[Asterisk-Dev] rfc2833 DTMFs sent with bad timestamps (patch)
lconroy
lconroy at insensate.co.uk
Sat Feb 26 19:19:54 MST 2005
Hi folks,
Cisco - not RFC-compliant? Surely not.
(Anyone not spent time "being VERY liberal in what you accept" to work
with a Cisco device?)
[e.g. having from: user name followed by the SIP AoR, with no < > to
separate them. Ah - hours of fun]
I believe, looking at 2833 (and the -bis8 draft) that each event SHOULD
be distinguished with a different timestamp, but that repetitions of a
single event SHOULD NOT use a different timestamp.
(That was certainly my understanding of how it was supposed to work
during the years of discussion we had getting it done in the first
place).
Of course, every frame should have an incremented serial number, or
they will appear to be duplicate packets created from nothing by wonky
intermediate routers.
In short, anyone sending multiple events all with the same timestamp is
pushing their luck.
If you want the definitive answer, the guy that's corralling the update
to 2833 is Tom Taylor of Nortel, who's a very helpful chap.
If you look through the IETF AVT mailing list archives (e.g. in January
this year) there's a number of threads he's been posting on details of
how it is supposed to work. They're quite useful in getting your head
around how it's supposed to work.
There is also an interesting snippet on setting the "M" flag with the
first frame of an event, and how to deal with that frame getting lost.
Worth a look, IMHO.
all the best,
Lawrence
On 26 Feb 2005, at 22:00, Brian West wrote:
> Ya i see this but when I did the last fix on the DTMF I used a lucent
> TNT and a 7960 which didn't ever increment the timestamps from digit
> to digit. So either these aren't totally compliant or the RFC warp
> them too! :P I posted on your bug on mantis too ;)
>
> /b
>
> On Feb 26, 2005, at 2:58 PM, Frank van Dijk wrote:
>
>> Frank van Dijk wrote:
>>
>>>
>>> Attached is a summary of an ethereal trace that shows the cisco
>>> gateway I talk to does change the ts from event to event but not for
>>> packets of the same event.
>>
>> *sigh*
>>
>> 1077 5.380739 <cisco> <asterisk> RTP Payload
>> type=ITU-T G.711 PCMA, SSRC=479656997, Seq=2025, Time=2577695885
>> Sequence number: 2025
>> Timestamp: 2577695885
>> 1083 5.420722 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4
>> Sequence number: 2026
>> Timestamp: 2577696285
>> 1086 5.430510 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4
>> Sequence number: 2027
>> Timestamp: 2577696285
>> 1087 5.440191 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4
>> Sequence number: 2028
>> Timestamp: 2577696285
>> 1092 5.470357 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4
>> Sequence number: 2029
>> Timestamp: 2577696285
>> 1097 5.521323 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4
>> Sequence number: 2030
>> Timestamp: 2577696285
>> 1098 5.521331 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4 (end)
>> Sequence number: 2031
>> Timestamp: 2577696285
>> 1105 5.528187 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4 (end)
>> Sequence number: 2032
>> Timestamp: 2577696285
>> 1106 5.540936 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Four 4 (end)
>> Sequence number: 2033
>> Timestamp: 2577696285
>> 1115 5.620470 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5
>> Sequence number: 2034
>> Timestamp: 2577697885
>> 1118 5.630274 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5
>> Sequence number: 2035
>> Timestamp: 2577697885
>> 1119 5.640401 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5
>> Sequence number: 2036
>> Timestamp: 2577697885
>> 1125 5.668845 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5
>> Sequence number: 2037
>> Timestamp: 2577697885
>> 1128 5.700174 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5 (end)
>> Sequence number: 2038
>> Timestamp: 2577697885
>> 1135 5.708510 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5 (end)
>> Sequence number: 2039
>> Timestamp: 2577697885
>> 1136 5.720080 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Five 5 (end)
>> Sequence number: 2040
>> Timestamp: 2577697885
>> 1143 5.780444 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6
>> Sequence number: 2041
>> Timestamp: 2577699165
>> 1146 5.790403 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6
>> Sequence number: 2042
>> Timestamp: 2577699165
>> 1147 5.800218 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6
>> Sequence number: 2043
>> Timestamp: 2577699165
>> 1152 5.828618 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6
>> Sequence number: 2044
>> Timestamp: 2577699165
>> 1155 5.860286 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6 (end)
>> Sequence number: 2045
>> Timestamp: 2577699165
>> 1162 5.868335 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6 (end)
>> Sequence number: 2046
>> Timestamp: 2577699165
>> 1163 5.880492 <cisco> <asterisk> RTP EVENT
>> Payload type=RTP Event, DTMF Six 6 (end)
>> Sequence number: 2047
>> Timestamp: 2577699165
>> _______________________________________________
>> Asterisk-Dev mailing list
>> Asterisk-Dev at lists.digium.com
>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>> To UNSUBSCRIBE or update options visit:
>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>
> _______________________________________________
> Asterisk-Dev mailing list
> Asterisk-Dev at lists.digium.com
> http://lists.digium.com/mailman/listinfo/asterisk-dev
> To UNSUBSCRIBE or update options visit:
> http://lists.digium.com/mailman/listinfo/asterisk-dev
>
>
---------------------------------------
lawrence conroy |tel:+44-1794-833666
More information about the asterisk-dev
mailing list