Re: [hylafax-users] Repost - HylaFAX not honoring T.31?

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

On 2004.11.26 18:31 Steve Underwood wrote:

> If my softmodem replies "0,1.0" faxgetty sticks as the point where it 
> is waiting for the reply to "+FCLASS=?", until a timeout occurs.
> If my softmodem replies "0,1" faxgetty is happy, and I can send and 
> receive faxes successfully with my softmodem.

Attached is the patch to fix this problem.  Thanks for reporting it.  
The patch has been committed to CVS-HEAD.

> Although I can get HylaFAX to work, "0,1" does not comply with the 
> spec. and I am not happy about that. I realise the spec is much 
> abused. Some modems document that "1" will give you a modem like the 
> T.31 spec. and "1.0" will give you a modem like the V.34 supplement 
> to T.31. They claim this complies with T.31, but T.31 does not concur 
> with that view.

It's a long story.

The "old" spec (EIA/TIA-578) indicated a response of "1" to AT+FCLASS=? 
in order to indicate support for that recommendation.

When T.31 was developed there were three new AT commands added to it 
(the only significant one, and the only one I can remember at the 
moment is +FAR).  And so to distinguish the new recommendation from the 
old one the specified response to AT+FCLASS=? was indicated as "1.0" 
instead.  Some time later Annex C was added to T.31, which enables 
V.34-Fax.  There may have been some subtle differences between 
EIA/TIA-578 and T.31, but I do not know of any except for various 
implementation clarifications that were made and other types of 
suggestions (mostly involved with timing and the usage of +FRS/+FTS).

So some manufacturers, notably Conexant, support "Class 1.0" and that 
means that the modem supports those three new commands that were new to 
T.31, but those chipsets, as of yet, do not support V.34-Fax.  Other 
manufacturers, notably Agere, support "Class 1.0" and that means that 
the modem supports V.34-Fax, but those chipsets, as of yet, do not 
support the three new AT commands.  So, again, we find a bit more 
disparity in the fax world.

In order to make your softmodem work reliably with other software 
packages I recommend that your response to the AT+FCLASS=? be 
"0,1,1.0".  This would be interpreted to indicate support for data (0) 
EIA/TIA-578 (1) and T.31 (1.0).  Of course, with the attached patch 
HylaFAX will have no problem with the response as you desire.

Thanks,  (and thanks for the work that you're doing now)

Lee.

Attachment: hylafax-servicetype.patch
Description: Binary data