Re: flexfax: Can only send one page to HP office fax/printer

[Robert Colquhoun <rjc@trump.net.au>]

Hi Damien,

At 12:06 AM 11/6/98 +1100, Damian Ivereigh wrote:
>I have a different modem (Multi-Tech) and I would be interested to see
>if this is a modem or a software problem. It seems so repeatable that it
>doubt it is just line noise. I know the standard answer has just been to
>revert to Class 1 operation, presumuably because the modem is not doing
>the right thing in Class 2 and Hylafax would do a better job. However
>Class 1 requires the s/w to do much more handholding of the protocol
>than class 2.
>
>I should add that I am not an expert with Hylafax, but I am willing to
>dive into the code if required.

I'm not an expert either, but i've got some documents here to help me :-)

>
>Actually I am not sure what I said earlier about RTN was right. Maybe
>one of the protocol experts can jump in and explain. Looking at the
>log:-
>
>> > > > Nov 04 03:25:11.95: [14694]: SENT 13269 bytes of data
>> > > > Nov 04 03:25:11.95: [14694]: SEND 1D RTC
>> > > > Nov 04 03:25:11.95: [14694]: <-- data [9]
>> > > > Nov 04 03:25:11.95: [14694]: SEND end page
>> > > > Nov 04 03:25:11.95: [14694]: SEND send MPS (more pages, same document)
>> > > > Nov 04 03:25:11.95: [14694]: <-- data [2]
>> > > > Nov 04 03:25:28.27: [14694]: --> [5:ERROR]
>
>Presumuably this is the remote saying (through the modem) it got an
>error.

ok from above:
1)Hylafax sent the page to the modem(13269 bytes) and hopefully the modem
sent it to the receiving fax.
2)Hylafax sent the RTC - which is a special sequence of bytes that tell the
fax at the other end the page is finished.  If the page data transfer is
out of control the receiving end can wait for this series of bytes to
appear on line to regain control of the session.
3)Hylafax sent the end of page which is a <dle> char + <mps> char to tell
the modem that the page was finished and we want to send more pages, it can
also send <eop> for no more pages to come.  The modem sends this to the
other fax machine.
4)The receiving modem then replies Message ConFirmation if everything is ok
to go ahead or if things went bad a number of error codes like ReTrain
Positve, Retrain Negative or a bunch of other errors.
MCF - stay in phase C do next page
RTP - go back into phase B renegotiate session then send next page(ie got
prev page)
RTN - go back into phase B renegotiate session then send same page
again.(ie did *not* get prev page)

>
>> > > > Nov 04 03:25:28.27: [14694]: SEND recv RTN (retrain negative)
>
>Is this Hylafax announcing its intention to *send* an RTN (drop the
>speed) to the modem?
>

The receiving modem has sent this because it hasn't got the previous page
properly, which could be for alot of different reasons like line noise,
trying to send a badly encoded document, some kind of buffer overrun
somewhere on the way.

>> > > > Nov 04 03:25:28.27: [14694]: <-- [7:AT+FDT\r]
>
>Is this Hylafax actually send the RTN is AT commands?

Hylafax is telling the modem to go back into phase b to renegotiate the
session.  In this situation Hylafax's class1 driver will automatically drop
the maximum speed of the connection, which will help if the cause of the
problem is line noise.  The class 2 driver tries again at the same speed.
Dropping speed will not help if the doc is badly encoded(except for some
receiving fax machines when the speed is lowered to 2400 will not reply
with a RTN and accept the page no matter what).


>
>> > > > Nov 04 03:25:34.28: [14694]: --> [20:+FCS:0,3,0,2,0,0,0,5]

This is the result of the phase b session - the new session parameters for
sending the next page.

- Robert