diff --git a/lang/Makefile b/lang/Makefile
index 1bf5e42c..35217c71 100644
--- a/lang/Makefile
+++ b/lang/Makefile
@@ -47,13 +47,10 @@ endif
 
 interpolate/parse.generated.go: interpolate/parse.rl
 	@echo "Generating: interpolation..."
-	ragel -Z -G2 -o interpolate/parse.generated.go interpolate/parse.rl
-	#@ROOT="$$( cd "$$( dirname "$${BASH_SOURCE[0]}" )" && cd .. && pwd )" && $$ROOT/misc/header.sh 'interpolate/parse.generated.go'
-	# XXX: I have no idea why I need to sed twice. I give up :P
+	ragel -Z -G2 -o $@ $<
 	# remove the ragel header so our header test passes
-	@sed -i -e "1d" 'interpolate/parse.generated.go'
-	@sed -i -e "1d" 'interpolate/parse.generated.go'
-	gofmt -s -w 'interpolate/parse.generated.go'
+	sed -i -e 1,2d $@
+	gofmt -s -w $@
 
 fuzz:
 	@$(MAKE) --quiet -C fuzz
diff --git a/lang/interpolate/parse.rl b/lang/interpolate/parse.rl
index 15a36edc..e006df4e 100644
--- a/lang/interpolate/parse.rl
+++ b/lang/interpolate/parse.rl
@@ -76,7 +76,7 @@ func Parse(data string) (out Stream, _ error) {
 		var = '${' var_name '}' ;
 
 		# Any special escape characters are matched here.
-		escaped_lit = '\\' ( 'a' | 'b' | 'f' | 'n' | 'r' | 't' | 'v' | '\\' | '"' | '$' )
+		escaped_lit = '\\' (any)
 		@{
 			switch s := data[fpc:fpc+1]; s {
 			case "a":
@@ -105,23 +105,16 @@ func Parse(data string) (out Stream, _ error) {
 			//	x = "\x00"
 			default:
 				//x = s // in case we want to avoid erroring
-				// this is a programming (parser) error I think
-				return nil, fmt.Errorf("unhandled escape sequence token: %s", s)
+				return nil, fmt.Errorf("unknown escape sequence: \\%s", s)
 			}
 			l = Literal{Value: x}
 		};
 
-		# XXX: explicitly try and add this one?
-		#escape_lit = '\\\\'
-		#@{
-		#	l = Literal{Value: "\\\\"}
-		#};
-
-		# Anything followed by a '$' that is not a '{' is used as-is
-		# with the dollar.
-		dollar_lit = '$' (any - '{')
+		# A lone dollar is a literal, if it is not a var. The `token` rule
+		# declares a var match is attempted first, else a `lit` and thus this.
+		dollar_lit = '$'
 		@{
-			l = Literal{Value: data[fpc-1:fpc+1]}
+			l = Literal{Value: data[fpc:fpc+1]}
 		};
 
 		# Literal strings that don't contain '$' or '\'.
diff --git a/lang/interpret_test/TestAstFunc2/escaping1.output b/lang/interpret_test/TestAstFunc2/escaping1.output
index 53fca49c..dd5a3486 100644
--- a/lang/interpret_test/TestAstFunc2/escaping1.output
+++ b/lang/interpret_test/TestAstFunc2/escaping1.output
@@ -19,3 +19,8 @@ Vertex: test[R: \This is r1 EOF]
 Vertex: test[S: \$ EOF]
 Vertex: test[T: newline
 EOF]
+Vertex: test[U: tab	\	tabEOF]
+Vertex: test[W: \$]
+Vertex: test[X: $This is x1 EOF]
+Vertex: test[Y: ${unused} EOF]
+Vertex: test[Z: $$$]
diff --git a/lang/interpret_test/TestAstFunc2/escaping1/main.mcl b/lang/interpret_test/TestAstFunc2/escaping1/main.mcl
index 87ff5475..dff9b054 100644
--- a/lang/interpret_test/TestAstFunc2/escaping1/main.mcl
+++ b/lang/interpret_test/TestAstFunc2/escaping1/main.mcl
@@ -46,23 +46,17 @@ test "S: \\$ EOF" {}
 
 test "T: newline\nEOF" {}
 
-# XXX: possible bugs or misunderstood expectations:
+test "U: tab\t\\\ttabEOF" {}
 
-#test "W: \\$" {}
-# got: <error>
-# exp: W: \$
+# test "V: " {}
 
-#$x1 = "This is x1"
-#test "X: $${x1} EOF" {}
-# got: X: $${x1} EOF
-# exp: X: $This is x1 EOF
+test "W: \\$" {}
 
-#$unused = "i am unused"
-#$y1 = "{unused}"
-#test "Y: $${y1} EOF" {} # check there isn't double parsing
-# got: Y: $${y1} EOF
-# exp: Y: ${unused} EOF
+$x1 = "This is x1"
+test "X: $${x1} EOF" {}
 
-#test "Z: $$$" {}
-# got: <error>
-# exp: Z: $$$ EOF
+$unused = "i am unused"
+$y1 = "{unused}"
+test "Y: $${y1} EOF" {} # check there isn't double parsing
+
+test "Z: $$$" {}
diff --git a/lang/interpret_test/TestAstFunc2/escaping4.output b/lang/interpret_test/TestAstFunc2/escaping4.output
index 3315675b..a75274ec 100644
--- a/lang/interpret_test/TestAstFunc2/escaping4.output
+++ b/lang/interpret_test/TestAstFunc2/escaping4.output
@@ -1 +1 @@
-# err: errInterpolate: parser failed: cannot parse string: X: \z
+# err: errInterpolate: parser failed: unknown escape sequence: \z
diff --git a/lang/interpret_test/TestAstFunc2/escaping5.output b/lang/interpret_test/TestAstFunc2/escaping5.output
index c2d5f06f..c81498b1 100644
--- a/lang/interpret_test/TestAstFunc2/escaping5.output
+++ b/lang/interpret_test/TestAstFunc2/escaping5.output
@@ -1 +1 @@
-# err: errInterpolate: parser failed: cannot parse string: X: there is no \j sequence
+# err: errInterpolate: parser failed: unknown escape sequence: \j