summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorGravatar Joey Hess <joey@kitenet.net>2011-03-27 12:41:53 -0400
committerGravatar Joey Hess <joey@kitenet.net>2011-03-27 12:41:53 -0400
commit3d84ec1bda71771c5d04c4c8ade37977eecf5a65 (patch)
tree5942be10cbe989b366e2f81034bbf26c58da9512
parent3aaf1fcc549791c19688a8d91cacc7c037506de5 (diff)
response
-rw-r--r--doc/bugs/git-annex_directory_hashing_problems_on_osx.mdwn17
-rw-r--r--doc/bugs/git-annex_has_issues_with_git_when_staging__47__commiting_logs.mdwn6
2 files changed, 23 insertions, 0 deletions
diff --git a/doc/bugs/git-annex_directory_hashing_problems_on_osx.mdwn b/doc/bugs/git-annex_directory_hashing_problems_on_osx.mdwn
index b11cfca42..d1a3c0607 100644
--- a/doc/bugs/git-annex_directory_hashing_problems_on_osx.mdwn
+++ b/doc/bugs/git-annex_directory_hashing_problems_on_osx.mdwn
@@ -12,4 +12,21 @@ $ git add -i
this has somewhat confused git when it tries to stage/merge files, I didn't notice this at first, but it is definately a problem for someone using case insensitive filesystems like the default OSX HFS+ formats or vfat/fat32.
+> I feel a bit stupid to not have considered case-insensative filesystems.
+> They are just so far from where I have lived for 20 years that it's hard
+> to keep them in mind.
+>
+> I guess that
+> [[git-annex_has_issues_with_git_when_staging__47__commiting_logs]] is
+> somehow a consequence (or cause?) of this, but I don't quite understand
+> how this is causing git to fail to stage files, or stage the same file
+> twice under different capitalizations. git-annex always will run git add
+> on the path with the "correct" capitalization. So unless something else
+> has added the path with the other capitalization (perhaps git add
+> .git-annex manually?) I don't understand how you get to this state.
+> --[[Joey]]
+
Also I came across this when I accidentally annexed some files in the .git-annex directory and it cause git-annex/git to be very unhappy when i pulled the repo to somewhere else. It might be worth teaching git-annex to disallow annex'ing of files inside the .git-annex/.git directories.
+
+> There is a guard against `git annex add .git-annex/foo`, but it doesn't
+> notice `cd .git-annex; git annex add foo`. --[[Joey]]
diff --git a/doc/bugs/git-annex_has_issues_with_git_when_staging__47__commiting_logs.mdwn b/doc/bugs/git-annex_has_issues_with_git_when_staging__47__commiting_logs.mdwn
index d4e2635f5..5c654ff2b 100644
--- a/doc/bugs/git-annex_has_issues_with_git_when_staging__47__commiting_logs.mdwn
+++ b/doc/bugs/git-annex_has_issues_with_git_when_staging__47__commiting_logs.mdwn
@@ -3,3 +3,9 @@ After a series of pretty convoluted copying files around between annex'd repos a
I tried running a *`git rm --cached -f -r *`* then *git add -u .git-annex/* or the usual *git add* then a commit fixes things for me. If I don't do that then my subsequent merges/pulls will fail and result in *no known copies of files* I suspect git-annex might have just touched some file modes and git picked up the changes but got confused since there was no content change. It might also just be a git on OSX thing and it doesn't affect linux/bsd users.
For now it's just a bit of extra work for me when it does occur but it does not seem to occur often.
+
+> What do you mean when you say that git "got wedged"? It hung somehow?
+>
+> If git-annex runs concurrently with another git command that locks
+> the repository its git add of log files can fail.
+> --[[Joey]]