flexfax: Retrain Negative errors

[Mike Brodbelt <m.brodbelt@acu.ac.uk>]

I'm having a number of problems with Hylafax jobs failing with retrain
negative errors, like this one below.

Feb 24 14:53:24.19: [23658]: <-- [7:AT+FDT\r]
Feb 24 14:53:24.61: [23658]: --> [7:CONNECT]
Feb 24 14:53:24.61: [23658]: SEND begin page
Feb 24 14:53:42.74: [23658]: SENT 21015 bytes of data
Feb 24 14:53:42.74: [23658]: SEND 1D RTC
Feb 24 14:53:42.74: [23658]: SEND end page
Feb 24 14:53:42.74: [23658]: SEND send EOP (no more pages or documents)
Feb 24 14:53:59.85: [23658]: --> [5:ERROR]
Feb 24 14:53:59.85: [23658]: SEND recv RTN (retrain negative)
Feb 24 14:53:59.85: [23658]: <-- [7:AT+FDT\r]
Feb 24 14:54:01.44: [23658]: --> [7:+FHS:00]
Feb 24 14:54:01.44: [23658]: REMOTE HANGUP: Normal and proper end of
connection (code 0)
Feb 24 14:54:01.44: [23658]: --> [2:OK]
Feb 24 14:57:01.44: [23658]: MODEM <Timeout>
Feb 24 14:57:01.44: [23658]: <-- [5:ATH0\r]
Feb 24 14:57:01.71: [23658]: --> [2:OK]
Feb 24 14:57:01.71: [23658]: SESSION END

Many of my users are trying to send faxes to remote locations such as
India, or Pakistan, as well as more technically modern countries, and so
I strongly suspect that the quality of remote phone lines is responsible
for many of these problems. Is there any way to make Hylafax more
tolerant of these errors, or is this just something I have to write off
as due to bad lines, and that I can't do anything about?

Mike.