[hylafax-devel] Re: My RH6.2 RPM hack & Q: Failure to train at 2400 bps or +FMINSP value

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

On Sun, 21 May 2000 10:05:36 -0700, joel.peshkin@conexant.com wrote:

Hello Joel!

>The place where class 1 and class 2/2.0 really diverge is when you have
>many (like 24 or more) fax modems being driven by the same machine or where
>the modems are seperated from the controlling program by a network.

:-) In this case, class 2/2.0 is much more timecritical than class 1 :-)

>Class 2/2.0 both handle the real-time processing required by the protocol

T.30 protocol is HALF DUPLEX, so there is not so much 'real time' neccessary ...

>in the modem so that any stalls on the DTE-DCE interface only result in flow
>control.

On the receiving side, I agree, but on the sending side (from the point of view of fax prog) 
there is not so much flow controll, because each command is followed from a response ...

>Class 1 requires the real-time processing to be done by the host
>system,

???? I repeat: there is no real-time ... there are only time windows (and these are as large as 
15 - 20 seconds).

>so any stalls result in deterioration of connect reliability and/or
>fax quality.

There is no difference while transmitt phase C data from/to modem regardless if class 1 or class 
2/2.0 On a certain hardware, it makes no difference, if phase C data is transmitted in class 1 or 
class 2. So fax quality is INDEPEND of class protocol. Well and reliability is a question, how 
well t.30 protocol is done by faxsoftware also regardless if class 1 or class 2/2.0

Only problem is the unknow state of modem in class 2/2.0, or do YOU know exactly, what modem is 
doing after you have sent the AT+FDT command? In class 1, you know it very well, in class 2/2.0 
you can only guess...

In case of any error, in class 2/2.0 you can only wait for +FHNG/+FHS and you can not do any 
error recover. In class 1, you can do any error recover sequence, T.30 (and you phantasy) allows 
:-)


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