--- a/tests/test-inhibit.t	Thu Nov 19 10:47:32 2015 -0800
+++ b/tests/test-inhibit.t	Tue Dec 01 10:25:38 2015 -0800
@@ -429,6 +429,41 @@
   |
   o  12:fe1634cbe235 add cK
   |
+  | o  11:53a94305e133 add cL
+  | |
+  | o  10:ad78ff7d621f add cK
+  | |
+  | | @  9:55c73a90e4b4 add cJ
+  | | |
+  | | | o  7:18214586bf78 add cJ
+  | | |/
+  | | o  6:cf5c4f4554ce add cH
+  | | |
+  | | o  5:5419eb264a33 add cG
+  | | |
+  | | o  4:98065434e5c6 add cE
+  | |/
+  o |  3:2db36d8066ff add cD
+  | |
+  o |  2:7df62a38b9bf add cC
+  | |
+  o |  1:02bcbc3f6e56 add cB
+  |/
+  o  0:54ccbc537fc2 add cA
+  
+
+Performing the rebase made ad78ff7d621f and 53a94305e133 visible, we prune
+them to make the graph more legible. It used to be the case that directaccess
+and inhibit were not making these revisions visible. Both behavior are
+acceptable but to keep the test changes minimal we prune those changesets.
+
+  $ hg strip ad78ff7d621f::
+  2 changesets pruned
+  $ hg log -G
+  o  13:2f7b7704d714 add cL
+  |
+  o  12:fe1634cbe235 add cK
+  |
   | @  9:55c73a90e4b4 add cJ
   | |
   | | o  7:18214586bf78 add cJ