diff --git a/esphttpdconfig.mk b/esphttpdconfig.mk
index e29783b..16155c4 100644
--- a/esphttpdconfig.mk
+++ b/esphttpdconfig.mk
@@ -51,4 +51,4 @@ GLOBAL_CFLAGS = \
     -DDEBUG_ANSI_NOIMPL=0 \
     -DHTTPD_MAX_BACKLOG_SIZE=8192 \
     -DDEBUG_INPUT=0 \
-    -DDEBUG_HEAP=1 \
+    -DDEBUG_HEAP=1
diff --git a/user/ansi_parser_callbacks.c b/user/ansi_parser_callbacks.c
index 35c9ea6..58a121f 100644
--- a/user/ansi_parser_callbacks.c
+++ b/user/ansi_parser_callbacks.c
@@ -10,6 +10,7 @@
 #include "uart_driver.h"
 #include "cgi_sockets.h"
 #include "version.h"
+#include "uart_buffer.h"
 
 /**
  * Send a response to UART0
@@ -18,7 +19,8 @@
 void ICACHE_FLASH_ATTR
 apars_respond(const char *str)
 {
-	UART_WriteString(UART0, str, UART_TIMEOUT_US);
+	UART_SendAsync(str, -1);
+	//UART_WriteString(UART0, str, UART_TIMEOUT_US);
 }
 
 /**
diff --git a/user/ascii.h b/user/ascii.h
index 55f5cdd..ac191f5 100644
--- a/user/ascii.h
+++ b/user/ascii.h
@@ -40,6 +40,10 @@ enum ASCII_CODES {
 	US = 31,
 	SP = 32,
 	DEL = 127,
+
+	// aliases
+	XON = DC1,
+	XOFF = DC3,
 };
 
 #endif //ESP_VT100_FIRMWARE_ASCII_H
diff --git a/user/cgi_sockets.c b/user/cgi_sockets.c
index 9a49cbc..a50e68f 100644
--- a/user/cgi_sockets.c
+++ b/user/cgi_sockets.c
@@ -5,6 +5,7 @@
 #include "cgi_sockets.h"
 #include "uart_driver.h"
 #include "screen.h"
+#include "uart_buffer.h"
 
 #define SOCK_BUF_LEN 1024
 static char sock_buff[SOCK_BUF_LEN];
@@ -113,13 +114,13 @@ void ICACHE_FLASH_ATTR updateSockRx(Websock *ws, char *data, int len, int flags)
 
 	if (strstarts(data, "STR:")) {
 		// pass string verbatim
-		UART_WriteString(UART0, data+4, UART_TIMEOUT_US);
+		UART_SendAsync(data+4, -1);
 	}
 	else if (strstarts(data, "BTN:")) {
 		// send button as low ASCII value 1-9
-		int btnNum = data[4] - '0';
+		u8 btnNum = (u8) (data[4] - '0');
 		if (btnNum > 0 && btnNum < 10) {
-			UART_WriteChar(UART0, (unsigned char)btnNum, UART_TIMEOUT_US);
+			UART_SendAsync((const char *) &btnNum, 1);
 		}
 	}
 	else if (strstarts(data, "TAP:")) {
diff --git a/user/serial.c b/user/serial.c
index 056a86c..67e7ba3 100644
--- a/user/serial.c
+++ b/user/serial.c
@@ -30,7 +30,7 @@ static void buf_pop(void *unused)
 		lb_ls++;
 		if (lb_ls >= LOGBUF_SIZE) lb_ls = 0;
 
-		if (OK == UART_WriteCharCRLF(UART1, logbuf[lb_ls], 2000)) {
+		if (OK == UART_WriteCharCRLF(UART1, logbuf[lb_ls], 1000)) {
 			quantity--;
 		} else {
 			// try another time
@@ -40,6 +40,11 @@ static void buf_pop(void *unused)
 	}
 }
 
