[rpd] End of LAST call

[Nishal Goburdhan]

On 1 Feb 2020, at 18:10, Anthony Ubah wrote:


> Hello Nishal Goburdhan,


hi,


> I quite support the proposition to track views shared by members of 

> the

> community, and responses provided by the Author(s) and community on 

> every

> policy proposed.

> This will go a long way in making decision making easier for the 

> Co-Chairs,

> Afrinioc staff and the community at large. It introduces more 

> transparency.

> That way, strong/weak/irrelevant or passive suggestions and objections 

> can

> be followed up and fully considered and/or addressed.

> I suggest a platform for this should be set aside and updated 

> regularly.

> This will shrink repetitions and save the community valuable time. We 

> all

> have different schedules, and everyone can not invest the minimum time

> required to revisit every comment posted in the rpd mailing list.


i am pretty certain i didn’t ask for an all-in-one tracking comment 
system (but hey, i’m not opposed to that, if that’s the route that 
the co-chairs and afrinic want to build .. )

all i’d like, is to see an explanation of _why_ something did not 
reach consensus, that can help both the authors, and the community, 
drive the proposed policy forward;  either to consensus that it’s good 
and needed, or bad, and unnecessary.
something as simple as:
# here were some of the concerns _about the policy_ that we, as 
co-chairs felt were not addressed.
and that needn’t be left to the end of the 14d period either.   i mean 
- i see nothing wrong with the co-chairs saying: “folks, you’re 
going in circles on issue Z;  whilst issue Y still lies unaddressed”.  
anyway.

fwiw, i’ve been a policy author that’s been sent back to the drawing 
board, for reasons that i *still* don’t quite understand  (so, i’m 
really grateful to my co-authors).  and having a concise list of 
objections, would have really been helpful to me, at the time.



> With respect to the above, a typical example would be my query on this

> RPKI ROAs for Unallocated and Unassigned AFRINIC Address Space policy

> proposal, a few of my inquiries were addressed, however, I questioned 

> the

> effectiveness of having such policy on a regional scale where the true

> benefit can only be enjoyed if implemented as a global policy.

> I ask again, why not push for a Global policy proposal for RPKI ROAs 

> for

> Unallocated and Unassigned resources? If I didn't miss the response to

> that, then it still hasn't been addressed.


i’m not an author, but i’ll try to answer.  there is zero reason for 
this to be a global policy _at all_.

it is true that if
* all the RIR communities did this, and;
* every one implemented policy like this, and;
* all major providers and IXPs dropped INVALID ROAs;
.. then the internet will be safer (route wise) overall.

but, it’s equally true, that if *only* one RIR did this, then, that 
single initiative would *still* make the internet safer;  just for 
resources that were allocated/assigned from that RIR.  so afrinic doing 
this, is a Good Thing for *our* resources (that afrinic is the custodian 
of).  and _that_ is what should matter;   ie.  is afrinic doing the Best 
Correct Thing Possible, to keep its constituents’ resources safe?

there’s nothing stopping someone from creating a global policy for 
this.  but, a global policy is _really not necessary_, to get the same 
nett effect as the intent of what the current policy under discussion in 
this region _will bring to our resources_.  and that is: “how do we 
make afrinic’s unassigned prefixes less susceptible to unauthorised 
use”.  focus on that, because that’s really the thing that warrants 
discussion!
to delay, in adopting something that will help prevent misuse of 
afrinic’s resource, on the presumption that there _might_ be a global 
policy someday, to do the same thing, is a Bad Way to think about policy 
(and network security).

in an earlier reply, i wrote:
“you are ignoring that there *is* already global consensus, at least 
in  the routing world, on what to do with AS0 ROAs. see:
https://tools.ietf.org/html/rfc6483#section-4. “
so the routing systems are, and have been, ready for this.  we just need 
the record keepers (ie. RIRs) to set the correct flags.

maybe, it would be better if you could explain why you think a global 
policy would be better;  because you might have a viewpoint that some of 
us are missing?

—n.