[asterisk-dev] [Code Review] Change T38 passthrough reinvite to use control frames
Joshua Colp
jcolp at digium.com
Wed Mar 18 11:08:27 CDT 2009
> On 2009-03-18 07:21:51, dimas wrote:
> > /branches/1.6.0/channels/chan_sip.c, line 5398
> > <http://reviewboard.digium.com/r/200/diff/1/?file=3116#file3116line5398>
> >
> > Are you sure the case should be for _NEGOTIATED and not _REQUEST_NEGOTIATE?
> > The app_fax sends _REQUEST_NEGOTIATE when it thinks it is time to switch to T38 and if we are in T38_PEER_REINVITE state at the moment - it is good idea to OK the transition.
> >
> > Also it seems to be good idea to pass _REQUEST_NEGOTIATE frame from channel driver to application too.
> > 1. It will allow app_fax to accept T38 at the moment the other side proposes without waiting for CNG. (app_fax needs to be updated for this)
> > 2. It will form some foundation for forwarding T38 requests further by the call chain - to IAX trunks etc.
Actually it should be for both, so I've changed it to be that way.
As for sending to applications this already happens with this code. It has no concept of what is on the other side 'nor does it care.
- Joshua
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviewboard.digium.com/r/200/#review568
-----------------------------------------------------------
On 2009-03-17 10:26:14, Joshua Colp wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviewboard.digium.com/r/200/
> -----------------------------------------------------------
>
> (Updated 2009-03-17 10:26:14)
>
>
> Review request for Asterisk Developers.
>
>
> Summary
> -------
>
> This patch changes the T38 passthrough reinvite method to use the existing T38 control frames. It also fixes an issue where T38 was accepted even though the remote side may be an application that was not aware of it.
>
>
> Diffs
> -----
>
> /branches/1.6.0/channels/chan_sip.c 182170
>
> Diff: http://reviewboard.digium.com/r/200/diff
>
>
> Testing
> -------
>
> The following scenarios have been tested:
>
> T38 Capable Device -> Asterisk -> Regular Phone - This scenario failed with a 488 which was expected
> T38 Capable Device -> Asterisk -> T38 Capable Device - This scenario worked with T38 negotiated on both sides
> T38 Capable Device -> Asterisk running an application - This scenario failed with a 488 which was expected
>
>
> Thanks,
>
> Joshua
>
>
More information about the asterisk-dev
mailing list