diff --git a/mendel/arduino.h b/mendel/arduino.h
index b2fc8ee..246f8fd 100644
--- a/mendel/arduino.h
+++ b/mendel/arduino.h
@@ -11,6 +11,12 @@
 #define		MASK(PIN)				(1 << PIN)
 #endif
 
+/*
+	magic I/O routines
+
+	now you can simply SET_OUTPUT(STEP); WRITE(STEP, 1); WRITE(STEP, 0);
+*/
+
 #define		_READ(IO)					(IO ## _RPORT & MASK(IO ## _PIN))
 #define		_WRITE(IO, v)			do { if (v) { IO ## _WPORT |= MASK(IO ## _PIN); } else { IO ## _WPORT &= ~MASK(IO ## _PIN); }; } while (0)
 #define		_TOGGLE(IO)				(IO ## _RPORT = MASK(IO ## _PIN))
@@ -33,6 +39,10 @@
 
 /*
 	ports and functions
+
+	added as necessary or if I feel like it- not a comprehensive list!
+
+	probably needs some #ifdefs for various chip types
 */
 
 // UART
diff --git a/mendel/clock.c b/mendel/clock.c
index d1df4d6..75e0d00 100644
--- a/mendel/clock.c
+++ b/mendel/clock.c
@@ -19,7 +19,8 @@ volatile uint32_t	clock = 0;
 
 // 1/4 second tick
 uint8_t						clock_counter_250ms = 0;
-volatile uint8_t	clock_flag_250ms = 0;
+uint8_t						clock_counter_1s = 0;
+volatile uint8_t	clock_flag = 0;
 
 void clock_setup() {
 	// use system clock
@@ -39,17 +40,19 @@ void clock_setup() {
 }
 
 ISR(TIMER2_COMPA_vect) {
-// 	WRITE(SCK, 0);
 	// global clock
 #ifdef	GLOBAL_CLOCK
 	clock++;
 #endif
 	// 1/4 second tick
 	if (++clock_counter_250ms == 250) {
-		clock_flag_250ms = 255;
+		clock_flag |= CLOCK_FLAG_250MS;
 		clock_counter_250ms = 0;
+		if (++clock_counter_1s == 4) {
+			clock_flag |= CLOCK_FLAG_1S;
+			clock_counter_1s = 0;
+		}
 	}
-// 	WRITE(SCK, 1);
 }
 
 #ifdef	GLOBAL_CLOCK
diff --git a/mendel/clock.h b/mendel/clock.h
index 2edf7ef..59612fb 100644
--- a/mendel/clock.h
+++ b/mendel/clock.h
@@ -9,25 +9,24 @@ void			clock_setup(void);
 uint32_t	clock_read(void);
 #endif
 
-extern volatile uint8_t	clock_flag_250ms;
+extern volatile uint8_t	clock_flag;
 
 #define	CLOCK_FLAG_250MS							1
-// #define	CLOCK_FLAG_250MS_TEMPCHECK		1
-// #define	CLOCK_FLAG_250MS_REPORT				2
-// #define	CLOCK_FLAG_250MS_STEPTIMEOUT	4
+#define	CLOCK_FLAG_1S									2
 
 /*
 	ifclock() {}
 
 	so we can do stuff like:
-	ifclock(CLOCK_FLAG_250MS_REPORT) {
+	ifclock(CLOCK_FLAG_250MS) {
 		report();
 	}
 
 	or:
-	ifclock(CLOCK_FLAG_250MS_STEPTIMEOUT)
+	ifclock(CLOCK_FLAG_1S)
 		disable_steppers();
 */
-#define	ifclock(F)	for (;clock_flag_250ms & (F);clock_flag_250ms &= ~(F))
+
+#define	ifclock(F)	for (;clock_flag & (F);clock_flag &= ~(F))
 
 #endif	/* _CLOCK_H */
diff --git a/mendel/dda.c b/mendel/dda.c
index 703e18b..044fd00 100644
--- a/mendel/dda.c
+++ b/mendel/dda.c
@@ -174,25 +174,6 @@ void dda_create(DDA *dda, TARGET *target) {
 			serwrite_uint32(dda->total_steps); serial_writechar(',');
 		}
 
-	// 	if (dda->f_delta > dda->total_steps) {
-	// 		dda->f_scale = dda->f_delta / dda->total_steps;
-	// 		if (dda->f_scale > 3) {
-	// 			dda->f_delta = dda->total_steps;
-	// 		}
-	// 		else {
-	// 			// if we boost the number of steps here, many will only be F-steps which take no time- maybe we should calculate move_distance first?
-	// 			dda->f_scale = 1;
-	// 			dda->total_steps = dda->f_delta;
-	// 		}
-	// 	}
-	// 	else {
-	// 		dda->f_scale = 1;
-	// 	}
-	//
-	// 	if (DEBUG) {
-	// 		serwrite_uint32(dda->total_steps); serial_writechar(',');
-	// 	}
-
 		dda->x_counter = dda->y_counter = dda->z_counter = dda->e_counter = dda->f_counter =
 			-(dda->total_steps >> 1);
 
@@ -206,8 +187,6 @@ void dda_create(DDA *dda, TARGET *target) {
 
 		if (distance < 2)
 			distance = dda->e_delta * UM_PER_STEP_E;
-	// 	if (distance < 2)
-	// 		distance = dda->f_delta;
 
 		if (DEBUG) {
 			serwrite_uint32(distance); serial_writechar(',');
@@ -241,12 +220,6 @@ void dda_create(DDA *dda, TARGET *target) {
 
 void dda_start(DDA *dda) {
 	// called from interrupt context: keep it simple!
-// 	if (
-// 			(current_position.X == dda->endpoint.X) &&
-// 			(current_position.Y == dda->endpoint.Y) &&
-// 			(current_position.Z == dda->endpoint.Z) &&
-// 			(current_position.E == dda->endpoint.E)
-// 		 ) {
 	if (dda->nullmove) {
 		// just change speed?
 		current_position.F = dda->endpoint.F;
@@ -310,123 +283,110 @@ void dda_step(DDA *dda) {
 	if (DEBUG)
 		serial_writechar('!');
 