+LOCAL void my_putc(char c)
+{
+	UART_WriteCharCRLF(UART1, (u8) c, 10);
+}
+
 /**
  * Init the serial ports
  */
@@ -50,8 +55,9 @@ void ICACHE_FLASH_ATTR serialInitBase(void)
 	UART_SetParity(UART1, PARITY_NONE);
 	UART_SetStopBits(UART1, ONE_STOP_BIT);
 	UART_SetBaudrate(UART1, BIT_RATE_115200);
-	//UART_SetPrintPort(UART1);
+	UART_SetPrintPort(UART1);
 	os_install_putc1(buf_putc);
+	//os_install_putc1(my_putc);
 	UART_SetupAsyncReceiver();
 
 	// 1 ms timer
diff --git a/user/uart_buffer.c b/user/uart_buffer.c
index 70965bc..952fde1 100644
--- a/user/uart_buffer.c
+++ b/user/uart_buffer.c
@@ -8,7 +8,7 @@
 #include <esp8266.h>
 #include <uart_register.h>
 
-#define UART_TX_BUFFER_SIZE 512  //Ring buffer length of tx buffer
+#define UART_TX_BUFFER_SIZE 256  //Ring buffer length of tx buffer
 #define UART_RX_BUFFER_SIZE 512 //Ring buffer length of rx buffer
 
 struct UartBuffer {
@@ -83,7 +83,6 @@ static void UART_WriteToAsyncBuffer(struct UartBuffer *pCur, const char *pdata,
 		pCur->pInPos = (pCur->pUartBuff + (pCur->pInPos - pCur->pUartBuff) % pCur->UartBuffSize);
 		pCur->Space -= (data_len - tail_len);
 	}
-
 }
 
 /******************************************************************************
@@ -98,6 +97,11 @@ void ICACHE_FLASH_ATTR UART_FreeAsyncBuffer(struct UartBuffer *pBuff)
 	os_free(pBuff);
 }
 
+u16 ICACHE_FLASH_ATTR UART_AsyncRxCount(void)
+{
+	return (u16) (pRxBuffer->UartBuffSize - pRxBuffer->Space);
+}
+
 /**
  * Retrieve some data from the RX buffer
  * @param pdata - target
@@ -139,6 +143,7 @@ UART_ReadAsync(char *pdata, uint16 data_len)
 			pRxBuffer->Space += len_tmp;
 		}
 	}
+	// this maybe shouldnt be here??
 	if (pRxBuffer->Space >= UART_FIFO_LEN) {
 		uart_rx_intr_enable(UART0);
 	}
@@ -146,13 +151,13 @@ UART_ReadAsync(char *pdata, uint16 data_len)
 }
 
