Re: [sipcore] WGLC: draft-ietf-sipcore-location-conveyance

[Robert Sparks <[e-mail]>]

Thanks for the reminder - I reread the text and talked with Jon. I think the only thing that needs to change
is the actual error code. 503 is to extreme - it would cut off all SIP traffic from the peer receiving the response,
not just traffic related to this request. I think the right code in this context is 500 (which can also use the Retry-After
mechanism).

More concisely, my suggestion is s/503/500/

RjS

On Mar 29, 2011, at 10:47 AM, James M. Polk wrote:

> At 03:45 PM 3/9/2011, Robert Sparks wrote:
>> There are a couple of larger and several small things to also address in this document before requesting publication.
>> 
>> The larger things:
>> 
>> 3) I can't figure out what the paragraph in section 4.4 that starts "Additionally, if a sip entity cannot..." and goes
>> onto talk about 503's is trying to say. I think some surrounding context must have been lost or not captured.
>> Why would we be recommending sending a 503 here?
> 
> pick on this one point, there was an error code for that basically said "couldn't process the location you just sent me, but it's ok to try again" *and* "couldn't process the location you just sent me, don't try to resend it for a while". Jon didn't like the idea of something that he feels duplicates the 503 with a retry after header field. He was on a mission to reduce the number of geolocation-error codes, and these two go whacked. That's where this text comes from. I'll get with him about writing this section better - or do you have an alternate plan we should pursue?
> 
> James