-	// turn 'L' light OFF so it's obvious if we froze in this routine
-// 	if (DEBUG)
-// 		WRITE(SCK, 0);
-
-// 	do {
-// 		WRITE(SCK, 0);
-
-		step_option &= ~(X_CAN_STEP | Y_CAN_STEP | Z_CAN_STEP | E_CAN_STEP | F_CAN_STEP);
+	step_option &= ~(X_CAN_STEP | Y_CAN_STEP | Z_CAN_STEP | E_CAN_STEP | F_CAN_STEP);
 // 		step_option |= can_step(x_min(), x_max(), current_position.X, dda->endpoint.X, dda->x_direction) & X_CAN_STEP;
 // 		step_option |= can_step(y_min(), y_max(), current_position.Y, dda->endpoint.Y, dda->y_direction) & Y_CAN_STEP;
 // 		step_option |= can_step(z_min(), z_max(), current_position.Z, dda->endpoint.Z, dda->z_direction) & Z_CAN_STEP;
-		step_option |= can_step(0      , 0      , current_position.X, dda->endpoint.X, dda->x_direction) & X_CAN_STEP;
-		step_option |= can_step(0      , 0      , current_position.Y, dda->endpoint.Y, dda->y_direction) & Y_CAN_STEP;
-		step_option |= can_step(0      , 0      , current_position.Z, dda->endpoint.Z, dda->z_direction) & Z_CAN_STEP;
-		step_option |= can_step(0      , 0      , current_position.E, dda->endpoint.E, dda->e_direction) & E_CAN_STEP;
-		step_option |= can_step(0      , 0      , current_position.F, dda->endpoint.F, dda->f_direction) & F_CAN_STEP;
+	step_option |= can_step(0      , 0      , current_position.X, dda->endpoint.X, dda->x_direction) & X_CAN_STEP;
+	step_option |= can_step(0      , 0      , current_position.Y, dda->endpoint.Y, dda->y_direction) & Y_CAN_STEP;
+	step_option |= can_step(0      , 0      , current_position.Z, dda->endpoint.Z, dda->z_direction) & Z_CAN_STEP;
+	step_option |= can_step(0      , 0      , current_position.E, dda->endpoint.E, dda->e_direction) & E_CAN_STEP;
+	step_option |= can_step(0      , 0      , current_position.F, dda->endpoint.F, dda->f_direction) & F_CAN_STEP;
 
-		if (step_option & X_CAN_STEP) {
-			dda->x_counter -= dda->x_delta;
-			if (dda->x_counter < 0) {
-				step_option |= REAL_MOVE;
+	if (step_option & X_CAN_STEP) {
+		dda->x_counter -= dda->x_delta;
+		if (dda->x_counter < 0) {
+			step_option |= REAL_MOVE;
 
-				x_step();
-				if (dda->x_direction)
-					current_position.X++;
-				else
-					current_position.X--;
+			x_step();
+			if (dda->x_direction)
+				current_position.X++;
+			else
+				current_position.X--;
 
-				dda->x_counter += dda->total_steps;
+			dda->x_counter += dda->total_steps;
+		}
+	}
+
+	if (step_option & Y_CAN_STEP) {
+		dda->y_counter -= dda->y_delta;
+		if (dda->y_counter < 0) {
+			step_option |= REAL_MOVE;
+
+			y_step();
+			if (dda->y_direction)
+				current_position.Y++;
+			else
+				current_position.Y--;
+
+			dda->y_counter += dda->total_steps;
+		}
+	}
+
+	if (step_option & Z_CAN_STEP) {
+		dda->z_counter -= dda->z_delta;
+		if (dda->z_counter < 0) {
+			step_option |= REAL_MOVE;
+
+			z_step();
+			if (dda->z_direction)
+				current_position.Z++;
+			else
+				current_position.Z--;
+
+			dda->z_counter += dda->total_steps;
+		}
+	}
+
+	if (step_option & E_CAN_STEP) {
+		dda->e_counter -= dda->e_delta;
+		if (dda->e_counter < 0) {
+			step_option |= REAL_MOVE;
+
+			e_step();
+			if (dda->e_direction)
+				current_position.E++;
+			else
+				current_position.E--;
+
+			dda->e_counter += dda->total_steps;
+		}
+	}
+
+	if (step_option & F_CAN_STEP) {
+		dda->f_counter -= dda->f_delta;
+		// since we don't allow total_steps to be defined by F, we may need to step multiple times if f_delta is greater than total_steps
+		while (dda->f_counter < 0) {
+
+			dda->f_counter += dda->total_steps;
+
+			if (dda->f_direction) {
+				current_position.F += 1;
+				if (current_position.F > dda->endpoint.F)
+					current_position.F = dda->endpoint.F;
 			}
-		}
-
-		if (step_option & Y_CAN_STEP) {
-			dda->y_counter -= dda->y_delta;
-			if (dda->y_counter < 0) {
-				step_option |= REAL_MOVE;
-
-				y_step();
-				if (dda->y_direction)
-					current_position.Y++;
-				else
-					current_position.Y--;
-
-				dda->y_counter += dda->total_steps;
+			else {
+				current_position.F -= 1;
+				if (current_position.F < dda->endpoint.F)
+					current_position.F = dda->endpoint.F;
 			}
+
+			step_option |= F_REAL_STEP;
 		}
+	}
 
-		if (step_option & Z_CAN_STEP) {
-			dda->z_counter -= dda->z_delta;
-			if (dda->z_counter < 0) {
-				step_option |= REAL_MOVE;
-
-				z_step();
-				if (dda->z_direction)
-					current_position.Z++;
-				else
-					current_position.Z--;
-
-				dda->z_counter += dda->total_steps;
-			}
-		}
-
-		if (step_option & E_CAN_STEP) {
-			dda->e_counter -= dda->e_delta;
-			if (dda->e_counter < 0) {
-				step_option |= REAL_MOVE;
-
-				e_step();
-				if (dda->e_direction)
-					current_position.E++;
-				else
-					current_position.E--;
-
-				dda->e_counter += dda->total_steps;
-			}
-		}
-
-		if (step_option & F_CAN_STEP) {
-			dda->f_counter -= dda->f_delta;
-			// since we don't allow total_steps to be defined by F, we may need to step multiple times if f_delta is greater than total_steps
-			while (dda->f_counter < 0) {
-
-				dda->f_counter += dda->total_steps;
-
-				if (dda->f_direction) {
-					current_position.F += 1;
-					if (current_position.F > dda->endpoint.F)
-						current_position.F = dda->endpoint.F;
-				}
-				else {
-					current_position.F -= 1;
-					if (current_position.F < dda->endpoint.F)
-						current_position.F = dda->endpoint.F;
-				}
-
-				step_option |= F_REAL_STEP;
-			}
-		}
-
-		// this generates too much debug output for all but the slowest step rates
-		if (0 && DEBUG) {
-			serial_writechar('[');
-			serwrite_hex8(step_option);
-			serial_writechar(':');
-			serwrite_int32(current_position.F);
-			serial_writechar('/');
-			serwrite_int32(dda->endpoint.F);
-			serial_writechar('#');
-			serwrite_uint32(dda->move_duration);
-			serial_writechar(']');
-		}
-
-// 		WRITE(SCK, 1);
-
-// 	} while (	((step_option & REAL_MOVE ) == 0)	&&
-// 						((step_option & F_CAN_STEP) != 0)	);
-// 	} while (0);
+	// this generates too much debug output for all but the slowest step rates
+	if (0 && DEBUG) {
+		serial_writechar('[');
+		serwrite_hex8(step_option);
+		serial_writechar(':');
+		serwrite_int32(current_position.F);
+		serial_writechar('/');
+		serwrite_int32(dda->endpoint.F);
+		serial_writechar('#');
+		serwrite_uint32(dda->move_duration);
+		serial_writechar(']');
+	}
 
