From fa15580d1a55aef612a4a973e74e1728c371e98f Mon Sep 17 00:00:00 2001 From: jimparis Date: Tue, 16 Aug 2016 15:14:37 +0000 Subject: [PATCH] --- ...34__git_add__34___vs___34__git_annex_add__34___in_v6.mdwn | 5 +++++ 1 file changed, 5 insertions(+) create mode 100644 doc/forum/__34__git_add__34___vs___34__git_annex_add__34___in_v6.mdwn diff --git a/doc/forum/__34__git_add__34___vs___34__git_annex_add__34___in_v6.mdwn b/doc/forum/__34__git_add__34___vs___34__git_annex_add__34___in_v6.mdwn new file mode 100644 index 0000000000..493344e52d --- /dev/null +++ b/doc/forum/__34__git_add__34___vs___34__git_annex_add__34___in_v6.mdwn @@ -0,0 +1,5 @@ +Prior to v6, I used `git annex add` for files that should go into the annex, and `git add` for files that should be tracked in git normally. With v6, `git add` silently adds to the annex instead, which surprised me because I didn't expect the file contents to still only reside only on the source computer after a push! I don't have `annex.largefiles` set, but it seems like this is intended behavior. Is there a way to return to v5 behavior for those two commands? Is it safe to just remove the smudge/clean filter? If so, is there a way to remove it in such a way that it propagates to all repositories? (It's currently set in `.git/info/attributes`, which overrides `.gitattributes`). + +Having to come up a generic rule for `annex.largefiles` to cover all of my distinctions is difficult, and it's tedious to re-specify it for each new file. I'd much rather just be able to continue using git as usual, and explicitly choose when to invoke git-annex. + +Jim