 //move data from uart fifo to rx buffer
-void UART_RxFifoDeq(void)
+void UART_RxFifoCollect(void)
 {
 	uint8 fifo_len, buf_idx;
 	uint8 fifo_data;
 	fifo_len = (uint8) ((READ_PERI_REG(UART_STATUS(UART0)) >> UART_RXFIFO_CNT_S) & UART_RXFIFO_CNT);
 	if (fifo_len >= pRxBuffer->Space) {
-		os_printf("buf full!!!\n\r");
+		UART_WriteChar(UART1, '%', 100);
 	}
 	else {
 		buf_idx = 0;
@@ -165,8 +170,8 @@ void UART_RxFifoDeq(void)
 			}
 		}
 		pRxBuffer->Space -= fifo_len;
+
 		if (pRxBuffer->Space >= UART_FIFO_LEN) {
-			//os_printf("after rx enq buf enough\n\r");
 			uart_rx_intr_enable(UART0);
 		}
 	}
@@ -175,29 +180,32 @@ void UART_RxFifoDeq(void)
 /**
  * Schedule data to be sent
  * @param pdata
- * @param data_len
+ * @param data_len - can be -1 for strlen
  */
 void ICACHE_FLASH_ATTR
-UART_SendAsync(char *pdata, uint16 data_len)
+UART_SendAsync(const char *pdata, int16_t data_len)
 {
-	if (pTxBuffer == NULL) {
-		info("\n\rnull, create buffer struct\n\r");
-		pTxBuffer = UART_AsyncBufferInit(UART_TX_BUFFER_SIZE);
-		if (pTxBuffer != NULL) {
-			UART_WriteToAsyncBuffer(pTxBuffer, pdata, data_len);
-		}
-		else {
-			error("uart tx MALLOC no buf \n\r");
-		}
-	}
-	else {
-		if (data_len <= pTxBuffer->Space) {
-			UART_WriteToAsyncBuffer(pTxBuffer, pdata, data_len);
+	u16 real_len = (u16) data_len;
+	if (data_len <= 0) real_len = (u16) strlen(pdata);
+
+//	if (pTxBuffer == NULL) {
+//		printf("init tx buf\n\r");
+//		pTxBuffer = UART_AsyncBufferInit(UART_TX_BUFFER_SIZE);
+//		if (pTxBuffer != NULL) {
+//			UART_WriteToAsyncBuffer(pTxBuffer, pdata, real_len);
+//		}
+//		else {
+//			printf("tx alloc fail\r\n");
+//		}
+//	}
+//	else {
+		if (real_len <= pTxBuffer->Space) {
+			UART_WriteToAsyncBuffer(pTxBuffer, pdata, real_len);
 		}
 		else {
-			error("UART TX BUF FULL!!!!\n\r");
+			UART_WriteChar(UART1, '^', 100);
 		}
-	}
+//	}
 
 	// Here we enable TX empty interrupt that will take care of sending the content
 	SET_PERI_REG_MASK(UART_CONF1(UART0), (UART_TX_EMPTY_THRESH_VAL & UART_TXFIFO_EMPTY_THRHD) << UART_TXFIFO_EMPTY_THRHD_S);
@@ -233,7 +241,7 @@ void UART_DispatchFromTxBuffer(uint8 uart_no)
 	uint8 len_tmp;
 	uint16 data_len;
 
-	if (pTxBuffer) {
+//	if (pTxBuffer) {
 		data_len = (uint8) (pTxBuffer->UartBuffSize - pTxBuffer->Space);
 		if (data_len > fifo_remain) {
 			len_tmp = fifo_remain;
@@ -244,8 +252,8 @@ void UART_DispatchFromTxBuffer(uint8 uart_no)
 			len_tmp = (uint8) data_len;
 			UART_TxFifoEnq(pTxBuffer, len_tmp, uart_no);
 		}
-	}
-	else {
-		error("pTxBuff null \n\r");
-	}
+//	}
+//	else {
+//		error("pTxBuff null \n\r");
+//	}
 }
diff --git a/user/uart_buffer.h b/user/uart_buffer.h
index f8beedf..f11cfd4 100644
--- a/user/uart_buffer.h
+++ b/user/uart_buffer.h
@@ -8,17 +8,19 @@
 #include <esp8266.h>
 
 // the init func
-void ICACHE_FLASH_ATTR UART_AllocBuffers(void);
+void UART_AllocBuffers(void);
 
 // read from rx buffer
-uint16 ICACHE_FLASH_ATTR UART_ReadAsync(char *pdata, uint16 data_len);
+uint16 UART_ReadAsync(char *pdata, uint16 data_len);
 
 // write to tx buffer
-void ICACHE_FLASH_ATTR UART_SendAsync(char *pdata, uint16 data_len);
+void UART_SendAsync(const char *pdata, int16_t data_len);
 
 //move data from uart fifo to rx buffer
-void UART_RxFifoDeq(void);
+void UART_RxFifoCollect(void);
 //move data from uart tx buffer to fifo
 void UART_DispatchFromTxBuffer(uint8 uart_no);
 
