Re: [hylafax-users] [hylafax-devel] Re: any body can help ?

[Dmitry Bely <dbely@mail.ru>]

Robert Colquhoun <rjc@trump.net.au> writes:

> > > I'm using Courier I modem and using ISDN line.
> 
> Have you tried the latest firmware on this?.....(also does it use any
> kind of special software driver?)
> 
> 
> > >     ---- Transcript of session follows ----
> > >
> > > Aug 27 13:10:14.58: [14207]: SESSION BEGIN 00000007
> > > 62226031855
> > > Aug 27 13:10:14.58: [14207]: SEND FAX: JOB 1 DEST 6031855
> > > COMMID 00000007
> > > Aug 27 13:10:14.58: [14207]: DELAY 2600 ms
> 
> Drops DTR...
> 
> > > Aug 27 13:10:17.17: [14207]: <-- [4:ATZ\r]
> > > Aug 27 13:10:22.17: [14207]: MODEM <Timeout>
> 
> and the ATZ command disappears. :-(
> ....
> > > Aug 27 13:10:23.54: [14207]: <-- [28:AT+FLI="JTSoft
> > > Db.App.Dev."\r]
> > > Aug 27 13:10:23.69: [14207]: --> [27:AT+FLI="JTSoft Db.App.Dev."]
> > > Aug 27 13:10:23.69: [14207]: --> [2:OK]
> > > Aug 27 13:10:23.70: [14207]: DIAL 6031855
> > > Aug 27 13:10:23.70: [14207]: <-- [12:ATDT6031855\r]
> > > Aug 27 13:10:33.29: [14207]: --> [11:ATDT6031855]
> > > Aug 27 13:10:33.29: [14207]: SEND FAILED: Unknown problem
> > > (check modem power)
> 
> Another problem here as well.

No -- that's just because Hylafax reset commands were not issued, switching 
the echo off. Hylafax expects only CONNECT, NO CARRIER, ... messages, not
ATD echo.

> > > Aug 27 13:10:33.29: [14207]: <-- [5:ATH0\r]
> > > Aug 27 13:10:33.29: [14207]: --> [4:+FCO]
> > > Aug 27 13:10:35.04: [14207]: --> [7:+FHS:02]
> > > Aug 27 13:10:35.04: [14207]: REMOTE HANGUP: Call aborted,
> > > from +FK or <CAN> (code 2)
> > > Aug 27 13:10:35.04: [14207]: --> [2:OK]
> > > Aug 27 13:10:35.04: [14207]: SESSION END
> >
> >Something wrong happened with the modem -- there was no "OK" in response
> >to ATZ! Of course, the best solution of the above problem is AT&F&W, but
> 
> As well as ATZ not working the modem refused to dial as well, possibly
> if we ignore the first error(the ATZ) then it will only reach ATDT and
> then stop again.

As I mentioned before, ATDT failed (more precisely, Hylafax thought that
it failed) just because no ModemEchoOffCmd (ATE0) was issued.

> Perhaps connect to the modem using minicom/cu/kermit and manually send
> the AT&F&W to see if it fixes things up....
> 
> 
> 
> >... maybe to change the following in ClassModem.c++
> >
> >     /*
> >      * Perform a soft reset as well to ensure the modem
> >      * is in a stable state before sending the additional
> >      * reset commands.
> >      */
> >     return atCmd(conf.softResetCmd, AT_OK, conf.resetDelay)
> >             && atCmd(resetCmds, AT_OK, ms);
> >
> >
> >to
> >
> >     (void) atCmd(conf.softResetCmd, AT_OK, conf.resetDelay);
> >     return atCmd(resetCmds, AT_OK, ms);
> >
> >
> >as I suggested initially? This way we could still recover in the situation
> >like that. What do you think, Robert?
> 
> Maybe:
> 
> if (!atCmd(conf.softResetCmd, AT_OK, conf.resetDelay)) {
>          //report error here
>          if (!atCmd("AT&F&W", AT_OK, conf.resetDelay)) {
>                  //there's really no coming back from here - crash and burn.
>          }
> }
> return atCmd(resetCmds, AT_OK, ms);

I don't like it :-))) The problem is that
a) IMHO silent rewriting of NVRAM by any software is not acceptable -- the
user will be very suprised :-)
b) Some modems restore proper factory defaults via AT&F1 or AT&F2, not AT&F.
c) atCmd(conf.softResetCmd, AT_OK, conf.resetDelay) is needed just to reset 
the modem in addition to DTR drop. Even if ATZ fails, other Hylafax reset
commands still may succeed.

Hope to hear from you soon,
Dmitry




____________________ HylaFAX(tm) Users Mailing List _______________________
 To unsub: mail -s unsubscribe hylafax-users-request@hylafax.org < /dev/null