Re: Summary of Closing the offer/answer rollback issue?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Hadriel,

I guess Gonzalo will provide his summary soon, but I don't think there
is currently a disagreement regarding pre-conditions. You call them
"conditional offers", Gao calls them "notifications", but that's just
wording... :)

I think the one of the main issues at the moment is what happens after
preconditions have been met on both sides:

1) Is the change now commited/in-use, and a re-INVITE failure would not
change that?                                       <----- "in-use"
alternative

OR

2) Would a re-INVITE failure cause a fallback (this is what is meant by
"late commitment")?                                <---- "late
commitment" alternative


If preconditions have NOT been met, and the re-INVITE fails, I think
most agree that there will be a rollback to the "last committed state".


Gao's draft also talk about other changes which would not be considered
as "real" SDP offers, for example if you reduce codecs. But, at least I
strongly prefer NOT to go into such details, because that would for sure
cause interop issues. I am not sure what Gao's view on that is at the
moment, though?


Regarding the race condition, I think we can avoid that with some BCP-
and guidance text.


Regards,

Christer

 

-----Original Message-----
From: Hadriel Kaplan [mailto:HKaplan@xxxxxxxxxxxxxx] 
Sent: Saturday, February 28, 2009 5:26 PM
To: Christer Holmberg; Gonzalo Camarillo; sipping
Subject: Summary of Closing the offer/answer rollback issue?


Is there an email in this long thread that summarizes the issues?
It's impossible to follow this thread. :)

I'm not clear on what the issues are with pre-conditions that make
failed offers concept break.  To me, pre-conditions are basically not
real SDP offers; they're conditional offers.  Until the *all* the
conditions are met, it's not "committed".  You continue using the last
committed state.  I know there are race conditions, but considering how
rare pre-conditions are in the real world (especially in re-Invite's),
that I'm having trouble imagining why we should care about corner cases
of such.  Regardless, I vote for any fixing that needs to happen because
of pre-conditions should be around changing pre-condition logic, even if
it means completely re-writing how pre-conditions works - don't change
normal SIP or SDP. (not that anyone is proposing it, I just can't tell
form this thread)

Re-Invites and SDP offer/answer have so many interop issues in the wild
even without pre-conditions, that this whole thread scares me. :(

-hadriel

_______________________________________________
Sipping mailing list  https://www.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip
Use sip@xxxxxxxx for new developments of core SIP

[Index of Archives]     [IETF Announce]     [IETF Discussion]     [Linux SCSI]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Big List of Linux Books]

  Powered by Linux