Roughly deal with ENOSPC condition despite disk being reported as full

(btrfs metadata problem).  Possibly needs more safety features.

1 files changed, 4 insertions, 0 deletions

diff --git a/TODO.md b/TODO.md
index 1743c23..ccb6de5 100755
--- a/TODO.md
+++ b/TODO.md
@@ -1,6 +1,10 @@
 TODO list for byteback
 ======================
 
+* byteback directory should be at the end of the load path, not start
+
+* change default retry parameters to cover 24 hour window
+
 * "backup could not be marked as complete" message unhelpful on client system - why?
 
 * how do we stop backups from crashing server's kernel?