 	if (step_option & REAL_MOVE)
 		// we stepped, reset timeout
@@ -453,8 +413,4 @@ void dda_step(DDA *dda) {
 	// turn off step outputs, hopefully they've been on long enough by now to register with the drivers
 	// if not, too bad. or insert a (very!) small delay here, or fire up a spare timer or something
 	unstep();
-
-	// turn 'L' light back on again
-// 	if (DEBUG)
-// 		WRITE(SCK, 1);
 }
diff --git a/mendel/dda.h b/mendel/dda.h
index 4249810..8a3679a 100644
--- a/mendel/dda.h
+++ b/mendel/dda.h
@@ -10,6 +10,7 @@
 	types
 */
 
+// target is simply a point in space/time
 typedef struct {
 	int32_t						X;
 	int32_t						Y;
@@ -18,8 +19,8 @@ typedef struct {
 	uint32_t					F;
 } TARGET;
 
+// this is a digital differential analyser data struct, holding both the initial values and the counters as it progresses
 typedef struct {
-// 	TARGET						currentpoint;
 	TARGET						endpoint;
 
 	uint8_t						x_direction		:1;
@@ -51,9 +52,13 @@ typedef struct {
 	variables
 */
 
+// steptimeout is set to zero when we step, and increases over time so we can turn the motors off when they've been idle for a while
 extern uint8_t	steptimeout;
 
+// startpoint holds the end position of the most recently created DDA, so we know where the next one starts
 extern TARGET startpoint;
+
+// current_position holds the machine's current position. this is only updated when we step, or when G92 (set home) is received.
 extern TARGET current_position;
 
 /*
diff --git a/mendel/dda_queue.c b/mendel/dda_queue.c
index b2dbc78..b768a08 100644
--- a/mendel/dda_queue.c
+++ b/mendel/dda_queue.c
@@ -8,7 +8,6 @@ uint8_t	mb_head = 0;
 uint8_t	mb_tail = 0;
 DDA movebuffer[MOVEBUFFER_SIZE];
 
-
 uint8_t queue_full() {
 	if (mb_tail == 0)
 		return mb_head == (MOVEBUFFER_SIZE - 1);
@@ -21,6 +20,7 @@ uint8_t queue_empty() {
 }
 
 void enqueue(TARGET *t) {
+	// don't call this function when the queue is full, but just in case, wait for a move to complete and free up the space for the passed target
 	while (queue_full())
 		delay(WAITING_DELAY);
 
@@ -31,24 +31,24 @@ void enqueue(TARGET *t) {
 
 	dda_create(&movebuffer[h], t);
 
-	// if queue only has one space left, stop transmition
-	if (((h + 2) & (MOVEBUFFER_SIZE - 1)) == mb_tail)
-		xoff();
-
 	mb_head = h;
 
+	#ifdef	XONXOFF
+		// if queue is full, stop transmition
+		if (queue_full())
+			xoff();
+	#endif
+
 	// fire up in case we're not running yet
 	enableTimerInterrupt();
 }
 
 void next_move() {
 	if (queue_empty()) {
-		// queue is empty
-// 		disable_steppers();
-// 		setTimer(DEFAULT_TICK);
 		disableTimerInterrupt();
 	}
 	else {
+		// next item
 		uint8_t t = mb_tail;
 		t++;
 		if (t == MOVEBUFFER_SIZE)
@@ -56,8 +56,11 @@ void next_move() {
 		dda_start(&movebuffer[t]);
 		mb_tail = t;
 	}
-	// restart transmission
-	xon();
+
+	#ifdef	XONXOFF
+		// restart transmission
+		xon();
+	#endif
 }
 
 void print_queue() {
diff --git a/mendel/dda_queue.h b/mendel/dda_queue.h
index 86a7185..c3176b0 100644
--- a/mendel/dda_queue.h
+++ b/mendel/dda_queue.h
@@ -7,6 +7,7 @@
 	variables
 */
 
+// this is the ringbuffer that holds the current and pending moves.
 extern uint8_t	mb_head;
 extern uint8_t	mb_tail;
 extern DDA movebuffer[MOVEBUFFER_SIZE];
@@ -15,10 +16,17 @@ extern DDA movebuffer[MOVEBUFFER_SIZE];
 	methods
 */
 
+// queue status methods
 uint8_t queue_full(void);
 uint8_t queue_empty(void);
+
+// add a new target to the queue
 void enqueue(TARGET *t);
+
+// called from step timer when current move is complete
 void next_move(void);
+
+// print queue status
 void print_queue(void);
 
 #endif	/* _DDA_QUEUE */
diff --git a/mendel/gcode.c b/mendel/gcode.c
index 594dc79..12479fa 100644
--- a/mendel/gcode.c
+++ b/mendel/gcode.c
@@ -69,6 +69,10 @@ int32_t	decfloat_to_int(decfloat *df, int32_t multiplicand, int32_t denominator)
 	return r;
 }
 
+/*
+	public functions
+*/
+
 void SpecialMoveXY(int32_t x, int32_t y, uint32_t f) {
 	TARGET t = startpoint;
 	t.X = x;
@@ -91,10 +95,6 @@ void SpecialMoveE(int32_t e, uint32_t f) {
 	enqueue(&t);
 }
 
-/*
-	public functions
-*/
-
 void scan_char(uint8_t c) {
 	static uint8_t last_field = 0;
 
@@ -102,38 +102,61 @@ void scan_char(uint8_t c) {
 	if (c >= 'a' && c <= 'z')
 		c &= ~32;
 
