Re: Starting in on a document

Hadmut Danisch wrote:
> 
> On Wed, Oct 15, 2003 at 02:11:28PM -0400, Alan DeKok wrote:
> >   My opinion is that we should describe the smallest design and
> > implementation which will address as much of the problem as possible.
> 
> No. For two reasons: First, we don't have a precise description of the
> problem yet.

Seems to me that we do; forgeries of SMTP MAIL FROM parameters and forgeries
of SMTP HELO parameters.

> Second: Keep the 80/20 rule in mind. You usually can solve 80% of the
> problem with 20% of the work/cost. Following your proposol, we'd have
> to describe a solution which is cheap (20%), but does solve only 80%
> of the problem. This won't sell.

Also keep in mind that implementing this is voluntary. If the solution
is too large, too complex, or causes too much collateral damage, it will
be ignored.

> > To me, this means Q(DOMAIN, IP), individual OK's, no new RR type, and
> > implemented via A RR.
> 
> And I received heaps of mails from people commenting my RMX drafts who
> told me that this is not enough.

I suspect that a large percentage of those comments were of the "I don't
want to have to do anything different" variety.


Raymond S Brand