diff --git a/devdocs/specs/core-search.txt b/devdocs/specs/core-search.txt
index 3f97eb55..16f4e98d 100644
--- a/devdocs/specs/core-search.txt
+++ b/devdocs/specs/core-search.txt
@@ -35,6 +35,9 @@ REQUIREMENTS
 
 - USE-CASE: Central text search for any match. Can include tags.  Can specify a type of object result desired.
 
+- MINIMUM 2 character words only are searchable
+- MAXIMUM 255 character words only are indexed (cut off at 255 if longer)
+
 - ? USE-CASE: In-object text search for the typeandid that user is in, e.g. when in Client info form can search on that client. 
     - This is problematic because it seems to require searching for related objects but  that's deprecated to priority 2 
     - Maybe more ideally is a search that can be something like "search all workorders FOR THIS CLIENT" or search all units "FOR THIS CLIENT" 
@@ -68,7 +71,6 @@ REQUIREMENTS
 - Coding: break this into separate discrete classes, the old v7 code is very monolithic and in-elegant
 
 //TODO:
-- SAMPLE DATA: Need a huge amount of sample data indexed to load test it
 - INDEXES: play with it and see what works best
 
 
diff --git a/devdocs/todo.txt b/devdocs/todo.txt
index dbfed66b..74786d61 100644
--- a/devdocs/todo.txt
+++ b/devdocs/todo.txt
@@ -29,11 +29,8 @@ IMMEDIATE ITEMS:
 ================
 
 	
-	- Search and search text indexing			
-		- Create a test for search that searches the widgets LOREM text 
-			- Test with huge dataset for performance testing
-		- Need to be able to specify max results, code that shit!
-		- bugbug: why is the single letter a being indexed?  Missing shortness filter, A not in stopwords for english??		
+	- Search and search text indexing					
+		- Can it be sped up with huge data test??	
 		- Update all the other routes to include search indexing (attachments, tags etc, anything with text in it)
 
 	
diff --git a/server/AyaNova/biz/Search.cs b/server/AyaNova/biz/Search.cs
index bee6c82e..7bc8141e 100644
--- a/server/AyaNova/biz/Search.cs
+++ b/server/AyaNova/biz/Search.cs
@@ -568,6 +568,7 @@ namespace AyaNova.Biz
             //Get stopwords and CJKIndex flag value
             LocaleWordBreakingData LocaleSearchData = GetLocaleSearchData(localeId);
             int MAXWORDLENGTH = 255;
+            int MINWORDLENGTH = 2;//A word isn't a word unless it's got at least two characters in it
             StringBuilder sbResults = new StringBuilder();
             //List to temporarily hold parsed words
             //used to easily ensure unique words only
@@ -824,10 +825,15 @@ namespace AyaNova.Biz
             //from the word list
             foreach (string s in tempParsedWords)
             {
-                //Add only non stopwords
-                if (!LocaleSearchData.StopWords.Contains(s))
+                //Filter out short words if we are breaking for indexing
+                //but keep them if they are part of a wildcard search phrase
+                if (s.Length > MINWORDLENGTH || (KeepWildCards && s.Contains('%')))
                 {
-                    ReturnList.Add(s);
+                    //Add only non stopwords
+                    if (!LocaleSearchData.StopWords.Contains(s))
+                    {
+                        ReturnList.Add(s);
+                    }
                 }
             }
 
diff --git a/test/raven-integration/Search/SearchOps.cs b/test/raven-integration/Search/SearchOps.cs
index 02a5e067..77ccfc01 100644
--- a/test/raven-integration/Search/SearchOps.cs
+++ b/test/raven-integration/Search/SearchOps.cs
@@ -451,7 +451,7 @@ namespace raven_integration
 
 
         [Fact]
-        public async void BigDataSearchShouldHonourMaxResultsAndBeRelativelyFast()
+        public async void ConstrainedBigDataSearchShouldHonourMaxResultsAndBeRelativelyFast()
         {
 
             //THIS test is a bit different in that it relies partly on the big dataset for testing
@@ -481,7 +481,44 @@ namespace raven_integration
             ResultCount.Should().BeLessOrEqualTo(1000);
 
             //1755ms is the longest I've seen in initial testing with all 1000 results so setting slightly above
-            TimeToSearch.Should().BeLessThan(1760, "Big data search should not be too slow");
+            TimeToSearch.Should().BeLessThan(1760, "Constrained big data search should not be too slow");
+
+        }//eot
+
+
+        [Fact]
+        public async void UnboundBigDataSearchShouldBeRelativelyFast()
+        {
+
+            //THIS test is a bit different in that it relies partly on the big dataset for testing
+            //so it has different paths depending upon if it's testing against the big data or not
+            const string TEST_SEARCH_PHRASE = "et*";
+
+            //Now see if can find those objects with a phrase search
+            dynamic SearchParameters = new JObject();
+
+            SearchParameters.phrase = TEST_SEARCH_PHRASE;
+            SearchParameters.nameOnly = false;
+            SearchParameters.typeOnly = 0;//no type
+            SearchParameters.maxResults = 0;//0=return all results
+
+            var watch = new System.Diagnostics.Stopwatch();
+            watch.Start();
+            ApiResponse a = await Util.PostAsync("Search", await Util.GetTokenAsync("manager", "l3tm3in"), SearchParameters.ToString());
+            watch.Stop();
+
+            var TimeToSearch = watch.ElapsedMilliseconds;
+
+            Util.ValidateDataReturnResponseOk(a);
+
+            //Now validate the return list
+            var ResultCount = ((JArray)a.ObjectResponse["result"]).Count;
+            //assert it's not unbounded
+            // ResultCount.Should().BeGreaterThan(1000);
+
+            //24072 ms is the longest I've seen in initial testing with all bigData seeded results so setting slightly above
+            TimeToSearch.Should().BeLessThan(24072, "Unconstrained big data search should not be too slow");
+            //Fastest is 17227 ms with 14143 results
 
         }//eot