-	// process field
+	// process previous field
 	if (last_field) {
+		// check if we're seeing a new field or end of line
 		if ((indexof(c, alphabet) >= 0) || (c == 10)) {
 			switch (last_field) {
 				case 'G':
 					next_target.G = read_digit.mantissa;
-					serwrite_uint8(next_target.G);
+					if (DEBUG)
+						serwrite_uint8(next_target.G);
 					break;
 				case 'M':
 					next_target.M = read_digit.mantissa;
-					serwrite_uint8(next_target.M);
+					if (DEBUG)
+						serwrite_uint8(next_target.M);
 					break;
 				case 'X':
-					next_target.target.X = decfloat_to_int(&read_digit, STEPS_PER_MM_X, 1);
-					serwrite_int32(next_target.target.X);
+					if (next_target.option_inches)
+						next_target.target.X = decfloat_to_int(&read_digit, STEPS_PER_IN_X, 1);
+					else
+						next_target.target.X = decfloat_to_int(&read_digit, STEPS_PER_MM_X, 1);
+					if (DEBUG)
+						serwrite_int32(next_target.target.X);
 					break;
 				case 'Y':
-					next_target.target.Y = decfloat_to_int(&read_digit, STEPS_PER_MM_Y, 1);
-					serwrite_int32(next_target.target.Y);
+					if (next_target.option_inches)
+						next_target.target.Y = decfloat_to_int(&read_digit, STEPS_PER_IN_Y, 1);
+					else
+						next_target.target.Y = decfloat_to_int(&read_digit, STEPS_PER_MM_Y, 1);
+					if (DEBUG)
+						serwrite_int32(next_target.target.Y);
 					break;
 				case 'Z':
-					next_target.target.Z = decfloat_to_int(&read_digit, STEPS_PER_MM_Z, 1);
-					serwrite_int32(next_target.target.Z);
+					if (next_target.option_inches)
+						next_target.target.Z = decfloat_to_int(&read_digit, STEPS_PER_IN_Z, 1);
+					else
+						next_target.target.Z = decfloat_to_int(&read_digit, STEPS_PER_MM_Z, 1);
+					if (DEBUG)
+						serwrite_int32(next_target.target.Z);
 					break;
 				case 'E':
-					next_target.target.E = decfloat_to_int(&read_digit, STEPS_PER_MM_E, 1);
-					serwrite_uint32(next_target.target.E);
+					if (next_target.option_inches)
+						next_target.target.E = decfloat_to_int(&read_digit, STEPS_PER_IN_E, 1);
+					else
+						next_target.target.E = decfloat_to_int(&read_digit, STEPS_PER_MM_E, 1);
+					if (DEBUG)
+						serwrite_uint32(next_target.target.E);
 					break;
 				case 'F':
 					// just use raw integer, we need move distance and n_steps to convert it to a useful value, so wait until we have those to convert it
-					next_target.target.F = decfloat_to_int(&read_digit, 1, 1);
-					serwrite_uint32(next_target.target.F);
+					if (next_target.option_inches)
+						next_target.target.F = decfloat_to_int(&read_digit, 254, 10);
+					else
+						next_target.target.F = decfloat_to_int(&read_digit, 1, 1);
+					if (DEBUG)
+						serwrite_uint32(next_target.target.F);
 					break;
 				case 'S':
 					// if this is temperature, multiply by 4 to convert to quarter-degree units
@@ -141,9 +164,13 @@ void scan_char(uint8_t c) {
 					// but it takes less code, less memory and loses no precision if we do it here instead
 					if (next_target.M == 104)
 						next_target.S = decfloat_to_int(&read_digit, 4, 1);
+					// if this is heater PID stuff, multiply by PID_SCALE because we divide by PID_SCALE later on
+					else if ((next_target.M >= 130) && (next_target.M <= 132))
+						next_target.S = decfloat_to_int(&read_digit, PID_SCALE, 1);
 					else
 						next_target.S = decfloat_to_int(&read_digit, 1, 1);
-					serwrite_uint16(next_target.S);
+					if (DEBUG)
+						serwrite_uint16(next_target.S);
 					break;
 				case 'P':
 					// if this is dwell, multiply by 1000 to convert seconds to milliseconds
@@ -151,9 +178,11 @@ void scan_char(uint8_t c) {
 						next_target.P = decfloat_to_int(&read_digit, 1000, 1);
 					else
 						next_target.P = decfloat_to_int(&read_digit, 1, 1);
-					serwrite_uint16(next_target.P);
+					if (DEBUG)
+						serwrite_uint16(next_target.P);
 					break;
 			}
+			// reset for next field
 			last_field = 0;
 			read_digit.sign = 0;
 			read_digit.mantissa = 0;
@@ -161,11 +190,13 @@ void scan_char(uint8_t c) {
 		}
 	}
 
-	// not in a comment?
+	// skip comments
 	if ((option_bitfield & OPTION_COMMENT) == 0) {
+		// new field?
 		if (indexof(c, alphabet) >= 0) {
 			last_field = c;
-			serial_writechar(c);
+			if (DEBUG)
+				serial_writechar(c);
 		}
 
 		// process character
@@ -223,6 +254,7 @@ void scan_char(uint8_t c) {
 			default:
 				// can't do ranges in switch..case, so process actual digits here
 				if (c >= '0' && c <= '9') {
+					// this is simply mantissa = (mantissa * 10) + atoi(c) in different clothes
 					read_digit.mantissa = (read_digit.mantissa << 3) + (read_digit.mantissa << 1) + (c - '0');
 					if (read_digit.exponent)
 						read_digit.exponent++;
@@ -230,14 +262,13 @@ void scan_char(uint8_t c) {
 		}
 	}
 
-	// got a command
+	// end of line
 	if (c == 10) {
 		serial_writechar(c);
 		// process
 		process_gcode_command(&next_target);
 
-		// save options
-// 		option_bitfield = next_target.option;
+		// reset 'seen comment'
 		option_bitfield &= ~OPTION_COMMENT;
 
 		// reset variables
@@ -252,7 +283,6 @@ void scan_char(uint8_t c) {
 }
 
 void process_gcode_command(GCODE_COMMAND *gcmd) {
-
 	// convert relative to absolute
 	if (gcmd->option_relative) {
 		gcmd->target.X += startpoint.X;
@@ -261,7 +291,7 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 		gcmd->target.E += startpoint.E;
 	}
 
-	// explicitly make unseen values equal to startpoint, otherwise relative position mode is a clusterfuck and who knows what other bugs could occur?
+	// explicitly make unseen values equal to startpoint, otherwise relative position mode is a clusterfuck
 	if (gcmd->seen_X == 0)
 		gcmd->target.X = startpoint.X;
 	if (gcmd->seen_Y == 0)
@@ -288,14 +318,20 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 				break;
 
 			//	G2 - Arc Clockwise
+			// unimplemented
 
 			//	G3 - Arc Counter-clockwise
+			// unimplemented
 
 			//	G4 - Dwell
 			case 4:
-				xoff();
+				#ifdef	XONXOFF
+					xoff();
+				#endif
 				delay_ms(gcmd->P);
-				xon();
+				#ifdef	XONXOFF
+					xon();
+				#endif
 				break;
 
 			//	G20 - inches as units
@@ -312,6 +348,7 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 			case 30:
 				enqueue(&gcmd->target);
 				// no break here, G30 is move and then go home
+
 			//	G28 - go home
 			case 28:
 				/*
@@ -321,6 +358,7 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 				SpecialMoveXY(-250L * STEPS_PER_MM_X, -250L * STEPS_PER_MM_Y, FEEDRATE_FAST_XY);
 				startpoint.X = startpoint.Y = 0;
 
