diff --git a/codeOptimizers/src/prog8/optimizer/ConstantFoldingOptimizer.kt b/codeOptimizers/src/prog8/optimizer/ConstantFoldingOptimizer.kt
index 63b8e31a5..7e9540d4f 100644
--- a/codeOptimizers/src/prog8/optimizer/ConstantFoldingOptimizer.kt
+++ b/codeOptimizers/src/prog8/optimizer/ConstantFoldingOptimizer.kt
@@ -87,7 +87,8 @@ class ConstantFoldingOptimizer(private val program: Program) : AstWalker() {
 
         if(expr.operator=="==" && rightconst!=null) {
             val leftExpr = expr.left as? BinaryExpression
-            if(leftExpr!=null) {
+            // only do this shuffling when the LHS is not a constant itself (otherwise problematic nested replacements)
+            if(leftExpr?.constValue(program) != null) {
                 val leftRightConst = leftExpr.right.constValue(program)
                 if(leftRightConst!=null) {
                     when (leftExpr.operator) {
diff --git a/docs/source/todo.rst b/docs/source/todo.rst
index a5d480918..377082c14 100644
--- a/docs/source/todo.rst
+++ b/docs/source/todo.rst
@@ -1,6 +1,11 @@
 TODO
 ====
 
+- is Jesko's circle algorithm faster than what we have now? https://en.wikipedia.org/wiki/Midpoint_circle_algorithm#Jesko's_Method
+  is it Correct??? Or better to use Casey's algorithm?
+- revisit the seek example to see if write-seeking now works?
+
+- remove unused interned strings in the resulting code (for example from removed if/else blocks)
 - [on branch: shortcircuit] investigate McCarthy evaluation again? this may also reduce code size perhaps for things like if a>4 or a<2 ....
 - [on branch: ir-less-branch-opcodes] IR: reduce the number of branch instructions such as BEQ, BEQR, etc (gradually), replace with CMP(I) + status branch instruction
 - IR: reduce amount of CMP/CMPI after instructions that set the status bits correctly (LOADs? INC? Bitwise operations, etc), but only after setting the status bits is verified!