey@kitenet.net>2014-10-21 16:08:53 -0400
committerGravatar Joey Hess <joey@kitenet.net>2014-10-21 16:08:53 -0400
commit4b7aa8745eae693af2be25688a37f7b71c5e4a0e (patch)
tree7a2cee26afae1f6711c28863f6d86ed44b6c24f8 /doc/todo/does_not_preserve_timestamps
parentd151f17c3d01a1c9e893c8ba36b638fb38239a53 (diff)
move to todo
Diffstat (limited to 'doc/todo/does_not_preserve_timestamps')
-rw-r--r--doc/todo/does_not_preserve_timestamps/comment_1_caf5e5cb17f4d05fff8c2fab661cd93f._comment8
-rw-r--r--doc/todo/does_not_preserve_timestamps/comment_2_c337fca1474b5b78f61ad6f421138ae4._comment12
-rw-r--r--doc/todo/does_not_preserve_timestamps/comment_3_9a3eeddc46e5a420575f00cb47caf703._comment8
-rw-r--r--doc/todo/does_not_preserve_timestamps/comment_4_99b064259fc2e3c6eb83c3da3b2d3bac._comment10
-rw-r--r--doc/todo/does_not_preserve_timestamps/comment_5_c95c8b9bd617830604500213c962fc7a._comment9
-rw-r--r--doc/todo/does_not_preserve_timestamps/comment_6_b99e00d0bc4258c4cb28b544b19ea3b8._comment12
6 files changed, 59 insertions, 0 deletions
diff --git a/doc/todo/does_not_preserve_timestamps/comment_1_caf5e5cb17f4d05fff8c2fab661cd93f._comment b/doc/todo/does_not_preserve_timestamps/comment_1_caf5e5cb17f4d05fff8c2fab661cd93f._comment
new file mode 100644
index 000000000..48ec44d2b
--- /dev/null
+++ b/doc/todo/does_not_preserve_timestamps/comment_1_caf5e5cb17f4d05fff8c2fab661cd93f._comment
@@ -0,0 +1,8 @@
+[[!comment format=mdwn
+ username="http://joeyh.name/"
+ ip="209.250.56.144"
+ subject="comment 1"
+ date="2014-09-23T20:27:25Z"
+ content="""
+The closest git comes to storing a timestamp is the date of the last commit of a file for mtime, and first commit for ctime. However, those are pretty expensive to look up for a given file. And git doesn't try to preserve timestamps in checkouts at all, which argues that git-annex, at least at the command line, should not either.
+"""]]
diff --git a/doc/todo/does_not_preserve_timestamps/comment_2_c337fca1474b5b78f61ad6f421138ae4._comment b/doc/todo/does_not_preserve_timestamps/comment_2_c337fca1474b5b78f61ad6f421138ae4._comment
new file mode 100644
index 000000000..4b5a750cf
--- /dev/null
+++ b/doc/todo/does_not_preserve_timestamps/comment_2_c337fca1474b5b78f61ad6f421138ae4._comment
@@ -0,0 +1,12 @@
+[[!comment format=mdwn
+ username="https://www.google.com/accounts/o8/id?id=AItOawmK0703vNSIQsP1mGf-4MAPnsBZiSc6yVo"
+ nickname="Emre"
+ subject="comment 2"
+ date="2014-09-23T20:58:10Z"
+ content="""
+Thanks Joey for the comment.
+
+But when syncing two repos, timestamps are critical at least for my use case. I can't lose this info. Even if it's expensive.
+
+Appreciate if you can consider to add it for direct mode repos, ie when a file is synced to another repo and created there, it shall carry at least the mtime of the file in source repo. Owncloud sync does it, btsync does it, although I know git-annex is different than those.