+				// move forward a bit
 				SpecialMoveXY(5 * STEPS_PER_MM_X, 5 * STEPS_PER_MM_Y, FEEDRATE_SLOW_XY);
 
 				// move back in to endstops slowly
@@ -339,7 +377,7 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 				SpecialMoveZ(-250L * STEPS_PER_MM_Z, FEEDRATE_FAST_Z);
 				startpoint.Z = 0;
 
-				// move out a bit
+				// move forward a bit
 				SpecialMoveZ(5 * STEPS_PER_MM_Z, FEEDRATE_SLOW_Z);
 
 				// move back into endstop slowly
@@ -354,7 +392,7 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 				/*
 					Home E
 				*/
-				// extruder only runs one way anyway
+				// extruder only runs one way and we have no "endstop", just set this point as home
 				startpoint.E = current_position.E = 0;
 
 				break;
@@ -374,7 +412,7 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 				startpoint.X = startpoint.Y = startpoint.Z = startpoint.E = 0;
 				break;
 
-			// TODO: spit an error
+			// unknown gcode: spit an error
 			default:
 				serial_writestr_P(PSTR("E: Bad G-code "));
 				serwrite_uint8(gcmd->G);
@@ -435,14 +473,29 @@ void process_gcode_command(GCODE_COMMAND *gcmd) {
 			case 106:
 				WRITE(FAN_PIN, 1);
 				break;
-
 			// M107- fan off
 			case 107:
 				WRITE(FAN_PIN, 0);
 				break;
 			#endif
 
-			// TODO: spit an error
+			// M130- heater P factor
+			case 130:
+				if (gcmd->seen_S)
+					p_factor = gcmd->S;
+				break;
+			// M131- heater I factor
+			case 131:
+				if (gcmd->seen_S)
+					i_factor = gcmd->S;
+				break;
+			// M132- heater D factor
+			case 132:
+				if (gcmd->seen_S)
+					d_factor = gcmd->S;
+				break;
+
+			// unknown mcode: spit an error
 			default:
 				serial_writestr_P(PSTR("E: Bad M-code "));
 				serwrite_uint8(gcmd->M);
diff --git a/mendel/gcode.h b/mendel/gcode.h
index 4a7ff24..b0187ea 100644
--- a/mendel/gcode.h
+++ b/mendel/gcode.h
@@ -5,12 +5,14 @@
 
 #include	"dda.h"
 
+// this is a very crude decimal-based floating point structure. a real floating point would at least have signed exponent
 typedef struct {
 	uint32_t	sign			:1;
 	uint32_t	mantissa	:24;
 	uint32_t	exponent	:7;
 } decfloat;
 
+// this holds all the possible data from a received command
 typedef struct {
 	uint8_t					seen_G	:1;
 	uint8_t					seen_M	:1;
@@ -30,16 +32,26 @@ typedef struct {
 	uint8_t						M;
 	TARGET						target;
 
-	uint16_t					S;
+	int16_t						S;
 	uint16_t					P;
 } GCODE_COMMAND;
 
+// the command being processed
 extern GCODE_COMMAND next_target;
 
+// utility functions
 int8_t indexof(uint8_t c, const char *string);
 int32_t	decfloat_to_int(decfloat *df, int32_t multiplicand, int32_t denominator);
 
+// this is where we construct a move without a gcode command, useful for gcodes which require multiple moves eg; homing
+void SpecialMoveXY(int32_t x, int32_t y, uint32_t f);
+void SpecialMoveZ(int32_t z, uint32_t f);
+void SpecialMoveE(int32_t e, uint32_t f);
+
+// accept the next character and process it
 void scan_char(uint8_t c);
+
+// when we have a whole line, feed it to this
 void process_gcode_command(GCODE_COMMAND *gcmd);
 
 #endif	/* _GCODE_H */
diff --git a/mendel/machine.h b/mendel/machine.h
index 9e29dc6..f3225f5 100644
--- a/mendel/machine.h
+++ b/mendel/machine.h
@@ -3,8 +3,8 @@
 
 /*
 	move buffer size, in number of moves
-		note that each move takes a fair chunk of ram (71 bytes as of this writing) so don't make the buffer too big - a bigger serial readbuffer may help more than increasing this unless your gcodes are more than 70 characters long on average.
-		however, a larger movebuffer will probably help with lots of short moves, as each move takes a bunch of math to set up
+		note that each move takes a fair chunk of ram (69 bytes as of this writing) so don't make the buffer too big - a bigger serial readbuffer may help more than increasing this unless your gcodes are more than 70 characters long on average.
+		however, a larger movebuffer will probably help with lots of short consecutive moves, as each move takes a bunch of math (hence time) to set up
 */
 #define	MOVEBUFFER_SIZE	8
 
@@ -12,9 +12,11 @@
 	axis calculations, adjust as necessary
 */
 
+// XY can have lots of precision and still move speedily, so microstepping is helpful
 #define	X_STEPS_PER_REV						3200.0
 #define	Y_STEPS_PER_REV						X_STEPS_PER_REV
-// we need more speed than precision on Z, turn off microstepping
+
+// we need far more speed than precision on Z due to the threaded rod drive, turn off microstepping
 #define	Z_STEPS_PER_REV						200.0
 
 #define	X_COG_CIRCUMFERENCE				(4.77 * 16.0)
@@ -22,6 +24,8 @@
 // also try:
 // #define	XY_COG_RADIUS			9.5
 // #define	XY_COG_CIRCUMFERENCE	(XY_COG_RADIUS * PI * 2)
+
+// this is the ratio between number of teeth on the Z motor gear, and teeth on the Z leadscrew base gear.
 #define	Z_GEAR_RATIO							1.0
 
 // we need more torque and smoothness at very low speeds on E, maximum microstepping
@@ -30,6 +34,26 @@
 #define	EXTRUDER_INLET_DIAMETER		3.0
 #define	EXTRUDER_NOZZLE_DIAMETER	0.8
 
