diff --git a/.gitmodules b/.gitmodules
new file mode 100644
index 0000000..8005543
--- /dev/null
+++ b/.gitmodules
@@ -0,0 +1,6 @@
+[submodule "submodules/zpp"]
+	path = submodules/zpp
+	url = ssh://git@gitea.pestctrl.io:2223/Zander/zpp.git
+[submodule "extern/zpp"]
+	path = extern/zpp
+	url = ssh://git@gitea.pestctrl.io:2223/Zander/zpp.git
diff --git a/cmdln/new.c b/cmdln/new.c
index b2c9c1a..af0e021 100644
--- a/cmdln/new.c
+++ b/cmdln/new.c
@@ -86,12 +86,14 @@ struct cmdln_flags* new_cmdln_flags(
                 break;
             }
             
+            #ifndef RELEASE_BUILD
             case 5: // --test-interactive-using-input-file
             {
                 test_interactive_using_input_file = optarg;
                 
                 break;
             }
+            #endif
             
             case 'c': // --command
             {
diff --git a/extern/zpp b/extern/zpp
new file mode 160000
index 0000000..89bb342
--- /dev/null
+++ b/extern/zpp
@@ -0,0 +1 @@
+Subproject commit 89bb34256e6d876ecebd8d8ca1175a0388c3ddfc
diff --git a/globals/terminalcolors.c b/globals/terminalcolors.c
index 2921255..48614d7 100644
--- a/globals/terminalcolors.c
+++ b/globals/terminalcolors.c
@@ -1,7 +1,7 @@
 
 #include "terminalcolors.h"
 
-#include "bin/globals/terminalcolors.cpp-4.c"
+#include "bin/e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"
 
 
 
@@ -15,13 +15,13 @@
 
 struct colors terminalcolors = {
     .precedences = {
-#include "bin/globals/terminalcolors.cpp-18.c"
+        #include "bin/94850162c0dceac6c6f9558e763457a6e8ac75c569da6ba6c986030bd21d4b1f"
 
 
 
     },
 
-#include "bin/globals/terminalcolors.cpp-24.c"
+    #include "bin/6a7d755dd4b2ea24593a9d09a7cfd18ae05601661eb1ff5416e878a887508378"
 
 
 
diff --git a/globals/terminalcolors.cpp b/globals/terminalcolors.cpp
index 2f83e80..3cb510b 100644
--- a/globals/terminalcolors.cpp
+++ b/globals/terminalcolors.cpp
@@ -1,7 +1,7 @@
 
 #include "terminalcolors.h"
 
-/*/ python
+/*/
 
 import colorsys;
 
diff --git a/makefile b/makefile
index dc8fd1f..dffda42 100644
--- a/makefile
+++ b/makefile
@@ -33,9 +33,9 @@ include srclist.mk
 
 .PRECIOUS: %.c
 
-%.c: %.cpp zog.py | bin/
+%.c: %.cpp ./extern/zpp/zpp | bin/
 	@echo "preprocessing ${<} ..."
-	@./zog.py -i $< --snippet-directory=bin -o $@
+	@./extern/zpp/zpp $< --snippet-directory=bin -o $@
 
 ${prefix}/%.o ${prefix}/%.d: %.c ${optionset} | ${prefix}/%/
 	@echo "compiling (${buildtype}) ${*}.c ..."
diff --git a/parse/tokenizer/next.c b/parse/tokenizer/next.c
index 8844fe0..e741890 100644
--- a/parse/tokenizer/next.c
+++ b/parse/tokenizer/next.c
@@ -26,6 +26,7 @@
 #include "struct.h"
 #include "next.h"
 
+#if 0
 static const enum state {
     s_error,
     
@@ -136,6 +137,7 @@ static const enum state {
         [s_reading_identifier]['0' ... '9'] = s_reading_identifier,
         [s_reading_identifier][128] = s_reading_identifier,
 };
+#endif
 
