Archive

Yes, why don't we just do it? This group seems to have stalled. I don't have a strong opinion on the information model, so why not take Cullen's suggestion as a starting point and try to get consensus (having a call if necessary)? SIP-CLF is a useful work item, but not one worth spending years trying to find the optimum solution.

John

> -----Original Message-----
> From: sip-clf-bounces@ietf.org 
> [mailto:sip-clf-bounces@ietf.org] On Behalf Of Cullen Jennings
> Sent: 12 January 2010 22:22
> To: SIP-CLF Mailing List
> Subject: [sip-clf] One possible Information model for SIP CLF
> 
> 
> David raises and excellent point that we are unclear about 
> our information model and that makes it harder for us to get 
> work done quickly. I encourage people to read RFC 3444.
> 
> I think there is a very clear and well define information 
> model for what we need in SIP CLF. 
> 
> For each SIP message that is logged, there is a Status-Line 
> or Request-Line, an ordered list of pairs of where each pair 
> has a header-name and a header-value, and finally there may 
> be zero or one message-bodies. We also have where the message 
> was coming from and going to from the transport layer as well 
> as a timestamp. These are defined in RFC 3261. Formal 
> language descriptions of one possible data model for them are 
> provided in section 25.1. If people feel the strong need to 
> have this information model in say UML, I suspect I could do 
> the UML for this in ASCII. 
> 
> This is it. 
> 
> As usual, I don't claim to have a good handle on where this 
> slides from being an IM to DM but it barely seems relevant 
> for getting the job done. Now of course we could slide the IM 
> differently, such as pulling out the SIP to and from tags as 
> elements in the information model. I don't really care how we 
> slice it apart but the key thing to me that doing so is 
> really easy and largely already defined. I think that on a 
> single 1 hour phone call we could probably come to consensus 
> on what the information model needed to be. 
> 
> I would strongly argue that what we don't need is an 
> extensible information model or a formal language schema for 
> describing and extensible data model. SIP 2.0 is not going to 
> be extended in a way that adds elements outside the above items. 
> 
> Keep it simple. You are trying to convince someone that 
> currently has an fprintf to replace the it with this so make it easy. 
> 
> 
> Cullen
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> sip-clf mailing list
> sip-clf@ietf.org
> https://www.ietf.org/mailman/listinfo/sip-clf
> 

Re: [sip-clf] One possible Information model for SIP CLF