+u16 UART_AsyncRxCount(void);
+
 #endif //ESP_VT100_FIRMWARE_UART_BUFFER_H
diff --git a/user/uart_driver.c b/user/uart_driver.c
index 9425113..a210668 100755
--- a/user/uart_driver.c
+++ b/user/uart_driver.c
@@ -84,7 +84,7 @@ bool ICACHE_FLASH_ATTR UART_CheckOutputFinished(UARTn uart_no, uint32 time_out_u
 	uint8 tx_fifo_len;
 
 	while (1) {
-		tx_fifo_len = UART_TxQueLen(uart_no);
+		tx_fifo_len = UART_TxFifoCount(uart_no);
 
 		// TODO If using output circbuf, check if empty
 
@@ -162,7 +162,7 @@ STATUS UART_WriteChar(UARTn uart_no, uint8 c, uint32 timeout_us)
 	uint32 t_s = system_get_time();
 
 	while ((system_get_time() - t_s) < timeout_us) {
-		uint8 fifo_cnt = UART_TxQueLen(uart_no);
+		uint8 fifo_cnt = UART_TxFifoCount(uart_no);
 
 		if (fifo_cnt < UART_TX_FULL_THRESH_VAL) {
 			WRITE_PERI_REG(UART_FIFO(uart_no), c);
diff --git a/user/uart_driver.h b/user/uart_driver.h
index 8a95c0d..c3d9bfd 100755
--- a/user/uart_driver.h
+++ b/user/uart_driver.h
@@ -172,8 +172,8 @@ extern UartDevice UartDev;
 //==============================================
 
 // FIFO used count
-#define UART_TxQueLen(uart_no) ((READ_PERI_REG(UART_STATUS((uart_no))) >> UART_TXFIFO_CNT_S) & UART_TXFIFO_CNT)
-#define UART_RxQueLen(uart_no) ((READ_PERI_REG(UART_STATUS((uart_no))) >> UART_RXFIFO_CNT_S) & UART_RXFIFO_CNT)
+#define UART_TxFifoCount(uart_no) ((READ_PERI_REG(UART_STATUS((uart_no))) >> UART_TXFIFO_CNT_S) & UART_TXFIFO_CNT)
+#define UART_RxFifoCount(uart_no) ((READ_PERI_REG(UART_STATUS((uart_no))) >> UART_RXFIFO_CNT_S) & UART_RXFIFO_CNT)
 
 STATUS UART_WriteCharCRLF(UARTn uart_no, uint8 c, uint32 timeout_us);
 STATUS UART_WriteChar(UARTn uart_no, uint8 c, uint32 timeout_us);
diff --git a/user/uart_handler.c b/user/uart_handler.c
index e969bae..59d8f67 100755
--- a/user/uart_handler.c
+++ b/user/uart_handler.c
@@ -19,11 +19,21 @@
 // messy irq/task based UART handling below
 
 static void uart0_rx_intr_handler(void *para);
+
 static void uart_recvTask(os_event_t *events);
+static void uart_processTask(os_event_t *events);
+
+// Those heavily affect the byte loss ratio
+#define PROCESS_CHUNK_LEN 1
+#define FIFO_FULL_THRES 4
 
 #define uart_recvTaskPrio        1
-#define uart_recvTaskQueueLen    15
+#define uart_recvTaskQueueLen    25
 static os_event_t uart_recvTaskQueue[uart_recvTaskQueueLen];
+//
+#define uart_processTaskPrio        0
+#define uart_processTaskQueueLen    25
+static os_event_t uart_processTaskQueue[uart_processTaskQueueLen];
 
 /** Clear the fifos */
 void ICACHE_FLASH_ATTR clear_rxtx(int uart_no)
@@ -62,14 +72,16 @@ void ICACHE_FLASH_ATTR UART_SetupAsyncReceiver(void)
 
 	// Start the Rx reading task
 	system_os_task(uart_recvTask, uart_recvTaskPrio, uart_recvTaskQueue, uart_recvTaskQueueLen);
+	system_os_task(uart_processTask, uart_processTaskPrio, uart_processTaskQueue, uart_processTaskQueueLen);
+
 	// set handler
 	ETS_UART_INTR_ATTACH((void *)uart0_rx_intr_handler, &(UartDev.rcv_buff)); // the buf will be used as an arg
 
