Re: [Ecrit] PhoneBCP

["Edge, Stephen" <[e-mail]>]

Bernard and others

This is an interesting view - an admission that the Ecrit solution will probably not be universally applied and when it is may be subject to modification. On the 3GPP/3GPP2 sides, solutions are defined with a specific and explicit scope and are expected to be - and nearly always actually are - supported exactly as defined in the appropriate spec.s for all scenarios covered by the scope. A solution whose deployment is unpredictable will be less useful because of the implied interworking problems between incompatible implementations. That is a main reason why I and others initially tried to include a more precise applicability/scope statement than the one we are now discussing. Specifically, the Ecrit solution will probably not be deployed by 3GPP/3GPP2 operators for cellular terminals because they have defined a somewhat different solution more suitable for cellular operation. That means that a terminal supporting only the Ecrit solution (based on an assertion of universal applicability) will encounter problems when its user attempts an emergency call on a 3GPP/3GPP2 network. If I can extend your analogies, this would be like not including a precise menu description for a meal containing meat and/or fish in a restaurant catering to both vegetarians and non-vegetarians.

Kind Regards

Stephen
________________________________
From: Bernard Aboba [mailto:[e-mail]]
Sent: Thursday, April 30, 2009 1:13 AM
To: Edge, Stephen; [e-mail]; [e-mail]
Subject: RE: [Ecrit] PhoneBCP

> Your interpretation of applicability is certainly extreme (yes or qualified yes to all questions) - removing any dependence on suitability or feasibility (which are the more normal benchmarks) and simply making a legalistic blanket assertion. (I can now more understand incidentally why the fictitious protocol police are so often invoked on matters of conformance.)
>
> But do others have any views on this?

Here is my 2 pence.   As noted in the Introduction, the Phone BCP document and framework documents are closely related documents that need to be read (and understood) together.  In looking at Phone BCP, there have been many situations in which  I wondered what the justification was for a requirement, only to find the (usually persuasive) explanation in the corresponding section of the framework document.

IMHO, the separation of the two documents is somewhat unnatural. It's a bit like taking a meal, freeze drying it, and then presenting the diner with a cup of powder and a cup of water and wondering why they are left with a quizical look on their faces.  Including cooking instructions along with the two cups is not a fully satisfactory answer.

Given this reality, if there is a need for text explaining the fundamental assumptions of the architecture, the place for that text is not in the BCP document, but in the framework document.  As has been noted by others, the entire BCP document is an applicability statement, and when viewed in that way, the proposed text is not only unnecessary, but somewhat confusing since required context for the BCP document belongs in the framework document.  If anything more is necesary, it would probably be some additional instructions for the reader  delving into the document set for the first time.

Like any IETF effort, the framework and BCP documents will be evaluated on their own merits. Athough this is an area with a considerable regulatory component, I see little evidence that the world at large is so in awe of the IETF that every word will be taken as immutable truth.  No doubt other SDOs will take issue with these documents in whole or in part and will explain the basis of their assessments and where necessary, their alternatives. As a result, some portions of the architecture may be adopted, some portions may be ignored, and others will be revised in an effort to address the issues encountered and improve their chances for deployment.

With respect to this process, the addition of cautionary text is a bit like the warning labels that we find on consumer products.  While it may be necessary in some legalistic sense, given the expertise of the individuals assessing these documents, one wonders what the practical effect will be.  Typically these are individuals who already understand that it is a good idea to double-cup hot coffee.