On 2011-01-30 09:52, Dave CROCKER wrote:
On 1/29/2011 12:10 PM, Brian E Carpenter wrote:
On 2011-01-27 16:29, Scott O. Bradner wrote:
4/ as part of #3 - the rules should also specifically deal with
the following pp from 2026
The requirement for at least two independent and interoperable
implementations applies to all of the options and features of the
...
Actually the draft does not appear to require interoperability testing
at all:
"* There are a significant number of implementations with
successful operational experience."
Is that intentional? I thought interop was generally regarded as
People are confusing testing with use. Those are two different kinds of
"interoperability", with the latter being far more stringent.
The new draft specifies the latter. And it quite intentionally does not
specify the former.
Please point to the text that requires *any* kind of interoperability
being demonstrated by running code. "successful operational experience"
does not state or imply interoperation between independent implementations.
This is a big change in principle from 2026, which is not what is advertised
on the box as "primarily a reduction from three IETF standards track
maturity levels to two."
I want a two stage process, but I don't want to lose interoperability as
an explicit criterion. To me, that's always been the meaning of the
"running code" slogan.
Brian
While "testing" is extremely important for when doing development, there
is no reason that the IETF should be required to include that very
intermediary activity within our standards process.
So the new proposal has two phases:
1) Specification
2) Use
That there are intermediate real-world phases, such as development,
testing and deployment is essential, of course. But there is nothing
essential in having the IETF mark completion of any of those
intermediate phases.
d/
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf