I still didn't find the source for 3 seconds, but I did more research, according to Wikipedia, http://en.wikipedia.org/wiki/Hayes_command_set, the default value for pause is 2 seconds, but it's configurable from 0-255. Finding such information inside 3GPP spec seems really hard, but so far it might be per protocol or per device decision. The UI should probably not care at all and instead leave it to FS to do the right thing when we fall back to encoding it ourself into the stream.

We shouldn't change DTMF things without input from people who know telephony. Cc'ing the telephonic cabal :-)

Nicolas, you got it right. The pause value of 3 seconds is from the 3GPP recommendation 22.101, A.21:

A.21 DTMF control digits separator

Provision has been made to enter DTMF digits with a telephone number, and upon the called party answering the UE shall send the DTMF digits automatically to the network after a delay of 3 seconds (± 20 %). The digits shall be sent according to the procedures and timing specified in 3GPP TS 24.008 [13].  The first occurrence of the "DTMF Control Digits Separator" shall be used by the ME to distinguish between the addressing digits (i.e. the phone number) and the DTMF digits. Upon subsequent occurrences of the separator, the UE shall pause again for 3 seconds (± 20 %) before sending any further DTMF digits. To enable the separator to be stored in the address field of an Abbreviated Dialling Number record in the SIM/USIM, the separator shall be coded as defined in 3GPP TS 31.102 [19]. The telephone number shall always precede the DTMF digits when stored in the SIM/USIM. The way in which the separator is entered and display in the UE, is left to the individual manufacturer's MMI. MEs which do not support this feature and encounter this separator in an ADN record of the SIM/USIM will treat the
character as "corrupt data" and act accordingly.

Perhaps the spec should mention that the pause is also protocol-dependent?

Thanks Pekka, I've update the title to reflect this conclusion. I was also thinking of it, and it might be fine to have a fix pause duration in Telepathy. 

In today's implementation, we very often link address book and contacts, which mean if one shares a voicemail contact between his phone and his laptop (which contain some p to automate the access) and want to use it over SIP on his laptop, the automated number might not work has SIP recommendation is 1 second which is a lot shorter. We could simply add some rational in the SPEC to mention this.

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/telepathy/telepathy-spec/issues/112.