On 01/07/2020 19:56, Matthias Merkel via db-wg wrote: > Given the small number of objects, I don't think implementing a filter > would have any significant impact (negative or positive) and may not be > worth the effort.
Perhaps as a result of my conceiving of it, there is a conceivable social engineering angle here, which could be put to use to create problems in the GRT. I would expect the minimum effort to protect against this - preventing any *new* mntners masquerading as aut-nums - as being a worthwhile investment of effort. As per Job's comment at ~18:41Z, this could be handled with a creation filter, of which there is precedent. It would certainly be worth putting a stop to the capability /before/ someone finds a way to exploit it, and we're all left looking at our feet. :) -- Tom