 	// fifo threshold config (max: UART_RXFIFO_FULL_THRHD = 127)
-	uint32_t conf = ((90) << UART_RXFIFO_FULL_THRHD_S);
+	uint32_t conf = ((FIFO_FULL_THRES & UART_RXFIFO_FULL_THRHD) << UART_RXFIFO_FULL_THRHD_S);
 	conf |= ((0x10 & UART_TXFIFO_EMPTY_THRHD) << UART_TXFIFO_EMPTY_THRHD_S);
 	// timeout config
-	conf |= ((0x02 & UART_RX_TOUT_THRHD) << UART_RX_TOUT_THRHD_S); // timeout threshold
+	conf |= ((0x06 & UART_RX_TOUT_THRHD) << UART_RX_TOUT_THRHD_S); // timeout threshold
 	conf |= UART_RX_TOUT_EN; // enable timeout
 	WRITE_PERI_REG(UART_CONF1(UART0), conf);
 
@@ -82,32 +94,12 @@ void ICACHE_FLASH_ATTR UART_SetupAsyncReceiver(void)
 
 	// Enable IRQ in Extensa
 	ETS_UART_INTR_ENABLE();
-
-	// Start the periodic reading event
-	system_os_post(uart_recvTaskPrio, 5, 0);
 }
 
 
 // ---- async receive stuff ----
 
 
-void uart_rx_intr_disable(uint8 uart_no)
-{
-	CLEAR_PERI_REG_MASK(UART_INT_ENA(uart_no), UART_RXFIFO_FULL_INT_ENA | UART_RXFIFO_TOUT_INT_ENA);
-}
-
-
-void uart_rx_intr_enable(uint8 uart_no)
-{
-	SET_PERI_REG_MASK(UART_INT_ENA(uart_no), UART_RXFIFO_FULL_INT_ENA | UART_RXFIFO_TOUT_INT_ENA);
-}
-
-
-/**
- * @brief get number of bytes in UART tx fifo
- * @param UART number
- */
-#define UART_GetRxFifoCount(uart_no) ((READ_PERI_REG(UART_STATUS((uart_no))) >> UART_RXFIFO_CNT_S) & UART_RXFIFO_CNT)
 
 //
 //void ICACHE_FLASH_ATTR UART_PollRx(void)
@@ -120,29 +112,67 @@ void uart_rx_intr_enable(uint8 uart_no)
 //	}
 //}
 
-
-static void ICACHE_FLASH_ATTR uart_recvTask(os_event_t *events)
+/**
+ * This is the system task handling UART Rx bytes.
+ * The function must be re-entrant.
+ *
+ * @param events
+ */
+static void uart_recvTask(os_event_t *events)
 {
-	static char buf[200];
+//#define PROCESS_CHUNK_LEN 64
+//	static char buf[PROCESS_CHUNK_LEN];
 
 	if (events->sig == 0) {
-		UART_RxFifoDeq();
+		UART_RxFifoCollect();
 
 		// clear irq flags
 		WRITE_PERI_REG(UART_INT_CLR(UART0), UART_RXFIFO_FULL_INT_CLR | UART_RXFIFO_TOUT_INT_CLR);
 		// enable rx irq again
 		uart_rx_intr_enable(UART0);
+
+		// Trigger the Reading task
+		system_os_post(uart_processTaskPrio, 5, 0);
 	}
 	else if (events->sig == 1) {
 		// ???
 	}
-	else if (events->sig == 5) {
+//	else if (events->sig == 5) {
+//		// Send a few to the parser...
+//		int bytes = UART_ReadAsync(buf, PROCESS_CHUNK_LEN);
+//		for (uint8 idx = 0; idx < bytes; idx++) {
+//			UART_HandleRxByte(buf[idx]);
+//		}
+//
+//		// ask for another run
+//		if (UART_AsyncRxCount() > 0) {
+//			system_os_post(uart_recvTaskPrio, 5, 0);
+//		}
+//	}
+}
+
+
+/**
+ * This is the system task handling UART Rx bytes.
+ * The function must be re-entrant.
+ *
+ * @param events
+ */
+static void uart_processTask(os_event_t *events)
+{
+	static char buf[PROCESS_CHUNK_LEN];
+
+	if (events->sig == 5) {
 		// Send a few to the parser...
-		int bytes = UART_ReadAsync(buf, 200);
+		int bytes = UART_ReadAsync(buf, PROCESS_CHUNK_LEN);
 		for (uint8 idx = 0; idx < bytes; idx++) {
 			UART_HandleRxByte(buf[idx]);
 		}
-		system_os_post(uart_recvTaskPrio, 5, 0); // call me later
+
+		// ask for another run
+		if (UART_AsyncRxCount() > 0) {
+			system_os_post(uart_processTaskPrio, 5, 0);
+		}
 	}
 }
 