+// these feedrates are used during homing and G0 rapid moves
+#define	FEEDRATE_FAST_XY					6000
+#define	FEEDRATE_SLOW_XY					300
+
+#define	FEEDRATE_FAST_Z						6000
+#define	FEEDRATE_SLOW_Z						300
+
+#define	FEEDRATE_FAST_E						1200
+
+// this is how many steps to suck back the filament by when we stop
+#define	E_STARTSTOP_STEPS					20
+
+// extruder settings
+#define	TEMP_HYSTERESIS						2
+
+/*
+	calculated values - you shouldn't need to touch these
+	however feel free to put in your own values if they can be more precise than the calculated approximations, remembering that they must end up being integers- floating point by preprocessor only thanks!
+*/
+
 #define	STEPS_PER_MM_X						((uint32_t) ((X_STEPS_PER_REV / X_COG_CIRCUMFERENCE) + 0.5))
 #define	STEPS_PER_MM_Y						((uint32_t) ((Y_STEPS_PER_REV / Y_COG_CIRCUMFERENCE) + 0.5))
 #define	STEPS_PER_MM_Z						((uint32_t) ((Z_STEPS_PER_REV * Z_GEAR_RATIO) + 0.5))
@@ -39,33 +63,19 @@
 // does this refer to filament or extrudate? extrudate depends on XY distance vs E distance.. hm lets go with filament
 #define	STEPS_PER_MM_E						((uint32_t) ((E_STEPS_PER_REV / (EXTRUDER_SHAFT_RADIUS * PI * EXTRUDER_INLET_DIAMETER)) + 0.5))
 
-#define	FEEDRATE_FAST_XY					6000
-#define	FEEDRATE_SLOW_XY					300
-#define	FEEDRATE_FAST_Z						6000
-#define	FEEDRATE_SLOW_Z						300
-
-#define	E_STARTSTOP_STEPS					20
-#define	FEEDRATE_FAST_E						1200
-
-/*
-	extruder settings
-*/
-
-#define	TEMP_HYSTERESIS						2
-
-/*
-	calculated values - you shouldn't need to touch these
-	however feel free to put in your own values if they can be more precise than the calculated approximations, remembering that they must end up being integers- floating point by preprocessor only thanks!
-*/
+// same as above with 25.4 scale factor
+#define	STEPS_PER_IN_X						((uint32_t) ((25.4 * X_STEPS_PER_REV / X_COG_CIRCUMFERENCE) + 0.5))
+#define	STEPS_PER_IN_Y						((uint32_t) ((25.4 * Y_STEPS_PER_REV / Y_COG_CIRCUMFERENCE) + 0.5))
+#define	STEPS_PER_IN_Z						((uint32_t) ((25.4 * Z_STEPS_PER_REV * Z_GEAR_RATIO) + 0.5))
+#define	STEPS_PER_IN_E						((uint32_t) ((25.4 * E_STEPS_PER_REV / (EXTRUDER_SHAFT_RADIUS * PI * EXTRUDER_INLET_DIAMETER)) + 0.5))
 
+// inverse, used in distance calculation during DDA setup
 #define	UM_PER_STEP_X			((uint32_t) ((1000.0 / STEPS_PER_MM_X) + 0.5))
 #define	UM_PER_STEP_Y			((uint32_t) ((1000.0 / STEPS_PER_MM_Y) + 0.5))
 #define	UM_PER_STEP_Z			((uint32_t) ((1000.0 / STEPS_PER_MM_Z) + 0.5))
 #define	UM_PER_STEP_E			((uint32_t) ((1000.0 / STEPS_PER_MM_E) + 0.5))
 
-/*
-	should be the same for all machines! ;)
-*/
+// should be the same for all machines! ;)
 #define	PI	3.1415926535
 
 #endif	/* _MACHINE_H */
diff --git a/mendel/mendel.c b/mendel/mendel.c
index 901e433..eecd4a9 100644
--- a/mendel/mendel.c
+++ b/mendel/mendel.c
@@ -72,8 +72,6 @@ inline void init(void) {
 	clock_setup();
 
 	// set up variables
-
-	// slow default
 	current_position.F = FEEDRATE_SLOW_Z;
 	memcpy(&startpoint, &current_position, sizeof(TARGET));
 	memcpy(&(next_target.target), &current_position, sizeof(TARGET));
@@ -83,14 +81,10 @@ inline void init(void) {
 
 	// say hi to host
 	serial_writestr_P(PSTR("Start\n"));
-
-	// start queue
-	//enableTimerInterrupt();
 }
 
+void clock_250ms(void);
 void clock_250ms() {
-	static uint8_t report = 0;
-
 	temp_tick();
 
 	if (steptimeout > (30 * 4))
@@ -98,18 +92,13 @@ void clock_250ms() {
 	else
 		steptimeout++;
 
-	report++;
-	if (report == 4) {
-		report = 0;
-
+	ifclock (CLOCK_FLAG_1S) {
 		if (DEBUG) {
 			// current move
 			serial_writestr_P(PSTR("DDA: f#"));
-			serwrite_int32(movebuffer[mb_head].f_counter);
+			serwrite_int32(movebuffer[mb_tail].f_counter);
 			serial_writechar('/');
-			// serwrite_uint16(movebuffer[mb_head].f_scale);
-			// serial_writechar('/');
-			serwrite_int16(movebuffer[mb_head].f_delta);
+			serwrite_int16(movebuffer[mb_tail].f_delta);
 			serial_writechar('\n');
 
 			// current position
@@ -155,7 +144,7 @@ int main (void)
 	for (;;)
 	{
 		// if queue is full, no point in reading chars- host will just have to wait
-		if (serial_rxchars() && !queue_full()) {
+		if ((serial_rxchars() != 0) && (queue_full() == 0)) {
 			uint8_t c = serial_popchar();
 			scan_char(c);
 		}
@@ -163,66 +152,5 @@ int main (void)
 		ifclock(CLOCK_FLAG_250MS) {
 			clock_250ms();
 		}
-
-// 		ifclock(CLOCK_FLAG_250MS_TEMPCHECK) {
-// 			temp_tick();
-// 		}
-//
-// 		ifclock(CLOCK_FLAG_250MS_STEPTIMEOUT) {
-// 			if (steptimeout > (30 * 4))
-// 				disable_steppers();
-// 			else
-// 				steptimeout++;
-// 		}
-//
-// 		ifclock(CLOCK_FLAG_250MS_REPORT) {
-// 			report++;
-// 			if (report == 4) {
-// 				report = 0;
-//
-// 				if (DEBUG) {
-// 					// current move
-// 					serial_writestr_P(PSTR("DDA: f#"));
-// 					serwrite_int32(movebuffer[mb_head].f_counter);
-// 					serial_writechar('/');
-// // 					serwrite_uint16(movebuffer[mb_head].f_scale);
-// // 					serial_writechar('/');
-// 					serwrite_int16(movebuffer[mb_head].f_delta);
-// 					serial_writechar('\n');
-//
-// 					// current position
-// 					serial_writestr_P(PSTR("Pos: "));
-// 					serwrite_int32(current_position.X);
-// 					serial_writechar(',');
-// 					serwrite_int32(current_position.Y);
-// 					serial_writechar(',');
-// 					serwrite_int32(current_position.Z);
-// 					serial_writechar(',');
-// 					serwrite_int32(current_position.E);
-// 					serial_writechar(',');
-// 					serwrite_uint32(current_position.F);
-// 					serial_writechar('\n');
-//
-// 					// target position
-// 					serial_writestr_P(PSTR("Dst: "));
-// 					serwrite_int32(movebuffer[mb_tail].endpoint.X);
-// 					serial_writechar(',');
-// 					serwrite_int32(movebuffer[mb_tail].endpoint.Y);
-// 					serial_writechar(',');
-// 					serwrite_int32(movebuffer[mb_tail].endpoint.Z);
-// 					serial_writechar(',');
-// 					serwrite_int32(movebuffer[mb_tail].endpoint.E);
-// 					serial_writechar(',');
-// 					serwrite_uint32(movebuffer[mb_tail].endpoint.F);
-// 					serial_writechar('\n');
-// 				}
-//
-// 				// Queue
-// 				print_queue();
-//
-// 				// temperature
-// 				temp_print();
-// 			}
-// 		}
 	}
 }
diff --git a/mendel/pinout.h b/mendel/pinout.h
index 246a71d..4d0a687 100644
--- a/mendel/pinout.h
+++ b/mendel/pinout.h
@@ -39,7 +39,7 @@
 #define	STEPPER_ENABLE_PIN	DIO9
 
