Re: [hylafax-users] Remote hangs up after ANSWER

[Lee Howard <faxguy@deanox.com>]

At 10:05 AM 8/31/01 +0200, Sjon Wijnolst wrote:
>> >Hello,
>> >
>> >I've installed HylaFAX 4.1 on a SuSE 7.1 linux box and
>> everythings worked
>> >fine during tests. So we decided to make this setup our new
>> client-server
>> >fax system and installed everything yesterday. After some
>> minor problems
>> >with clients submitting faxes everything looks OK now but one of our
>> >customers called that they can't send a fax. Their fax
>> machine says there's
>> >no answer on the other line - according to the log-file of
>> the session I say
>> >that their machine just hangs up!?
>> >
>> >Session log:
>> >Aug 30 14:16:11.32: [  334]: SESSION BEGIN 00000033 31786197475
>> >Aug 30 14:16:11.32: [  334]: <-- [4:ATA\r]
>> >Aug 30 14:16:19.95: [  334]: --> [5:+FCON]
>> >Aug 30 14:16:19.95: [  334]: ANSWER: FAX CONNECTION
>> >Aug 30 14:16:19.95: [  334]: RECV FAX: begin
>> >Aug 30 14:16:21.01: [  334]: --> [8:+FHNG: 0]
>> >Aug 30 14:16:21.01: [  334]: REMOTE HANGUP: Normal and proper end of
>> >connection (code 0)
>> >Aug 30 14:16:22.83: [  334]: --> [2:OK]
>> >Aug 30 14:16:22.83: [  334]: RECV FAX: Normal and proper end
>> of connection
>> >Aug 30 14:16:22.83: [  334]: RECV FAX: end
>> >Aug 30 14:16:22.83: [  334]: SESSION END
>>
>> Tried using Class 1 ?
>>
>> Lee.
>>
>
>No, that could be an option. But, what does it mean for sending and
>receiving faxes when I use Class 1 instead of Class 2? Are there any
>disadvantages?

Simplified... Class 1 is fax protocol that is mostly run by the software.
Class 2 is fax protocol that is mostly run by the hardware (firmware).
That's the basics of the differences.

In practice... Class 1 tends to be easier for us to work with because we
can see, edit, and review most of the code which runs the protocol -
because it's embedded into HylaFAX.  In the past, people have been worried
about using Class 1 because of CPU usage (especially when running several
modems in one system) affecting the critical timing that fax protocol
requires.

Frankly, unless they were running 286 or 8088/8086 systems (and even
then...) or the system was weak and normally using +80% CPU, I don't think
this is a valid concern.  So, in the past, when people have felt like Class
1 was less reliable than Class 2 because of timing problems - well, I think
that their fingers were pointing at the nature of the Class 1 protocol
rather than where it should have been pointing: their implementation of the
Class 1 protocol.

So, these days Class 1 in HylaFAX tends to be much more reliable than Class
2/2.0.  Why?  Because we've worked out most of the bugs that so plagued
such doomsayers before.  If you find a flaw with Class 2/2.0 protocol,
chances are good that HylaFAX can't fix it - because the code to fix is
embedded inside the chips on your modem.  So in order to make Class 2 work
on all modems perfectly, HylaFAX has to make its Class 2 code full of
workarounds and fixes for faulty firmware - not a fun job, and for what
when we can simply just switch to Class 1?  If we find a Class 1 bug - in
most cases we just have to fix the code that's causing the problem.  You
can see that the latter option is much more friendly.

But - there have been instances where I've found that certain modems cannot
CONNECT with certain other modems or fax machines after ATA or ATDT when
using one fax Class or another.  So, that's why in your case I recommended
using Class 1 to test.  If it still doesn't connect, then we try something
else - like disabling adaptive answer.

Lee.



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