diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
index 0b191b1..2161e9b 100644
--- a/.github/workflows/release.yml
+++ b/.github/workflows/release.yml
@@ -11,7 +11,10 @@ on:
 
 jobs:
   build_and_release:
-    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        os: [ubuntu-latest, macos-latest, windows-latest]
+    runs-on: ${{ matrix.os }}
 
     steps:
       - name: Checkout code
@@ -28,15 +31,18 @@ jobs:
         run: |
           python -m pip install --upgrade pip
           pip install conan
+        shell: ${{ matrix.os == 'windows-latest' && 'pwsh' || 'bash' }}
 
       - name: Configure Conan
-        run: |
-          conan profile detect
+        run: conan profile detect
+        shell: ${{ matrix.os == 'windows-latest' && 'pwsh' || 'bash' }}
 
       - name: Install dependencies and build with Conan
         run: |
           conan install . --build=missing
           conan build .
+        shell: ${{ matrix.os == 'windows-latest' && 'pwsh' || 'bash' }}
+
       - name: Determine if prerelease
         id: prerelease_check
         run: |
@@ -47,16 +53,39 @@ jobs:
             echo "false" > prerelease.txt
           fi
           echo "prerelease=$(cat prerelease.txt)" >> $GITHUB_OUTPUT
+        shell: ${{ matrix.os == 'windows-latest' && 'pwsh' || 'bash' }}
+
       - name: Archive build folder
         run: |
           TAG=${GITHUB_REF##refs/tags/}
-          ARCH=$(uname -m)
-          FOLDER_NAME="chloride-$TAG-linux-${ARCH}"
-          TAR_NAME="$FOLDER_NAME.tar.gz"
-          mv build/bin "$FOLDER_NAME"
-          cp LICENSE "$FOLDER_NAME"
-          tar -czf "$TAR_NAME" "$FOLDER_NAME"
+          if [[ "$RUNNER_OS" == "Linux" ]]; then
+            OS=linux
+            ARCH=$(uname -m)
+            FOLDER_NAME="chloride-$TAG-$OS-$ARCH"
+            TAR_NAME="$FOLDER_NAME.tar.gz"
+            mv build/bin "$FOLDER_NAME"
+            cp LICENSE "$FOLDER_NAME"
+            tar -czf "$TAR_NAME" "$FOLDER_NAME"
+          elif [[ "$RUNNER_OS" == "macOS" ]]; then
+            OS=macos
+            ARCH=$(uname -m)
+            FOLDER_NAME="chloride-$TAG-$OS-$ARCH"
+            TAR_NAME="$FOLDER_NAME.tar.gz"
+            mv build/bin "$FOLDER_NAME"
+            cp LICENSE "$FOLDER_NAME"
+            tar -czf "$TAR_NAME" "$FOLDER_NAME"
+          else
+            # Windows
+            $TAG = $env:GITHUB_REF -replace 'refs/tags/', ''
+            $ARCH = if ([Environment]::Is64BitOperatingSystem) { 'x64' } else { 'x86' }
+            $FOLDER_NAME = "chloride-$TAG-windows-$ARCH"
+            $TAR_NAME = "$FOLDER_NAME.zip"
+            Rename-Item build\bin $FOLDER_NAME
+            Copy-Item LICENSE $FOLDER_NAME
+            Compress-Archive -Path $FOLDER_NAME -DestinationPath $TAR_NAME
+          fi
           echo "TAR_NAME=$TAR_NAME" >> $GITHUB_ENV
+        shell: ${{ matrix.os == 'windows-latest' && 'pwsh' || 'bash' }}
 
       - name: Create GitHub Release
         id: create_release
diff --git a/src/dynamic_array/darray.c b/src/dynamic_array/darray.c
index fcd1174..7818b8d 100644
--- a/src/dynamic_array/darray.c
+++ b/src/dynamic_array/darray.c
@@ -26,7 +26,7 @@ void darray_resize(DArray *arr, size_t new_size) {
 
