Re: flexfax: Modified Huffman protocol

[David Woolley <david@djwhome.demon.co.uk>]

> 
> Nope, the error is from getty, which is run to initiate the fax connection

You should be running faxgetty, not getty.

> > To get any further, we need detailed Hylafax session logs.
> > 
> > What fax class are you using?
> 
> Fax class is 2.0. here's one log:

You are aware that Sportsters have problems in anything but Class 1?

> Jan 10 21:15:23.88: [   91]: SESSION BEGIN 00000005 ********* (fax number
> edited out)
> Jan 10 21:15:23.89: [   91]: <-- [4:ATA\r]
> Jan 10 21:16:17.30: [   91]: --> [10:NO CARRIER]
> Jan 10 21:16:17.31: [   91]: ANSWER: Ring detected without successful
> handshake
> Jan 10 21:16:17.31: [   91]: SESSION END

Hylafax wasn't involved there, except in its initial programming of the modem.


> Jan 10 21:17:22.15: [   91]: ANSWER: DATA CONNECTION
> Jan 10 21:17:22.15: [   91]: GETTY: START "/sbin/getty -h cua1 dx_38400",
> pid 4$

faxgetty has handed off to the normal voice getty because the modem thinks
the incoming call is a data one.

> Jan 10 21:17:22.28: [   91]: GETTY: exit status 0400

The data getty has failed rather fast, I think because of the inappropriate
use of -h, which I believe forces a modem hangup when the getty is started,
which is not a sensible thing to do when you have already answered an
incoming call.  If this is correct, data calls could never have worked,
so I would wan't to know why you have configured a data getty, given the
security risks.

I probably failed to find this in the source because I was looking for
'getty exit', not 'GETTY: exit'.
                        ^

Are you sure that the sending modem is generating fax calling tone?