From 901a4dc9a729829f4f5cdd0d212855c16b47cfef Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Fri, 24 Apr 2015 13:32:59 -0400 Subject: [PATCH] comment --- ...8_d8620ce7b3dbb81c0d3d0b09ded1deb0._comment | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) create mode 100644 doc/forum/Adding_files_with_wildcard_on_Mac_Yosemite/comment_8_d8620ce7b3dbb81c0d3d0b09ded1deb0._comment diff --git a/doc/forum/Adding_files_with_wildcard_on_Mac_Yosemite/comment_8_d8620ce7b3dbb81c0d3d0b09ded1deb0._comment b/doc/forum/Adding_files_with_wildcard_on_Mac_Yosemite/comment_8_d8620ce7b3dbb81c0d3d0b09ded1deb0._comment new file mode 100644 index 0000000000..ac95deb7fe --- /dev/null +++ b/doc/forum/Adding_files_with_wildcard_on_Mac_Yosemite/comment_8_d8620ce7b3dbb81c0d3d0b09ded1deb0._comment @@ -0,0 +1,18 @@ +[[!comment format=mdwn + username="joey" + subject="""comment 8""" + date="2015-04-24T17:26:01Z" + content=""" +The old behavior was accidental, and was never documented. When you use +such undocumented behaviors, you're taking the risk of bugfixes breaking +things. It's not fair to call that a "regression". If I had to worry about +every bugfix breaking users who relied on the old buggy behavior in some +way, I could just stop working now. + +It might be reasonable to add the "expand wildcards rather than letting the +shell do it" feature, as an option. Of course, it would need to be tested +for every git-annex command and problems like the one that caused this bug +to be noticed in the first place dealt with, for every git-annex command. +Using --include and --exclude, which already work seems pretty reasonable +instead. +"""]]