[hylafax-devel] My RTN patch and many other changes/enhancements, third edition :-)

[Dr. Harald Pollack <Harald.Pollack@Datanews.at>]

On Fri, 14 Apr 2000 08:52:34 +0200, Bernd Proissl wrote:


Hello Bernd!

>for me the following parameters look reasonable:
>
>
>parameter 1: number of times the page which got RTN will be retransmitted
>before lowering signalling rate  (default=0)

Keep in mind, that RTN implies a RETRAIN, and this is the only point of fax protocol, where a 
bitrate election should be made (on base if remote either response with FTT or CFR). In case of 
FTT response, first an 'intra modulation' fallback should be done and after this an 'inter 
modulation' fallback.

Intra modulation: WITHIN a distinct modulation scheme (V.17, V.29)
Inter modulation: from one scheme to another (V.17 to V.27t or V.29 to V.27t)

So it is obvious, that in bad line situations, V.17 should only be selected carefully, because it 
may be impossible to fallback till V.27t (14400 - 12000 - 9600 - 7200 ... and normally timeout 
hits) while starting from V.29 will always resulting in V.27t (9600 - 7200 - 4800 - 2400).

>parameter 2: lowest possible signalling rate which is allowd after RTN
>             hylafax should give up if it gets an RTN while transmitting at this rate
>             (default=4800, default=2400 for far distance calls)

In case of retrain (which is mandatory after RTN) sender have to start with DIS/DCS negotiation 
and therefore with highest bitrate - both - are able to use. Fallback should be done till 2400 
bps down and than training at 2400 should be repeated till timeout hits (or maybe after one or 
two repetitions).

>parameter 3: How many calls will be made additionally if the current session
>was aborted due to RTN. (default=0, default=1 for far distance calls)

In practice none :-) but after Dmitrys fixes it is more than possible, that such RTN showers will 
be history at all :-)

>can this parameters be made dependend on the number dialed and/or on the
>access code dialed?

SHould be done better on (recognized) remote fax equipment ...

>this would allow differnet handling for certain problematic machines or for
>far distance calls (which are theoretically more likely to line noise)

Line NOISE is rarely the reason for RTN. Line noise influences signal/noise ratio, which will 
influence training at higher bitrates. So line noise is normaly detected by trainings sequence.


Mit herzlichen Gruessen / Yours sincerely

Dr. Harald Pollack

Harald.Pollack@DATAnews.at



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