+"""]]
diff --git a/doc/todo/does_not_preserve_timestamps/comment_3_9a3eeddc46e5a420575f00cb47caf703._comment b/doc/todo/does_not_preserve_timestamps/comment_3_9a3eeddc46e5a420575f00cb47caf703._comment
new file mode 100644
index 000000000..ba34823d4
--- /dev/null
+++ b/doc/todo/does_not_preserve_timestamps/comment_3_9a3eeddc46e5a420575f00cb47caf703._comment
@@ -0,0 +1,8 @@
+[[!comment format=mdwn
+ username="https://www.google.com/accounts/o8/id?id=AItOawmK0703vNSIQsP1mGf-4MAPnsBZiSc6yVo"
+ nickname="Emre"
+ subject="comment 3"
+ date="2014-09-23T21:15:29Z"
+ content="""
+Btw, git storing the last commit time as Mtime is not enough, it shall store the original timestamp of the file, not the date of commit. Hope I could explain and hope this is something doable.
+"""]]
diff --git a/doc/todo/does_not_preserve_timestamps/comment_4_99b064259fc2e3c6eb83c3da3b2d3bac._comment b/doc/todo/does_not_preserve_timestamps/comment_4_99b064259fc2e3c6eb83c3da3b2d3bac._comment
new file mode 100644
index 000000000..08de75619
--- /dev/null
+++ b/doc/todo/does_not_preserve_timestamps/comment_4_99b064259fc2e3c6eb83c3da3b2d3bac._comment
@@ -0,0 +1,10 @@
+[[!comment format=mdwn
+ username="http://svario.it/gioele"
+ nickname="gioele"
+ subject="comment 4"
+ date="2014-09-24T07:15:09Z"
+ content="""
+You can try to store the timestamps just before commit and restore them on checkout.
+
+Have a look at [metastore](https://github.com/przemoc/metastore): it is a ready-made solution for plain git. Maybe you can adapt it to git-annex.
+"""]]
diff --git a/doc/todo/does_not_preserve_timestamps/comment_5_c95c8b9bd617830604500213c962fc7a._comment b/doc/todo/does_not_preserve_timestamps/comment_5_c95c8b9bd617830604500213c962fc7a._comment
new file mode 100644
index 000000000..7632548cd
--- /dev/null
+++ b/doc/todo/does_not_preserve_timestamps/comment_5_c95c8b9bd617830604500213c962fc7a._comment
@@ -0,0 +1,9 @@
+[[!comment format=mdwn
+ username="https://www.google.com/accounts/o8/id?id=AItOawlM_DRhi_5pJrTA0HbApHR25iAgy-NBXTY"
+ nickname="Tor Arne"
+ subject="comment 5"
+ date="2014-10-01T22:02:43Z"
+ content="""
+I was just about to start using git-annex for my data when I saw this bug report, which unfortunately is a blocker. Persisting basic metadata of files that are synced seems like a core feature of a file sync/transfer tool, so I'm really hoping this can be solved somehow :/
+
+"""]]
diff --git a/doc/todo/does_not_preserve_timestamps/comment_6_b99e00d0bc4258c4cb28b544b19ea3b8._comment b/doc/todo/does_not_preserve_timestamps/comment_6_b99e00d0bc4258c4cb28b544b19ea3b8._comment
new file mode 100644
index 000000000..67fe25595
--- /dev/null
+++ b/doc/todo/does_not_preserve_timestamps/comment_6_b99e00d0bc4258c4cb28b544b19ea3b8._comment
@@ -0,0 +1,12 @@
+[[!comment format=mdwn
+ username="https://www.google.com/accounts/o8/id?id=AItOawlM_DRhi_5pJrTA0HbApHR25iAgy-NBXTY"
+ nickname="Tor Arne"
+ subject="comment 6"
+ date="2014-10-01T22:39:01Z"
+ content="""
+Isn't this what the metadata feature does though? http://git-annex.branchable.com/design/metadata/
+
+With annex.genmetadata true set, it should store year and mont (but not day/time? if so why not?
+
+Is the missing piece of the puzzle to apply the metadata again on checkout?
+"""]]