summaryrefslogtreecommitdiff
path: root/doc/bugs
diff options
context:
space:
mode:
Diffstat (limited to 'doc/bugs')
-rw-r--r--doc/bugs/Drop_--from_always_trusts_local_repository.mdwn44
-rw-r--r--doc/bugs/assistant_eats_all_CPU.mdwn7
-rw-r--r--doc/bugs/assistant_eats_all_CPU/comment_23_48a4c8d9dcc6cec243c6072090f26b6d._comment12
-rw-r--r--doc/bugs/git-annex_fails_to_initialize_under_Windows/comment_2_b9a3a0104bc56f9110fc58c9df140f12._comment34
-rw-r--r--doc/bugs/git_annex_sync_in_direct_mode_does_not_honor_skip-worktree/comment_5_cb98789c50c58f01055183dbaf7b4eba._comment11
5 files changed, 108 insertions, 0 deletions
diff --git a/doc/bugs/Drop_--from_always_trusts_local_repository.mdwn b/doc/bugs/Drop_--from_always_trusts_local_repository.mdwn
new file mode 100644
index 000000000..a1b27cf5b
--- /dev/null
+++ b/doc/bugs/Drop_--from_always_trusts_local_repository.mdwn
@@ -0,0 +1,44 @@
+### Please describe the problem.
+
+The command `git annex drop --from` always trusts the local repository, even if
+it is marked as untrusted.
+
+
+### What steps will reproduce the problem?
+[[!format sh """
+mkdir t u; cd t; git init; git commit --allow-empty -m "Initial commit"; git annex init "Trusted"; date > file; git annex add file; git commit -m "Add file"; cd ../u; git init; git remote add t ../t; git fetch t; git merge t/master; git annex init "Untrusted"; git annex untrust .; git annex get file; cd ../t; git remote add u ../u; git fetch u; cd ..
+"""]]
+
+Create two repositories, *t* (trusted) and *u* (untrusted). A file is in both
+repositories. When performing `git annex drop file` in repository *t*, `git
+annex` will abort because there are not enough copies. But when performing `git
+annex drop --from t file` in *u*, git annex will delete the copy.
+
+
+### What version of git-annex are you using? On what operating system?
+
+Bug was introduced with 6c31e3a8 and still exists in current master (d955cfe7).
+
+
+### Please provide any additional information below.
+
+The following change seems to solve the problem. (First time working with
+Haskell, please excuse the crude code.)
+
+[[!format diff """
+diff --git a/Command/Drop.hs b/Command/Drop.hs
+index 269c4c2..09ea99a 100644
+--- a/Command/Drop.hs
++++ b/Command/Drop.hs
+@@ -82,8 +82,9 @@ performRemote key afile numcopies remote = lockContent key $ do
+ (remotes, trusteduuids) <- Remote.keyPossibilitiesTrusted key
+ present <- inAnnex key
+ u <- getUUID
++ level <- lookupTrust u
+ let have = filter (/= uuid) $
+- if present then u:trusteduuids else trusteduuids
++ if present && level <= SemiTrusted then u:trusteduuids else trusteduuids
+ untrusteduuids <- trustGet UnTrusted
+ let tocheck = filter (/= remote) $
+ Remote.remotesWithoutUUID remotes (have++untrusteduuids)
+"""]]
diff --git a/doc/bugs/assistant_eats_all_CPU.mdwn b/doc/bugs/assistant_eats_all_CPU.mdwn
index 4939bf456..719fca425 100644
--- a/doc/bugs/assistant_eats_all_CPU.mdwn
+++ b/doc/bugs/assistant_eats_all_CPU.mdwn
@@ -520,3 +520,10 @@ $ ps -O start xf | grep git-annex
13761 23:56:38 Z ? 00:00:00 \_ [git] <defunct>
6252 12:56:59 S ? 00:01:09 /usr/bin/emacs23
"""]]
+
+#### This bug is fixed
+
+> [[fixed|done]]. This was a Cronner bug, triggered when you had a
+> scheduled fsck job that runs monthly at any time, and the last time it ran was on a day of a
+> month > 12. Workaround: Disable scheduled fsck jobs, or change them to
+> run on a specific day of the month. Or upgrade. --[[Joey]]
diff --git a/doc/bugs/assistant_eats_all_CPU/comment_23_48a4c8d9dcc6cec243c6072090f26b6d._comment b/doc/bugs/assistant_eats_all_CPU/comment_23_48a4c8d9dcc6cec243c6072090f26b6d._comment
new file mode 100644
index 000000000..b6d2dec76
--- /dev/null
+++ b/doc/bugs/assistant_eats_all_CPU/comment_23_48a4c8d9dcc6cec243c6072090f26b6d._comment
@@ -0,0 +1,12 @@
+[[!comment format=mdwn
+ username="http://joeyh.name/"
+ ip="209.250.56.191"
+ subject="comment 23"
+ date="2014-04-11T16:32:47Z"
+ content="""
+<maxy> the clock_gettime(0x2 and clock_gettime(0x3 are consistent with getCurrentTime and getTimeZone of nextTime
+
+So, that strongly points to the Cronner thread, and I doubt this is specific to stable at all.
