From c2a5f3c44fbd2da0d8794229f3aa94c2f5095f9d Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Tue, 15 Dec 2020 14:05:54 -0400 Subject: [PATCH] followup --- ..._8e35d10c208d0d1e93254a1a275fcba6._comment | 20 +++++++++++++++++++ 1 file changed, 20 insertions(+) create mode 100644 doc/forum/What_is_going_on_here__63__/comment_5_8e35d10c208d0d1e93254a1a275fcba6._comment diff --git a/doc/forum/What_is_going_on_here__63__/comment_5_8e35d10c208d0d1e93254a1a275fcba6._comment b/doc/forum/What_is_going_on_here__63__/comment_5_8e35d10c208d0d1e93254a1a275fcba6._comment new file mode 100644 index 0000000000..9a25ac9a23 --- /dev/null +++ b/doc/forum/What_is_going_on_here__63__/comment_5_8e35d10c208d0d1e93254a1a275fcba6._comment @@ -0,0 +1,20 @@ +[[!comment format=mdwn + username="joey" + subject="""comment 5""" + date="2020-12-15T17:53:40Z" + content=""" +Seems that 5.20151208 (I tried it) attempts to get content from the remote +specified with --from even though the git-annex branch's location log +doesn't say the remote contains that content. It stats the expected +location on the remote, notices there's a file there, and ignores the +location log. + +That was changed in [[!commit cd1676d6048ff7307518df94ea22a23e711be29d]], +and the change had a good reason motivating it. It would be possible to add +back a check of Remote.hasKeyCheap when the location log doesn't think it +has a copy, but that would make every copy from a directory remote slower +to handle this edge case and I don't think that's warranted. + +As noted, annex-speculate-present does work if you really need that for +some reason. +" ""]]