Re: [Gen-art] Genart last call review of draft-ietf-pim-drlb-13

["Pete Resnick" <[e-mail]>]

Hi Alvaro,

On 7 Nov 2019, at 10:35, Alvaro Retana wrote:

> On November 5, 2019 at 6:58:52 PM, Pete Resnick wrote:
>
>>> Yes, IPR has been declared and the WG has been notified.
>> ----
>>
>> That seems to indicate that nobody had any comment about the IPR 
>> declaration.
>> But I also see noted in the shepherd report, "Cisco has an 
>> implementation of
>> this protocol. No other vendors have indicated plan to implement the
>> specification".
>
> Not having discussions about IPR is normal in the Routing Area.  I
> would consider having a discussion about it curious. ;-)

Heh.

> The sentence you quoted above from the Shepherd report is incomplete,
> it says: “No other vendors have indicated plan to implement the
> specification but they support publication of this draft.”

Well, the statement that someone "supports publication of this draft" I 
always take as a NOOP. Do they support publication because they like the 
author and think the author should have more publications? Or because 
they think the writing is poetic? Or because they want the author to 
support their own draft next time? Why would they support publication if 
they have no plans to implement it?

> I don’t
> know the reasons why other vendors are not implementing — another
> obvious possibility is simply that they don’t have immediate 
> customer
> demand.

Sure, if they think eventually there will be demand and that they'll 
implement later, that's a fine reason, and that would indicate that they 
don't think they'll have any IPR worries later. That said, the IPR 
declaration for this document still shows, "Licensing Declaration to be 
Provided Later"; perhaps everybody is fine with RAND for this case.

> The extension in this draft requires signaling and coordination
> between multiple routers; the dynamic nature of agreeing on which
> router is the GDR is what requires interoperability between the
> different routers.  This type of load balancing can be useful 
> wherever
> multiple PIM DRs exist, so I don’t think it can be considered a
> private extension, or applicable only in single-vendor deployments.

If only one vendor implements this and everybody else ignores it, it 
still works, right?

> Thanks for the careful review and for including interesting notes.

Always willing to cause interesting discussion. :-)

pr
-- 
Pete Resnick https://www.episteme.net/
All connections to the world are tenuous at best