[sr-dev] about registrar outbound modes

Juha Heinanen jh at tutpro.com
Sat Mar 9 07:52:10 CET 2013


from registrar README:

3.24. outbound_mode (integer)

   If set to 0 then this module will accept REGISTER requests that do not
   contain a Supported: header with the outbound options-tag. The 200 OK
   response to REGISTER requests that this module generates will not
   contain Require: or Supported: headers with the outbound options tag.

   If set to 1 then this module will accept REGISTER requests that do not
   contain a Supported: header with the outbound options-tag. The 200 OK
   response to REGISTER requests that this module generates will contain a
   Supported: header with the outbound options tag.

   If set to 2 then this module will reject REGISTER requests that do not
   contain a Supported: header with the outbound options-tag. The 200 OK
   response to REGISTER requests that this module generates will contain
   Require: and Supported: headers with the outbound options tag.

looks likes there is no mode that could be used by registrar that
intends to support registrations from both outbound capable and
non-outbound capable UAs.

i would like to change mode 1 as follows:

   If set to 1 then this module will accept REGISTER requests that do not
   contain a Supported: header with the outbound options-tag. The 200 OK
   response to REGISTER requests that this module generates will contain a
   Require: or Supported: header with the outbound options tag
   depending on if REGISTER requests contain or do not contain,
   respectively, a Supported: header with the outbound options-tag.

without this modification UAs that support outbound don't get back a
response that tells them that registrar supports outbound:

4.2.1.  Initial Registrations
   ...
   If outbound registration succeeded, as indicated by the presence of
   the outbound option-tag in the Require header field of a successful
   registration response, the UA begins sending keep-alives as described
   in Section 4.4.

i would thus like to consider the current definition of mode 1 as a bug,
since it does not achieve anything more than mode 0.

comments?

-- juha



More information about the sr-dev mailing list