ietf
[Top] [All Lists]

Re: Interoperability testing and the scope of IETF activities

2007-12-24 22:37:04

----- Original Message ----- From: "Dave Crocker" <dhc2(_at_)dcrocker(_dot_)net>
To: "Fred Baker" <fred(_at_)cisco(_dot_)com>
Cc: <ietf(_at_)ietf(_dot_)org>
Sent: Saturday, December 22, 2007 11:11 AM
Subject: Interoperability testing and the scope of IETF activities




Fred Baker wrote:
I found Dave Crocker's comment that the IETF never does interoperability testing equally amazing. The IETF has never offered the room (that is something a host does), but to describe the semi-annual PPP-fests that occurred throughout the 1990's, the OSPF testing that resulted in RFC 1246, and all the interoperability testing that has been done since, as unrelated to IETF activity is ... interesting. Here, I thought we were all IETF participants testing IETF specifications and products based on them, providing test results as called for in RFC 2026 and working group document feedback, and in the end demonstrating the "running code" part of the IETF mantra.


Fred,

There seems to be some confusion about the boundaries between "IETF activities" and "Internet activities." I am assuming that no rational person would assert that they are 1:1.

When a company is selling a product that incorporates an IETF standard, is that an IETF activity?

Sure if that product suite is one of the instances used to meet the IETF's two interoperable copies...


When a conference is held that discusses IETF standards, is that an IETF activity?

yes it is. And in fact if that instance generates NoteWell instances they need to be captured and submitted to the IP Archive that represents the materials used to create that IETF Standard.


When a company has employees attending the IETF and they get together for a company-specific dinner, is that an IETF activity?

Yes it is... At the very least ANYTIME money is spent on an IETF activity whether its dinner, drinks or meeting's where IETF IP is discussed or worked on, those are exactly IETF functions per the definition of NoteWell. Also NoteWell would pertain to two or members of the IETF having a discussion about IETF owned IP in the process of vetting it too so just a convo between two engineer's over a drink is exactly IP Development and covered under the IP ownership rules of the IETF so there is NO ISSUE as to whether this is IETF property right?


The list of things that are "related" to the IETF but are not part of it is probably infinite. Content, timing, location and participant names have pretty much nothing to do with whether it is an IETF activity. The question is whether is it "run" by the IETF.

No - actually the real issue is under NoteWell whether that IP is controlled or not.


Successful IETF work begins somewhere before the IETF.

meaning that the scope of IETF work also extends past the IETF itself and its meeting's. For instance if three people are sitting at a pizza parlor in Cambridge and they are discussing a NoteWell controlled IP and are designing their next response to the last I-D that was posted, would that then constitute IP owned by the IETF under its broad NoteWell definition? - the answer is YES... meaning that NoteWell is actually a curse that no one thought through but now that we are stuck with it, you folks need to know how broadly your definition of what is and isnt the IETF's IP under NoteWell is.

Bluntly NO ONE can touch anything the IETF does without stepping on NoteWell to some extent but hey - I bet that was the intent.


It is a process that discovers and assesses community need for some added capability and for willingness to work on it.

I havent laughed that hard in some time - thanks David... you funny.

In rare cases, an IETF BOF serves that role, but the usual experience is that waiting until them pretty much dooms the result.

Lesson 1: Without real community interest in using IETF output, it won't get used. So making sure the interest is there is paramount.

Why? - What is the Point? - The IETF isnt supposed to be selling anything - its supposed to allow people to create fair and open standards models that are all tested to some set of nebulous interoperability standards.


After the IETF does its part, the work continues among other people and groups. The IETF is the shepherd for only a part of the sequence. If any of the other parts fail to do their job, the technology will not get used. We seem to forget that.

Hmmm - Generally since the IETF le a trailing-edge standards process I would disagree here. The IETF does its thing generally after some protocol already exists.


     Lesson 2: The IETF does not create or sustain community interest.

Amen - but look at the statement following. It admits that the IETF intentionally 'ignores' efforts when it wants to. There is a clear damage that this does to others in the Standards process, and the IETF and its management needs to be accountable for that damage.

Witness how thoroughly we ignore post-Proposed efforts, except for the rare and half-hearted Draft exercise. The IETF operates open loop, with respect to the actual success of its work.

Which is another way of saying 'Look at what we prevent from becoming a standard'... and that 'we only allow standards for things we are interested in' - I wonder how many of the Sponsor's Lawyers are going to feel comfortable when they get a copy of this emailing...

I am betting the number is much larger than most of us think.


It is not a criticism of the IETF to observe that its role is limited. It is a criticism to observe that an serious constituency thinks otherwise.

And a criticism of the IETF for not building controls which would assure its fair and open practices were a reality instead of what I refer to as "boilerplate fantasy's" IMHO.

Todd Glassey

d/

--

  Dave Crocker
  Brandenburg InternetWorking
  bbiw.net

_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf


_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf