Mike,
Yea, rewording the definitions makes sense then.
Paraphrasing --
fax-modem-carrier-lost: Lost connection to the receiver during send. (Receiver detected loss of carrier and could not acknowledge reception of carrier).
fax-modem-training-failure: The sender and receiver were unable to successfully negotiate a data rate. (Receiver was unable to agree with the data rate suggested by the sender (or) Receiver unable to receive the facsimile per the suggested data rate)
Or something better ....
Daniel.
-----Original Message-----
From: Michael Sweet [mailto:msweet at apple.com]
Sent: Thursday, June 27, 2013 6:22 PM
To: Manchala, Daniel
Cc: 'fx OHTAKE SHIN'; 'ipp at pwg.org'
Subject: Re: [IPP] Re: Fuji Xerox has reviewed the IPP FaxOut specification and has comments
Daniel,
On 2013-06-27, at 8:26 PM, "Manchala, Daniel" <Daniel.Manchala at xerox.com> wrote:
> I think renaming both 'fax-modem-carrier-lost' and 'fax-modem-training-failure' to something like 'fax-modem-confirmation-not-received' or 'fax-out-modem-confirmation-not-received' would help. The reasoning is as follows:
These names have already been approved (5107.3) and are in use (CUPS, which provided prototyping specifically for the preliminary fax support via efax).
Like I've said on many occasions, we can change the definitions all we like, but unless we want the mess of supporting deprecated old names plus semantically identical new names I don't see the point in continuing discussion of changing the names.
As for fax-in-* and fax-out-*, there are enough common states between receiver and sender that I would not want to double things up.
> As part of training (speed or data rate such as 9600bps) and at the end of it, a TCF (Training Check Flag) is sent from the fax-out (sending) device to a fax-in (receiving) device. If there is a training failure, an FTT (Failure to Train) is sent back from the fax-in receiving device to the fax-out sending device. If there is no training failure, a CFR (Confirmation to Receive) is sent from the fax-in receiving device to the fax-out sending device. After sufficient number of training failures (usually 3), a DCN (Disconnect) signal is sent from fax-out device to fax-in device and the transmission stops.
>> Likewise, if carrier is lost, there is no response received in terms of MCF (Message Confirmation) from fax-in device at the fax-out device.
>> In either of the two cases, a failure to detect/receive CFR or MCF from the fax-in device at the fax-out device can trigger 'fax-out-modem-confirmation-not-received'. This is how the sending end will indicate that it lost its connection to the receiver.
>> We still could have 'fax-modem-carrier-lost', but on the receiving device, and probably it is good to rename it as 'fax-in-modem-carrier-lost' indicating that this is an error on the fax-in device. Likewise, we could have 'fax-modem-training-failure' but on the receiving device only and it is good to rename it (something) as 'fax-in-modem-training-failure' or 'fax-in-modem-failure-to-train.
>> I like to have the 'fax-out-*' and 'fax-in-*' syntax as it would be less confusing as to what mode a device is acting (as a sender or receiver).
>> Thanks,
> Daniel.
>>>> -----Original Message-----
> From: ipp-bounces at pwg.org [mailto:ipp-bounces at pwg.org] On Behalf Of
> Michael Sweet
> Sent: Thursday, June 27, 2013 5:07 AM
> To: fx OHTAKE SHIN
> Cc: ipp at pwg.org> Subject: [IPP] Re: Fuji Xerox has reviewed the IPP FaxOut
> specification and has comments
>> Shin,
>> Thanks for your comments!
>> Responses inline...
>> On 2013-06-26, at 10:25 PM, fx OHTAKE SHIN <shin.ohtake at fujixerox.co.jp> wrote:
>> Greetings,
>>>> Fuji Xerox has comments for the IPP FaxOut specification final call.
>>>> ----
>> 8.2 job-state-reasons
>> 1. fax-modem-carrier-lost
>> See ITU-T T.30(09/2005)- Annex A Examples 13, FAX MSG carrier is lost
>> by Called terminal. We think fax-modem-carrier-lost should be removed from the specification.
>> If the specification supposes another case, tell us an example on the facsimile protocol sequence.
>> But on the sending end we still need an indication that we lost the connection to the receiver. Would changing the definition here to something like "Lost connection to the receiver during send." be more descriptive?
>> (in AT command set parlance, this would be a "NO CARRIER" response,
> received when the connection is lost)
>>> 2. fax-modem-training-failure
>> See ITU-T T.30(09/2005)-Appendix IV Examples 4, Training failure is
>> detected by Called terminal. We think fax-modem-training-failure should be removed from the specification.
>> If the specification supposes another case, tell us an example on the facsimile protocol sequence.
>> But on the sending end we still need an indication that we were unable to connect to the receiver after the receiver answered the call. This definition has already been updated once, but I am happy to reword it again to make this clear - basically, 'fax-modem-training-failure' means that the receiver answered, the sender got the carrier tone, but the sender and receiver were unable to successfully negotiate a data rate.
>> (in AT command set parlance, this is also a "NO CARRIER" response,
> received after dialing)
>>> 3. fax-modem-voice-detected
>> If fax-modem-voice-detected specified as a job state, it should
>> change name to job-calling. Facsimile always detect sound other than
>> a carrier tone while facsimile dialing numbers to detecting any
>> facsimile signals, because facsimile may hear ringing tone or machine's voice(voice answer system), human's voice(fax manual receive). Therefore, fax-modem-voice-detected cannot specify as an error job reason, because of ringing tone.
>> I'm not sure I understand your comment, but a couple things:
>> 1. We can't change the names of these keywords: they were defined and
> approved last year as part of the Printer MIB and IPP MFD Alerts spec
> (PWG 5107.3-2012). Which of course I don't have referenced and am
> re-defining the names in the IANA section (editorial changes at
> least...) [NOTE TO SELF: Add 5107.3 IANA definitions to IPP
> registrations]
>> 2. We do want a way to indicate when the receiver answers but does not respond with a carrier within the connection timeout setup in the modem. Again, falling back on the AT command set I would probably just get a "NO CARRIER" response and a hang-up, but some modems did/do report "DELAYED" for the additional rings - we *could* add a keyword for that ("fax-modem-delayed"? "fax-modem-waiting"?) but I'd like to leave this one as-is even if many implementations can't support 'fax-modem-voice-detected'.
>> 3. None of these are unconditionally required to implement - certain
> keywords depend on hardware features that may not be supported by your
> products, and so naturally you won't report those (similar to how you
> might not have a duplexing unit in your printer and thus don't have to
> report or support the "sides" attribute...)
>> _________________________________________________________
> Michael Sweet, Senior Printing System Engineer, PWG Chair
>>> --
> This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.
>> _______________________________________________
> ipp mailing list
>ipp at pwg.org>https://www.pwg.org/mailman/listinfo/ipp
_________________________________________________________
Michael Sweet, Senior Printing System Engineer, PWG Chair
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.