Hylafax Mailing List Archives
|
[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
[hylafax-users] HylaFAX-4.1beta3
Hi HylaFAX devolpers,
thanks a lot for your new release (4.1beta3)!
There are some problems in compiling HylaFAX-4.1beta3 on Linux
(glibc-2.1.2, gcc-2.95.2) with utmpx support:
diff -ur hylafax-4.1beta3.ORIG/faxd/GettySysV.c++
hylafax-4.1beta3/faxd/GettySysV.c++
--- hylafax-4.1beta3.ORIG/faxd/GettySysV.c++ Mon Aug 23 13:19:33
1999
+++ hylafax-4.1beta3/faxd/GettySysV.c++ Thu Feb 22 12:21:58 2001
@@ -30,7 +30,6 @@
#include <termios.h>
#include <sys/ioctl.h>
extern "C" {
-#include <utmp.h>
#if HAS_UTMPX
#if !HAS_EXIT_STATUS
/* workaround SVR4.0.3 header braindamage */
@@ -39,7 +38,9 @@
short e_exit; /* Process exit status */
};
#endif /* !HAS_EXIT_STATUS */
+#define __USE_GNU
#include <utmpx.h>
+#undef __USE_GNU
#define utmp utmpx
#undef ut_time
@@ -51,6 +52,8 @@
#define pututline pututxline
#define setutent setutxent
#define endutent endutxent
+#else
+#include <utmp.h>
#endif /* HAS_UTMPX */
}
With utmpx support enabled, it would be better to include 'utmpx.h'
alone, not 'utmp.h' too. Otherwise you'll get a problem with
'struct exit_status'
The '#define __USE_GNU' is required to switch on the declaration of
'updwtmpx' in 'utmpx.h'.
Linux with glibc-2.1.2 doesn't have a member 'ut_time' in 'struct
utmpx',
instead there is a member 'ut_tv' with type 'struct timeval':
@@ -169,7 +172,9 @@
ut.ut_exit.e_exit = 0;
ut.ut_exit.e_termination = 0;
#endif
- ut.ut_time = Sys::now();
+// ut.ut_time = Sys::now();
+ ut.ut_tv.tv_sec = Sys::now();
+ ut.ut_tv.tv_usec = 0;
// mark utmp entry as a login
strncpy(ut.ut_user, "LOGIN", sizeof (ut.ut_user));
/*
@@ -212,7 +217,9 @@
ut->ut_exit.e_exit = (exitStatus >> 8) & 0xff; //
XXX
ut->ut_exit.e_termination = exitStatus & 0xff; //
XXX
#endif
- ut->ut_time = Sys::now();
+// ut->ut_time = Sys::now();
+ ut->ut_tv.tv_sec = Sys::now();
+ ut->ut_tv.tv_usec = 0;
pututline(ut);
writeWtmp(ut);
break;
There is a very small problem in 'config.site' but it's there for
several years:
#LOCKS="ansi"
should be
#LOCKS="ascii"
I have found that it would be appropriate to raise the value of
IXO_XMITTIMEOUT to 45 seconds as in the UCP protocol specification:
diff -ur hylafax-4.1beta3.ORIG/faxd/ixo.h hylafax-4.1beta3/faxd/ixo.h
--- hylafax-4.1beta3.ORIG/faxd/ixo.h Mon Oct 12 22:47:49 1998
+++ hylafax-4.1beta3/faxd/ixo.h Thu Feb 22 10:09:14 2001
@@ -41,7 +41,7 @@
#define IXO_LOGINTIMEOUT 15 // timeout on login
response
#define IXO_GATIMEOUT 30 // timeout waiting
for go-ahead msg
#define IXO_XMITRETRIES 3 // 3 attempts to send
message block
-#define IXO_XMITTIMEOUT 15 // timeout for
message block xmit resp
+#define IXO_XMITTIMEOUT 45 // timeout for
message block xmit resp
#define IXO_ACKTIMEOUT 30 // timeout waiting
for transaction ack
The charcter ETX should be discarded too in receiving SMS responses:
diff -ur hylafax-4.1beta3.ORIG/faxd/pageSendApp.c++
hylafax-4.1beta3/faxd/pageSendApp.c++
--- hylafax-4.1beta3.ORIG/faxd/pageSendApp.c++ Mon Aug 23 13:19:35
1999
+++ hylafax-4.1beta3/faxd/pageSendApp.c++ Thu Feb 22 10:08:51
2001
@@ -416,7 +416,7 @@
int c = getModemChar(0);
if (c == EOF)
break;
- if (c == '\r') {
+ if (c == '\r' || c == '\003') {
if (buf.getLength() > 0) // discard leading
\r's
break;
} else if (c != '\n') // discard all \n's
Thanks again and many greetings from the Cologne Carneval
Ado
--
------------------------------------------------------------------------
Heinz-Ado Arnolds Ado.Arnolds@dhm-systems.de
DHM GmbH & Co. KG +49 2234 18400 (voice)
Max-Planck-Strasse 2, 50858 Koeln, Germany +49 2234 184040 (fax)
____________________ HylaFAX(tm) Users Mailing List _______________________
To unsub: mail -s unsubscribe hylafax-users-request@hylafax.org < /dev/null