@@ -172,7 +202,7 @@ uart0_rx_intr_handler(void *para)
 		WRITE_PERI_REG(UART_INT_CLR(UART0), UART_RXFIFO_FULL_INT_CLR);
 
 		// run handler
-		system_os_post(uart_recvTaskPrio, 0, 0); /* -> notify the polling thread */
+		system_os_post(uart_recvTaskPrio, 0, 1); /* -> notify the polling thread */
 	}
 
 	if (status_reg & UART_RXFIFO_TOUT_INT_ST) {
@@ -187,10 +217,13 @@ uart0_rx_intr_handler(void *para)
 	if (status_reg & UART_TXFIFO_EMPTY_INT_ST) {
 		CLEAR_PERI_REG_MASK(UART_INT_ENA(UART0), UART_TXFIFO_EMPTY_INT_ENA);
 		UART_DispatchFromTxBuffer(UART0);
-//		WRITE_PERI_REG(UART_INT_CLR(UART0), UART_TXFIFO_EMPTY_INT_CLR); - is called by the dispatch func if more data is to be sent.
+//		WRITE_PERI_REG(UART_INT_CLR(UART0), UART_TXFIFO_EMPTY_INT_CLR); //- is called by the dispatch func if more data is to be sent.
 	}
 
 	if (status_reg & UART_RXFIFO_OVF_INT_ST) {
 		WRITE_PERI_REG(UART_INT_CLR(UART0), UART_RXFIFO_OVF_INT_CLR);
+
+		// overflow error
+		UART_WriteChar(UART1, '!', 100);
 	}
 }
diff --git a/user/uart_handler.h b/user/uart_handler.h
index 5200a91..80ef4e6 100755
--- a/user/uart_handler.h
+++ b/user/uart_handler.h
@@ -18,11 +18,25 @@ void UART_Init(void);
 /** Configure async Rx on UART0 */
 void UART_SetupAsyncReceiver(void);
 
-///** User must provide this func for handling received bytes */
+/** User must provide this func for handling received bytes */
 extern void UART_HandleRxByte(char c);
 
-void uart_rx_intr_disable(uint8 uart_no);
+static inline void uart_rx_intr_disable(uint8 uart_no)
+{
+	CLEAR_PERI_REG_MASK(UART_INT_ENA(uart_no), UART_RXFIFO_FULL_INT_ENA | UART_RXFIFO_TOUT_INT_ENA);
+}
 
-void uart_rx_intr_enable(uint8 uart_no);
+static inline void uart_rx_intr_enable(uint8 uart_no)
+{
+	SET_PERI_REG_MASK(UART_INT_ENA(uart_no), UART_RXFIFO_FULL_INT_ENA | UART_RXFIFO_TOUT_INT_ENA);
+}
+
+/**
+ * @brief get number of bytes in UART tx fifo
+ * @param UART number
+ */
+static inline u8 UART_GetRxFifoCount(u8 uart_no) {
+	return (u8) ((READ_PERI_REG(UART_STATUS((uart_no))) >> UART_RXFIFO_CNT_S) & UART_RXFIFO_CNT);
+}
 
 #endif