Re: [hylafax-users] [SOLVED] RE: factors in bad scan lines

[Dmitry Bely <dbely@mail.ru>]

Lee Howard <faxguy@deanox.com> writes:

> >The correct settings are:
> >
> >Class1:
> >ModemSendFillOrder      LSB2MSB
> >ModemRecvFillOrder      LSB2MSB
> >ModemFrameFillOrder     LSB2MSB
> >
> >Class2:
> >ModemSendFillOrder      LSB2MSB
> >ModemRecvFillOrder      MSB2LSB (infamous Rockwell bug)
> >ModemFrameFillOrder     Not applicable
> >
> >Initially Rockwell was introduced this bug in its modem's data pump (Class2
> >also assumed LSB2MSB), which most software and hardware makers had to
> >follow. I know the only exception -- non-Rockwell Multitech modems, that
> >implements the original Class2 specs and use LSB2MSB for ModemRecvFillOrder.
> >
> >Class2.0:
> >ModemSendFillOrder      LSB2MSB
> >ModemRecvFillOrder      LSB2MSB
> >ModemFrameFillOrder     Not applicable
> 
> Thank you!
> 
> So, should a "default" config setting for Class 2 be 'ModemRecvFillOrder:
> MSB2LSB' ?
> This "default" wouldn't work for Multitech, but for everything else?

Yes. I've changed the sources to autodetect this parameter the following
way (will be commited a bit later):

faxd/Class2.c++
[---cut---]
    /*
     * Class2/2.0 defaults for send & receive bit order
     */
    sendFillOrder = (conf.sendFillOrder != 0)?
        conf.sendFillOrder :
        FILLORDER_LSB2MSB;
    recvFillOrder = (conf.recvFillOrder != 0)?
        conf.recvFillOrder : 
        (serviceType == SERVICE_CLASS20 || modemMfr.find(0, "MULTI-TECH") == 0)?
            /*
	     * Multitech report the manufacturer as
	     * "Multi-Tech Sys" or "Multi-Tech Systems"
	     *
	     * it seems to be the only Class2 modem which
	     * does not clone the infamous Rockwell bug
	     * (which has become "standard" now)
	     */
            FILLORDER_LSB2MSB :
            FILLORDER_MSB2LSB;

    rtcRev = TIFFGetBitRevTable(sendFillOrder == FILLORDER_LSB2MSB);

    setupClass2Parameters();			// send parameters to the modem
    return (true);
}
[---cut---]

Hope to hear from you soon,
Dmitry




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