+
+Please run git-annex vicfg, and paste all the \"schedule\" lines, from a repository that has the problem. That should allow me to reproduce and fix this bug.
+"""]]
diff --git a/doc/bugs/git-annex_fails_to_initialize_under_Windows/comment_2_b9a3a0104bc56f9110fc58c9df140f12._comment b/doc/bugs/git-annex_fails_to_initialize_under_Windows/comment_2_b9a3a0104bc56f9110fc58c9df140f12._comment
new file mode 100644
index 000000000..1fcfc5a51
--- /dev/null
+++ b/doc/bugs/git-annex_fails_to_initialize_under_Windows/comment_2_b9a3a0104bc56f9110fc58c9df140f12._comment
@@ -0,0 +1,34 @@
+[[!comment format=mdwn
+ username="ayutheos"
+ ip="49.124.177.13"
+ subject="comment 2"
+ date="2014-04-10T07:52:36Z"
+ content="""
+I'm getting this error too.
+
+ user@NOTEBOOK /d/pictures
+ $ git annex init \"photos\"
+ init photos
+ Detected a filesystem without fifo support.
+
+ Disabling ssh connection caching.
+
+ Detected a crippled filesystem.
+
+ Enabling direct mode.
+ fatal: index file open failed: Invalid argument
+ git-annex: git [Param \"checkout\",Param \"-q\",Param \"-B\",Param \"annex/direct/master\"] failed
+
+git-annex version:
+
+ user@NOTEBOOK /d/pictures
+ $ git annex version
+ git-annex version: 5.20140403-gdfa17fc
+ build flags: Assistant Webapp Webapp-secure Pairing Testsuite S3 WebDAV DNS Feeds Quvi TDFA CryptoHash
+ key/value backends: SHA256E SHA1E SHA512E SHA224E SHA384E SKEIN256E SKEIN512E SHA256 SHA1 SHA512 SHA224 SHA384 SKEIN256 SKEIN512 WORM URL
+ remote types: git gcrypt S3 bup directory rsync web webdav tahoe glacier hook external
+ local repository version: 5
+ supported repository version: 5
+ upgrade supported from repository versions: 2 3 4
+
+"""]]
diff --git a/doc/bugs/git_annex_sync_in_direct_mode_does_not_honor_skip-worktree/comment_5_cb98789c50c58f01055183dbaf7b4eba._comment b/doc/bugs/git_annex_sync_in_direct_mode_does_not_honor_skip-worktree/comment_5_cb98789c50c58f01055183dbaf7b4eba._comment
new file mode 100644
index 000000000..d40d2d789
--- /dev/null
+++ b/doc/bugs/git_annex_sync_in_direct_mode_does_not_honor_skip-worktree/comment_5_cb98789c50c58f01055183dbaf7b4eba._comment
@@ -0,0 +1,11 @@
+[[!comment format=mdwn
+ username="EvanDeaubl"
+ ip="24.251.129.149"
+ subject="comment 5"
+ date="2014-04-09T03:28:24Z"
+ content="""
+I'm afraid I abandoned this patch. It worked, but was still fidgety for being able to ignore parts of the tree. I found another way to do what I wanted by loading an indirect repo into /data and taking advantage of a surprise side effect in how the /sdcard filesystem translated the symlinks from the ext4 filesystem.
+
+I can probably scare it up from my archives, but it hasn't been kept up to date. The good news is (as I recall) the patch was pretty small.
+
+"""]]