Re: [hylafax-users] Problem going from 4.1.5->4.1.7 for topic?

[Lee Howard <faxguy@howardsilvan.com>]

On 2003.07.31 07:41 Giulio Orsero wrote:
> In the past days I upgraded various systems (rockwell, lucent) to
> 4.1.6/4.1.7 with no problems.
> 
> Today I upgraded a topic chipset system (send-only system) [system A],
> and
> did some tests sending to HylaFAX-4.1.7(rockwell) [system B].

....[error after upgrade]
> Jul 31 16:10:48.21: [15262]: SEND training at v.17 14400 bit/s
> Jul 31 16:10:48.21: [15262]: <-- [9:AT+FTH=3\r]
> Jul 31 16:10:48.26: [15262]: --> [7:CONNECT]
> Jul 31 16:10:48.26: [15262]: <-- data [23]
> Jul 31 16:10:48.26: [15262]: <-- data [2]
> Jul 31 16:10:48.28: [15262]: --> [1:]
> Jul 31 16:10:48.28: [15262]: Error sending T.30 prologue frames

....[same part before upgrade]
> Jul 31 16:13:58.25: [15412]: SEND training at v.17 14400 bit/s
> Jul 31 16:13:58.25: [15412]: <-- [9:AT+FTH=3\r]
> Jul 31 16:13:58.30: [15412]: --> [7:CONNECT]
> Jul 31 16:13:58.30: [15412]: <-- data [23]
> Jul 31 16:13:58.30: [15412]: <-- data [2]
> Jul 31 16:13:58.32: [15412]: --> [1:]
> Jul 31 16:13:58.33: [15412]: --> [7:CONNECT]
> Jul 31 16:13:58.33: [15412]: <-- data [6]
> Jul 31 16:13:58.33: [15412]: <-- data [2]
> Jul 31 16:14:00.50: [15412]: --> [2:OK]

The problem was with:

http://bugs.hylafax.org/bugzilla/attachment.cgi?id=414&action=view

this will not happen with 4.1.6/4.1.7 if you back out the last patch 
there, but you could have faxgetty hang with USRs.
Please try the attached patch against 4.1.6/4.1.7 for a better 
resolution to Bug 414.

Thanks.

Lee.


diff -Nru hylafax.orig/faxd/ClassModem.c++ hylafax/faxd/ClassModem.c++
--- hylafax.orig/faxd/ClassModem.c++	Mon Jul 14 09:38:25 2003
+++ hylafax/faxd/ClassModem.c++	Thu Jul 31 09:29:02 2003
@@ -61,7 +61,7 @@
     "",				// 7
     "\"Voice\"",		// SERVICE_VOICE
 };
-const char* ClassModem::ATresponses[14] = {
+const char* ClassModem::ATresponses[15] = {
     "Nothing",			// AT_NOTHING
     "OK",			// AT_OK
     "Connection established",	// AT_CONNECT
@@ -74,7 +74,8 @@
     "Command error",		// AT_ERROR
     "<Empty line>",		// AT_EMPTYLINE
     "<Timeout>",		// AT_TIMEOUT
-    "<xonxoff>",		// AT_XONXOFF
+    "<dle+etx>",		// AT_DLEETX
+    "<xon>",			// AT_XON
     "<Unknown response>"	// AT_OTHER
 };
 const char* ClassModem::callTypes[5] = {
@@ -770,8 +771,12 @@
 		lastResponse = AT_RING;
 	    break;
 	case '\020':
-	    if (streq(buf, "\020\003"))		// DC1/DC3 (XON/XOFF)
-		lastResponse = AT_XONXOFF;
+	    if (streq(buf, "\020\003"))		// DLE/ETX
+		lastResponse = AT_DLEETX;
+	    break;
+	case '\021':
+	    if (streq(buf, "\021"))		// DC1 (XON)
+		lastResponse = AT_XON;
 	    break;
 	}
     }
diff -Nru hylafax.orig/faxd/ClassModem.h hylafax/faxd/ClassModem.h
--- hylafax.orig/faxd/ClassModem.h	Mon Jul 14 09:38:25 2003
+++ hylafax/faxd/ClassModem.h	Thu Jul 31 09:28:55 2003
@@ -166,8 +166,9 @@
 	AT_ERROR	= 9,	// "ERROR" response
 	AT_EMPTYLINE	= 10,	// empty line (0 characters received)
 	AT_TIMEOUT	= 11,	// timeout waiting for response
-	AT_XONXOFF	= 12,	// xon/xoff characters
-	AT_OTHER	= 13	// unknown response (not one of above)
+	AT_DLEETX	= 12,	// dle/etx characters
+	AT_XON		= 13,	// xon character
+	AT_OTHER	= 14	// unknown response (not one of above)
     };
 private:
     ModemServer& server;	// server for getting to device
@@ -191,7 +192,7 @@
 
     static const char* serviceNames[9];	 // class 2 services
     static const char* callStatus[10];	 // printable call status
-    static const char* ATresponses[14];
+    static const char* ATresponses[15];
 
     ClassModem(ModemServer&, const ModemConfig&);