Re: [pim] draft-ietf-pim-explicit-tracking-03 to standards track?

[Hitoshi Asaeda <[e-mail]>]

Lenny,

> At the time this was initially submitted, it was suggested that explicit 
> tracking was already implemented for years in a number of implementations 
> with no apparent interop problems and it was not clear why this 
> functionality needed to be specified and ought to be left up to the 
> implementation.  The fear was that unnecessary over-specification can 
> reduce the flexibility of implementations and could produce a spec that 
> was not backward compatible with existing implementations that worked just 
> fine.  The proposed compromise was to adopt the doc as informational, so 
> folks could understand how explicit tracking worked, but not lead to 
> over-specification and reduced flexibility for implementations.  I still 
> agree with this decision.

There are some optimization funstions such as specific query
suppression, which may not be implemented in some routers. You may say
these functions are over-spec. But they are not over-spec. They are
necessary functions and eager to be implememted especially in wireless
environments. You can see the fact in rfc6636,
draft-liu-multimob-igmp-mld-wireless-mobile, which has just started
adoption call, and other draft proposed in multimob.
Note this draft does not prohibit not to implement such features.
All these features, which can be omitted in some environments/
conditions, are defined as "optional". Then you can keep your
implementation simple.
Decision which features should be optional (i.e. use "MAY") was made
by the discussions in the previous PIM WG meetings. Please read the
latest draft and give me comments if you feel some other features,
which you feel over-spec, should be also optional. We can then start
the discussion about it in the list again.

And the explicit tracking function will be the important feature for
the future standard version of IGMP/MLD. There is no reason to keep it
as informational. Specification of the explicit tracking function to
be standard is necessary for the current and future needs.

I believe they are the reason at the last meeting we agreed on moving
it to proposed standard.
(For your information, please check the minutes of the last meeting;
http://www.ietf.org/proceedings/85/minutes/minutes-85-pim)

Regards,
--
Hitoshi Asaeda