diff --git a/src/index/fsindexer.cpp b/src/index/fsindexer.cpp
index 8234dd7d..9df1ce78 100644
--- a/src/index/fsindexer.cpp
+++ b/src/index/fsindexer.cpp
@@ -48,12 +48,14 @@
 
 
 // When using extended attributes, we have to use the ctime, because
-// this is all that gets set when the attributes are modified. This
-// is quite an expensive price to pay...
+// this is all that gets set when the attributes are modified. 
+// As of 1.19 we use ctime in all cases as this allows to detect a
+// file renamed into an existing file (e.g. when shifting logs or
+// other archives).
 #ifdef RCL_USE_XATTR
 #define RCL_STTIME st_ctime
 #else
-#define RCL_STTIME st_mtime
+#define RCL_STTIME st_ctime
 #endif // RCL_USE_XATTR
 
 using namespace std;
diff --git a/website/release-1.19.html b/website/release-1.19.html
index 36dbe24c..4cd3f368 100644
--- a/website/release-1.19.html
+++ b/website/release-1.19.html
@@ -54,6 +54,11 @@
         that no debris from older releases remain (e.g.: old stemming
         files which are not used any more).</p>
 
+      <p>Installing 1.19 over an 1.18 index will force a lot of
+        reindexing anyway because we switched from using st_ctime instead
+        of st_mtime to detect file modifications, meaning that all files
+        which were modified since created will be updated.</p>
+
       <p><span class="important">Viewer exceptions</span>: 
         There is a new list of mime types that should be opened with
         the locally configured application even when <em>Use
@@ -85,6 +90,11 @@
           inside the simple search string, which was often more confusing
           and ennoying than useful.</li>
 
+        <li>Extended file attributes are now indexed by default. As a
+          side effect, recoll now uses st_ctime, not st_mtime to detect
+          file changes. This means that installing 1.19 will reindex
+          many files (all those that were modified since created).</li>
+
         <li>The software can now be built so that indexing is runs with
         multiple threads. This can be a major performance boost for
         people with multiprocessor machines and big indexes. It is not