Re: [sidr] pCNT & (AS_PATH) prepending: Is it in scope?

[Randy Bush <[e-mail]>]

not your fault, but very hard to find your text amid the top posting,
lack of line wraps, quoting disasters, ...

> I do not think that it was simplification that led to the deprecation
> of AS-SET, rather the difficulty or impossibility of securing it to
> the same extent as a single AS; which led to the discovery that AS_SET
> were rare and that their loss would not affect almost all of the
> Internet.

we knew as-sets were rare.  but before moving to deprecate we decided to
actually measure how rare.

note that as-sets affect both origin validation and path validation.

> Question is; how common is prepending?  I thought that it was
> widespread and 'normal' but there would have to be hard data first,
> before deprecation could be contemplated.

we could measure.  but given that we can see that it is quite common,
and we have reasonable ways to deal with it, why should we spend the
time?  what might we learn?

randy