 // list of PWM-able pins and corresponding timers
-// timer1 is reserved for step timing
+// timer1 is used for step timing so don't use OC1A/OC1B (DIO9/DIO10)
 // OC0A											DIO6
 // OC0B											DIO5
 // OC1A											DIO9
@@ -132,7 +132,7 @@
 */
 
 #ifdef	STEPPER_ENABLE_PIN
-	// for connection to stepper driver ENABLE pins
+	// for connection to stepper driver ENABLE pins (negative asserted)
 // 	#define	enable_steppers()		WRITE(STEPPER_ENABLE_PIN, 0)
 // 	#define	disable_steppers()	WRITE(STEPPER_ENABLE_PIN, 1)
 	// for connection to ATX PSU PWR_ON signal
diff --git a/mendel/ringbuffer.c b/mendel/ringbuffer.c
index de4ce94..2f2636e 100644
--- a/mendel/ringbuffer.c
+++ b/mendel/ringbuffer.c
@@ -43,7 +43,6 @@ void ringbuffer_writechar(ringbuffer *buf, uint8_t data)
 	}
 }
 
-
 uint8_t ringbuffer_peekchar(ringbuffer *buf, RB_BITS index)
 {
 	return buf->data[_rb_mod(buf->read_pointer + index, buf->size)];
diff --git a/mendel/ringbuffer.h b/mendel/ringbuffer.h
index d618d3b..7635ed8 100644
--- a/mendel/ringbuffer.h
+++ b/mendel/ringbuffer.h
@@ -4,6 +4,7 @@
 #include	<stdint.h>
 #include	<avr/interrupt.h>
 
+// ringbuffer head/tail/length precision. change to uint16_t if you want a buffer bigger than 252 bytes or so
 #define	RB_BITS	uint8_t
 
 typedef struct {
@@ -13,15 +14,19 @@ typedef struct {
 	volatile RB_BITS		data[];
 } ringbuffer;
 
+// initialize a ringbuffer
 void ringbuffer_init(ringbuffer *buf, RB_BITS bufsize);
 
+// return how many bytes can be read or written
 RB_BITS ringbuffer_canread(ringbuffer *buf);
 RB_BITS ringbuffer_canwrite(ringbuffer *buf);
 
+// read bytes
 uint8_t ringbuffer_readchar(ringbuffer *buf);
 uint8_t ringbuffer_peekchar(ringbuffer *buf, RB_BITS index);
 RB_BITS ringbuffer_readblock(ringbuffer *buf, uint8_t *newbuf, RB_BITS size);
 
+// write bytes
 void ringbuffer_writechar(ringbuffer *buf, uint8_t data);
 RB_BITS ringbuffer_writeblock(ringbuffer *buf, uint8_t *data, RB_BITS size);
 
diff --git a/mendel/serial.c b/mendel/serial.c
index d895e82..c431bce 100644
--- a/mendel/serial.c
+++ b/mendel/serial.c
@@ -121,15 +121,17 @@ void serial_writestr_P(PGM_P data)
 		serial_writechar(r);
 }
 
-void xoff() {
-	flowflags = FLOWFLAG_SEND_XOFF;
-	// enable TX interrupt so we can send this character
-	UCSR0B |= (1 << UDRIE0);
-}
+#ifdef	XONXOFF
+	void xon() {
+		if (flowflags & FLOWFLAG_SENT_XOFF)
+			flowflags = FLOWFLAG_SEND_XON;
+		// enable TX interrupt so we can send this character
+		UCSR0B |= (1 << UDRIE0);
+	}
 
-void xon() {
-	if (flowflags & FLOWFLAG_SENT_XOFF)
-		flowflags = FLOWFLAG_SEND_XON;
-	// enable TX interrupt so we can send this character
-	UCSR0B |= (1 << UDRIE0);
-}
+	void xoff() {
+		flowflags = FLOWFLAG_SEND_XOFF;
+		// enable TX interrupt so we can send this character
+		UCSR0B |= (1 << UDRIE0);
+	}
+#endif
diff --git a/mendel/serial.h b/mendel/serial.h
index 19597f7..2abc440 100644
--- a/mendel/serial.h
+++ b/mendel/serial.h
@@ -5,23 +5,31 @@
 #include	<avr/io.h>
 #include	<avr/pgmspace.h>
 
+// initialise serial subsystem
 void serial_init(void);
 
+// return number of characters in the receive and send buffer
 uint16_t serial_rxchars(void);
 uint16_t serial_txchars(void);
 
+// read one character
 uint8_t serial_popchar(void);
+// send one character
 void serial_writechar(uint8_t data);
 
+// read/write many characters
 uint16_t serial_recvblock(uint8_t *block, int blocksize);
 void serial_writeblock(void *data, int datalen);
 
+// write from flash
 void serial_writechar_P(PGM_P data);
 void serial_writeblock_P(PGM_P data, int datalen);
-
 void serial_writestr_P(PGM_P data);
 
-void xoff(void);
-void xon(void);
+#ifdef	XONXOFF
+	// XON/XOFF flow control
+	void xoff(void);
+	void xon(void);
+#endif
 
 #endif	/* _SERIAL_H */
diff --git a/mendel/sermsg.h b/mendel/sermsg.h
index 647a053..5566f66 100644
--- a/mendel/sermsg.h
+++ b/mendel/sermsg.h
@@ -3,11 +3,13 @@
 
 #include	<stdint.h>
 
+// functions for sending hexadecimal
 void serwrite_hex4(uint8_t v);
 void serwrite_hex8(uint8_t v);
 void serwrite_hex16(uint16_t v);
 void serwrite_hex32(uint32_t v);
 