   // Determine number of full chunks needed to store new_size elements
   size_t required_bytes = new_size * arr->element_size;
-  size_t new_capacity_bytes = ((required_bytes + CHUNK_SIZE - 1) / CHUNK_SIZE) * CHUNK_SIZE;
+  size_t new_capacity_bytes = required_bytes*2;
   size_t new_capacity = new_capacity_bytes / arr->element_size;
   if (!new_capacity) {
     return;
diff --git a/src/main.c b/src/main.c
index 0c99c27..c481178 100644
--- a/src/main.c
+++ b/src/main.c
@@ -158,7 +158,7 @@ int load_cache(Translated *translated_dest, char *joined_paths, uint64_t hash,
                char *source_path) {
   FILE *bytecode_file = fopen(joined_paths, "rb");
   if (!bytecode_file) {
-    printf("cache doesnt exist... compiling from source.\n");
+    fprintf(stderr,"cache doesnt exist... compiling from source.\n");
     return 1;
   }
 
@@ -194,7 +194,7 @@ int load_cache(Translated *translated_dest, char *joined_paths, uint64_t hash,
   XXH64_freeState(state);
 
   if (calc_hash != stored_hash) {
-    printf("cache hash mismatch (corrupted?)\n");
+    fprintf(stderr,"cache hash mismatch (corrupted?)\n");
     goto FAILED;
   }
 
@@ -268,11 +268,11 @@ int load_cache(Translated *translated_dest, char *joined_paths, uint64_t hash,
     goto FAILED;
   }
 
-  printf("cache exists and is valid, so will be used.\n");
+  fprintf(stderr,"cache exists and is valid, so will be used.\n");
   fclose(bytecode_file);
   return 0;
 FAILED:
-  printf("cache is invalid... compiling from source.\n");
+  fprintf(stderr,"cache is invalid... compiling from source.\n");
   fclose(bytecode_file);
   return 1;
 }
@@ -346,7 +346,7 @@ Execution execute(char *path, Stack *stack) {
     end = clock();
     time_spent = (double)(end - start) / CLOCKS_PER_SEC;
     total_time_spent += time_spent;
-    printf("Lexer time taken: %f seconds\n", time_spent);
+    fprintf(stderr,"Lexer time taken: %f seconds\n", time_spent);
     fclose(state.file);
 
     DArray ast;
@@ -363,7 +363,7 @@ Execution execute(char *path, Stack *stack) {
     end = clock();
     time_spent = (double)(end - start) / CLOCKS_PER_SEC;
     total_time_spent += time_spent;
-    printf("Parser time taken: %f seconds\n", time_spent);
+    fprintf(stderr,"Parser time taken: %f seconds\n", time_spent);
     darray_free(&tokens, free_token);
 
     start = clock();
@@ -380,7 +380,7 @@ Execution execute(char *path, Stack *stack) {
     end = clock();
     time_spent = (double)(end - start) / CLOCKS_PER_SEC;
     total_time_spent += time_spent;
-    printf("Translation time taken: %f seconds\n", time_spent);
+    fprintf(stderr,"Translation time taken: %f seconds\n", time_spent);
 
     darray_free(&ast, free_parsed);
 #if defined(__linux__)
@@ -452,8 +452,8 @@ Execution execute(char *path, Stack *stack) {
   end = clock();
   time_spent = (double)(end - start) / CLOCKS_PER_SEC;
   total_time_spent += time_spent;
-  printf("Execution time taken: %f seconds\n", time_spent);
-  printf("total time taken: %f seconds\n", total_time_spent);
+  fprintf(stderr,"Execution time taken: %f seconds\n", time_spent);
+  fprintf(stderr,"total time taken: %f seconds\n", total_time_spent);
 