 void tokenizer_next(
     struct tokenizer* this)
diff --git a/repl/main.c b/repl/main.c
new file mode 100644
index 0000000..cb4d2fe
--- /dev/null
+++ b/repl/main.c
@@ -0,0 +1,135 @@
+
+#include <assert.h>
+
+#include <debug.h>
+
+#include "main.h"
+
+void repl_main(void)
+{
+    ENTER;
+    
+    TODO;
+    
+    // if we're actually connected to a terminal:
+        // set terminal options.
+    // otherwise, open the file we're pretending is a terminal.
+    
+    // create input string
+        // which is an actual string that the history object has a reference to?
+    
+    // create history:
+        // I think this is a list of strings?
+    
+    // create terminal tokenizer
+    
+    // create terminal buffer:
+        // which keeps track of what characters we want where (not nessarily \\
+           the entire height of the terminal, but certianly the entire width) \\
+           characters can have foreground color (and maybe boldness?).
+        // has a insert column and line. handles word-wrap.
+        // keeps track of what's on the screen and only edits as nessisary.
+        // starts with assuming we're only editing one line (full width), but \\
+           if the expression line gets too long or printing error messages \\
+           might push us beyond that, then --even if we cleared-- we would \\
+           need to consider the new editing height.
+        // this terminal buffer also keeps track of where the cursor should \\
+           be vs. where it is and will update that apprioately as well.
+    
+    // loop:
+        // read next token
+        
+        // somehow also check if the terminal resized. If it did:
+            // tell the terminal buffer.
+                // internally this will completely reset what's live.
+        
+        // match token.kind:
+            // (almost) every case clears and redraws what we want the buffer \\
+               to contain.
+            
+            // case up, down:
+                // selects history index?
+                // which selects the string we're editing?
+            
+            // case left, right:
+                // just move cursor position, no clear needed.
+            
+            // case 'a-zA-Z0-9', etc.:
+                // insert the letter into the string
+                
+                // create parsing tokenizer
+                
+                // parse the string
+                    // all statements and expressions keep references to the \\
+                       tokens.
+                
+                // tell the statement to color tokens
+                
+                // the tokenizer keeps references to all tokens too.
+                
+                // clear the buffer
+                
+                // set buffer insert position to 0, 0.
+                
+                // maybe the prompt + line is longer than the terminal width.
+                    // some maybe we word-wrap?
+                
+                // draw the prompt line (either green or red)
+                
+                // iterate through tokenizer's tokens:
+                    // each token knows how many whitespace characters were \\
+                       read since the previous one.
+                    // print their text into the buffer with the token's color.
+                    // the buffer will handle word-wrapping.
+                    // some tokens don't have an expression or statement that \\
+                       owns them.
+                    // some part of the input text may not be part of a \\
+                       token? like a comment?
+                        // I think comments will need to be considered \\
+                           tokens that the parser doesn't pick up. That way \\
+                           it at least appears in the output line.
+            
+            // case delete, backspace:
+                // remove the letter at location
+                
+                // run through basically the same parsing logic as inserting \\
+                   a character
+            
+            // case enter:
+                // push a newline
+                // if line is not empty:
+                    // evaluate statement
+                        // which might print a value
+                            // into the terminal buffer? or straight out?
+                                // I think straight out.
+                // terminal buffer.clear(true-reset = True):
+                    // true reset will clear what's "live" and the buffer, \\
+                       and set the height back to 1.
+                // new string line
+                // new history entry
+                // redraw prompt.
+            
+        // update buffer:
+            // compare (buffer of what we want vs. what's live)
+    
+    TODO;
+    
+    EXIT;
+}
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
diff --git a/repl/main.h b/repl/main.h
new file mode 100644
index 0000000..fb0f38b
--- /dev/null
+++ b/repl/main.h
@@ -0,0 +1,2 @@
+
+void repl_main(void);
diff --git a/repl/tokenizer/free.c b/repl/tokenizer/free.c
new file mode 100644
index 0000000..e69de29
diff --git a/repl/tokenizer/free.h b/repl/tokenizer/free.h
new file mode 100644
index 0000000..e69de29
diff --git a/repl/tokenizer/new.c b/repl/tokenizer/new.c
new file mode 100644
index 0000000..e69de29
diff --git a/repl/tokenizer/new.h b/repl/tokenizer/new.h
new file mode 100644
index 0000000..e69de29
diff --git a/repl/tokenizer/next.c b/repl/tokenizer/next.c
new file mode 100644
index 0000000..e69de29
diff --git a/repl/tokenizer/next.h b/repl/tokenizer/next.h
new file mode 100644
index 0000000..e69de29
diff --git a/repl/tokenizer/struct.h b/repl/tokenizer/struct.h
new file mode 100644
index 0000000..e69de29
diff --git a/zog.py b/zog.py
deleted file mode 100755
index 695cad7..0000000
--- a/zog.py
+++ /dev/null
@@ -1,113 +0,0 @@
-#!/usr/bin/env python3
-
-import os;
-import sys;
-import io;
-import argparse;
-
-parser = argparse.ArgumentParser(prog = 'zog');
-
-parser.add_argument('-i', '--input', required = True);
-parser.add_argument('-o', '--output', required = True);
-parser.add_argument('-s', '--snippet-directory', default = "/tmp/");
-
-args = parser.parse_args(sys.argv[1:]);
-
-input_file = args.input;
-output_file = args.output;
-snippet_directory = args.snippet_directory;
-
-while input_file.startswith("./"):
-    input_file = input_file[2:];
-
-while snippet_directory.startswith("./"):
-    snippet_directory = snippet_directory[2:];
-
-while snippet_directory.endswith("/"):
-    snippet_directory = snippet_directory[:-1];
-
-myglobals = dict();
-
-istream = open(input_file, "r");
-ostream = open(output_file, "w");
-
-lines = istream.readlines();
-
-i = 0;
-n = len(lines);
-
-while i < n:
-    if "/*/" in lines[i]:
-        indentation = lines[i].index("/*/");
-        
-        start_index = i;
-        
-        i += 1;
-        
-        source = ""
-        
-        while "/*/" not in lines[i]:
-            source += lines[i][indentation:];
-            i += 1;
-            
-        end_index = i;
-        i += 1;
-        
-        real_stdout = sys.stdout
-        sys.stdout = captured_stdout = io.StringIO()
-        eval(compile(source, "<zog-snippet>", "exec"), myglobals);
-        sys.stdout = real_stdout
-        
-        outstring = captured_stdout.getvalue();
-        
-        if outstring and outstring[-1] != "\n":
-            outstring += "\n"
-        
-        temppath = f"{snippet_directory}/{input_file}-{start_index+1}.c";
-        
-        tempdir = temppath[:temppath.rindex('/')];
-        
-        os.makedirs(tempdir, exist_ok = True);
-        
-        with open(temppath, "w") as stream:
-            for outline in outstring.split("\n"):
-                stream.write((" " * indentation) + outline + '\n');
-        
-        ostream.write(f'#include "{temppath}"' + "\n");
-        ostream.write("\n" * (end_index - start_index));
-    else:
-        ostream.write(lines[i]);
-        i += 1;
-
-istream.close();
-ostream.close();
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-