+// functions for sending decimal
 #define	serwrite_uint8(v)		serwrite_uint32(v)
 #define	serwrite_int8(v)		serwrite_int32(v)
 #define	serwrite_uint16(v)	serwrite_uint32(v)
diff --git a/mendel/temp.c b/mendel/temp.c
index 524f18e..56f9c86 100644
--- a/mendel/temp.c
+++ b/mendel/temp.c
@@ -39,8 +39,6 @@ uint8_t		temp_flags		= 0;
 #define		TEMP_FLAG_PRESENT		1
 #define		TEMP_FLAG_TCOPEN		2
 
-#define		PID_SCALE			1024L
-
 uint16_t temp_read() {
 	uint16_t temp;
 	SPCR = MASK(MSTR) | MASK(SPE);
@@ -55,6 +53,10 @@ uint16_t temp_read() {
 	for (;(SPSR & MASK(SPIF)) == 0;);
 	temp |= SPDR;
 
+	WRITE(SS, 0);
+
+	SPCR = 0;
+
 	temp_flags = 0;
 	if ((temp & 0x8002) == 0) {
 		// got "device id"
@@ -69,10 +71,6 @@ uint16_t temp_read() {
 		}
 	}
 
-	WRITE(SS, 0);
-
-	SPCR = 0;
-
 	return 0;
 }
 
@@ -110,15 +108,26 @@ void temp_tick() {
 	uint16_t last_temp = current_temp;
 	temp_read();
 
-	int16_t	t_delta = target_temp - current_temp;
+	int16_t	t_error = target_temp - current_temp;
 
 	// PID stuff
-	heater_p = t_delta;
-	heater_i += t_delta;
-	// note: D follows temp rather than error so there's no large derivative when the target temperature changes
+	// proportional
+	heater_p = t_error;
+
+	// integral
+	heater_i += t_error;
+	// prevent integrator wind-up
+	if (heater_i > I_LIMIT)
+		heater_i = I_LIMIT;
+	else if (heater_i < -I_LIMIT)
+		heater_i = -I_LIMIT;
+
+	// derivative
+	// note: D follows temp rather than error so there's no large derivative when the target changes
 	heater_d = (current_temp - last_temp);
 
-	uint8_t pid_output = (
+	// combine factors
+	int32_t pid_output_intermed = (
 			(
 				(((int32_t) heater_p) * p_factor) +
 				(((int32_t) heater_i) * i_factor) +
@@ -126,6 +135,15 @@ void temp_tick() {
 			) / PID_SCALE
 		);
 
+	// rebase and limit factors
+	uint8_t pid_output;
+	if (pid_output_intermed > 127)
+		pid_output = 255;
+	else if (pid_output_intermed < -128)
+		pid_output = 0;
+	else
+		pid_output = (pid_output_intermed + 128);
+
 #ifdef	HEATER_PIN_PWMABLE
 	HEATER_PIN_PWMABLE = pid_output
 #else
diff --git a/mendel/temp.h b/mendel/temp.h
index 7b8312a..4ca4dbb 100644
--- a/mendel/temp.h
+++ b/mendel/temp.h
@@ -3,11 +3,31 @@
 
 #include	<stdint.h>
 
+// extruder heater PID factors
+// google "PID without a PHD" if you don't understand this PID stuff
+extern int32_t p_factor;
+extern int32_t i_factor;
+extern int32_t d_factor;
+
+#define		PID_SCALE			1024L
+#define		I_LIMIT				4000
+
+// read temperature from sensor
 uint16_t temp_read(void);
+
+// set target temperature
 void temp_set(uint16_t t);
+
+// return last read temperature
 uint16_t temp_get(void);
+
+// true if last read temp is close to target temp, false otherwise
 uint8_t	temp_achieved(void);
+
+// send current temperature to host
 void temp_print(void);
+
+// periodically read temperature and update heater with PID
 void temp_tick(void);
 
 #endif	/* _TIMER_H */
diff --git a/mendel/timer.c b/mendel/timer.c
index 0dce8f9..de63983 100644
--- a/mendel/timer.c
+++ b/mendel/timer.c
@@ -7,7 +7,6 @@
 #include	"dda.h"
 
 ISR(TIMER1_COMPA_vect) {
-// 	WRITE(SCK, 0);
 	if (movebuffer[mb_tail].live) {
 		// this interrupt can be interruptible
 		// TODO: remove when not debugging
@@ -19,19 +18,14 @@ ISR(TIMER1_COMPA_vect) {
 // 		cli();
 // 		enableTimerInterrupt();
 	}
-// 	WRITE(SCK, 1);
 
-	// perhaps we can fall directly into dda_start instead of waiting for another step
-	if (movebuffer[mb_tail].live == 0) {
+	// fall directly into dda_start instead of waiting for another step
+	if (movebuffer[mb_tail].live == 0)
 		next_move();
-	}
-
 }
 
 void setupTimerInterrupt()
 {
-	//clear the registers
-
 	// no outputs
 	TCCR1A = 0;
 	// CTC mode
@@ -43,6 +37,8 @@ void setupTimerInterrupt()
 	setTimer(10000);
 }
 
+// the following are all from reprap project 5D firmware
+
 uint8_t getTimerResolution(const uint32_t delay)
 {
 	// these also represent frequency: 1000000 / delay / 2 = frequency in hz.
@@ -142,7 +138,6 @@ void delay_ms(uint32_t delay) {
 	delayMicrosecondsInterruptible(delay * 1000);
 }
 
-// from reprap project 5D firmware
 void delayMicrosecondsInterruptible(uint16_t us)
 {
   // for a one-microsecond delay, simply return.  the overhead
diff --git a/mendel/timer.h b/mendel/timer.h
index 27410c3..a85ee1d 100644
--- a/mendel/timer.h
+++ b/mendel/timer.h
@@ -8,13 +8,16 @@
 #define	US	* (F_CPU / 1000000)
 #define	MS	* (F_CPU / 1000)
 
-#define	DEFAULT_TICK	(100 US)
+// #define	DEFAULT_TICK	(100 US)
 #define	WAITING_DELAY	(10 MS)
 
 void setupTimerInterrupt(void);
+
 uint8_t getTimerResolution(const uint32_t delay);
 void setTimerResolution(uint8_t r);
+
 uint16_t getTimerCeiling(const uint32_t delay);
+#define setTimerCeiling(c)		OCR1A = c
 
 void setTimer(uint32_t delay);
 
@@ -28,6 +31,4 @@ void delayMicrosecondsInterruptible(unsigned int us);
 #define enableTimerInterrupt()	do { TIMSK1 |= (1<<OCIE1A); } while (0)
 #define disableTimerInterrupt() do { TIMSK1 &= ~(1<<OCIE1A); } while (0)
 
-#define setTimerCeiling(c)		OCR1A = c
-
 #endif	/* _TIMER_H */