   return (Execution){err, *main_scope};
 }
diff --git a/src/parser/parser.c b/src/parser/parser.c
index 41a1020..1d69034 100644
--- a/src/parser/parser.c
+++ b/src/parser/parser.c
@@ -23,7 +23,6 @@
 #include "return/return.h"
 #include "string/string.h"
 #include <gmp.h>
-#include <gmp.h>
 #include <stdbool.h>
 #include <stddef.h>
 #include <stdio.h>
@@ -75,6 +74,8 @@ ParsedValueReturn parse_token_full(char *file, DArray *tokens, size_t *index,
       return parse_if(file, tokens, index);
     case TOKEN_RETURN:
       return parse_return(file, tokens, index);
+    case TOKEN_LET:
+      return parse_declaration(file, tokens, index);
     default:
       break;
     };
@@ -121,9 +122,6 @@ ParsedValueReturn parse_token_full(char *file, DArray *tokens, size_t *index,
     (*index)++;
     output = parse_number(token, file);
     break;
-  case TOKEN_LET:
-    output = parse_declaration(file, tokens, index);
-    break;
   case TOKEN_DO:
     output = parse_dowrap(file, tokens, index);
     break;
diff --git a/src/runtime/internals/dynamic_array_armem/darray_armem.c b/src/runtime/internals/dynamic_array_armem/darray_armem.c
index 7d3202e..cf1842e 100644
--- a/src/runtime/internals/dynamic_array_armem/darray_armem.c
+++ b/src/runtime/internals/dynamic_array_armem/darray_armem.c
@@ -36,8 +36,7 @@ void darray_armem_resize(darray_armem *arr, size_t new_size) {
   }
 
   size_t required_bytes = new_size * arr->element_size;
-  size_t new_capacity_bytes =
-      ((required_bytes + CHUNK_SIZE - 1) / CHUNK_SIZE) * CHUNK_SIZE;
+  size_t new_capacity_bytes =required_bytes*2;
   size_t new_capacity = new_capacity_bytes / arr->element_size;
 
   if (!new_capacity) {
diff --git a/src/runtime/objects/string/string.c b/src/runtime/objects/string/string.c
index e5d719c..e81f98c 100644
--- a/src/runtime/objects/string/string.c
+++ b/src/runtime/objects/string/string.c
@@ -10,12 +10,14 @@
 #include <stdio.h>
 #include <string.h>
 #include "string.h"
+#include "../number/number.h"
 
 ArgonObject *ARGON_STRING_TYPE = NULL;
 
 ArgonObject *new_string_object(char*data, size_t length) {
   ArgonObject * object = new_object();
   add_field(object, "__class__", ARGON_STRING_TYPE);
+  add_field(object, "length", new_number_object_from_long(length, 1));
   object->type = TYPE_STRING;
   object->value.as_str.data = ar_alloc_atomic(length);
   memcpy(object->value.as_str.data, data, length);
diff --git a/src/translator/call/call.c b/src/translator/call/call.c
index 89426bc..41c3b7b 100644
--- a/src/translator/call/call.c
+++ b/src/translator/call/call.c
@@ -14,6 +14,7 @@ size_t translate_parsed_call(Translated *translated, ParsedCall *call,
   }
   push_instruction_byte(translated, OP_INIT_CALL);
   push_instruction_code(translated, call->args.size);
+  push_instruction_byte(translated, OP_NEW_SCOPE);
 
   DArray *old_return_jumps = translated->return_jumps;
   translated->return_jumps = NULL;
@@ -28,6 +29,7 @@ size_t translate_parsed_call(Translated *translated, ParsedCall *call,
   }
 
   translated->return_jumps = old_return_jumps;
+  push_instruction_byte(translated, OP_POP_SCOPE);
 
   push_instruction_byte(translated, OP_SOURCE_LOCATION);
   push_instruction_code(translated, call->line);
diff --git a/testing.ar b/testing.ar
index 158c203..a1dd833 100644
--- a/testing.ar
+++ b/testing.ar
@@ -1 +1,10 @@
-term.log("hello "+"world")
\ No newline at end of file
+let say_hi(name) = do
+    let z(y) = do
+        return y
+    let u = z(
+    do
+        return name
+    )
+    return "hello "+u+", how are you?"
+
+term.log(say_hi("william"))
\ No newline at end of file