RE: Late Last Call Gen-ART review of draft-ietf-avt-rfc4749-dtx-update-01

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

 



Hi

The current text in section 3 of my draft is mostly a copy/paste from RFC 3551. Are you suggesting to replace the whole block by a reference to RFC3551 section 4.1?

With you comment that it is specific to G.729.1, I see no reason for an implementation to not set M=1 as expected. So I'm also OK to replace the SHOULD by a MUST in the current text.

What option seems the best?

Aurelien


> -----Message d'origine-----
> De : Roni Even [mailto:ron.even.tlv@xxxxxxxxx] 
> Envoyé : mardi 23 septembre 2008 12:46
> À : 'Spencer Dawkins'; SOLLAUD Aurelien RD-CORE-LAN
> Cc : fluffy@xxxxxxxxx; ietf@xxxxxxxx; gen-art@xxxxxxxx; 
> tom.taylor@xxxxxxxxxx
> Objet : RE: Late Last Call Gen-ART review of 
> draft-ietf-avt-rfc4749-dtx-update-01
> 
> Hi,
> This text is just relevant to G.729.1 and not to other 
> codecs, so if you want to add a reason for the SHOULD it 
> should be based on RFC4749 implementations that would not 
> send M=1. Maybe the text should say that if DTX is offered 
> the sender must set the M bit according to RFC 3551.
> Roni
> 
> -----Original Message-----
> From: Spencer Dawkins [mailto:spencer@xxxxxxxxxxxxxxxxx]
> Sent: Tuesday, September 23, 2008 12:45 PM
> To: aurelien.sollaud@xxxxxxxxxxxxxxxxxx
> Cc: fluffy@xxxxxxxxx; ietf@xxxxxxxx; gen-art@xxxxxxxx; 
> ron.even.tlv@xxxxxxxxx; tom.taylor@xxxxxxxxxx
> Subject: Re: Late Last Call Gen-ART review of
> draft-ietf-avt-rfc4749-dtx-update-01
> 
> Hi, Aurelien,
> 
> I think we're good on almost everything in your response.
> 
> On the SHOULD/MUST below, it could very well be OK to have 
> this as SHOULD, but we've been asking for some indication of 
> reasons why SHOULDs might not be implemented - which could be 
> as simple as "there is a lot of deployed code that didn't 
> implement this, because it was a SHOULD in RFC 3551".
> 
> If you leave this as SHOULD, you might want to say something 
> about the effect on receivers, since conformant sender 
> implementations aren't doing something that the spec assumed 
> they would be doing. In this case, you're saying that the 
> receiver can't look at the M-bit to identify the beginning of 
> a talkspurt, right?
> 
> If you get any other feedback about SHOULD/MUST here, please 
> take that into account, of course...
> 
> Thanks,
> 
> Spencer
> 
> ----- Original Message -----
> From: <aurelien.sollaud@xxxxxxxxxxxxxxxxxx>
> To: <spencer@xxxxxxxxxxxxxxxxx>
> Cc: <fluffy@xxxxxxxxx>; <ietf@xxxxxxxx>; <gen-art@xxxxxxxx>; 
> <ron.even.tlv@xxxxxxxxx>; <tom.taylor@xxxxxxxxxx>
> Sent: Tuesday, September 23, 2008 4:20 AM
> Subject: RE: Late Last Call Gen-ART review of
> draft-ietf-avt-rfc4749-dtx-update-01
> 
> 
> Hi
> 
> Thank you for the review.
> Below some answers.
> 
> Aurelien
> 
> > 3.  RTP Header Usage
> >
> >    If DTX is used, the first packet of a talkspurt, that 
> is, the first
> >    packet after a silence period during which packets have not been
> >    transmitted contiguously, SHOULD be distinguished by 
> setting the M
> >
> > Spencer (review): why not MUST here?
> >
> 
> [AS] It is the wording from RFC 3551 (4.1).
> It could be a MUST, but I saw no reason to be stronger than 
> the RTP spec. 
> 
> 
> 
_______________________________________________

Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf


[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]