ietf-mxcomp
[Top] [All Lists]

Re: Ruminations

2004-03-28 15:32:05

--Dave Crocker <dhc(_at_)dcrocker(_dot_)net> wrote:


Gordon,


GF> A breakage deemed acceptable by some is blasphemous, in the most
religious GF> sense, to others.  We need to come to a consensus as to
what should be broken GF> and what should not before we can proceed[1].
Patrik said this, Ted said GF> this, Dave said this.

I do not recall endorsing breaking the mail service.


I think Gordon's claim was that you said something like "we need to come to consensus.." not necessarily that "breaking stuff is ok". (Just a guess... he may have been talking about a different Dave altogether :)

Speaking for myself, I *want* the current system to be as broken as possible for forgeries, while breaking as little as possible for legitimate uses. Herein lies the rub, of course, since there is a sliding scale between "break nothing" and "break a lot".

Perhaps a long-term plan is called for where we have some immediate stuff to phase in (for example, stuff that will break nothing at all and enforces pre-existing RFCs) and some other stuff that phases in later (like, stating that a previous practice is "deprecated" and *should* be phased out but leaving a sunset interval for it)

Even so, there will be some tradeoffs to make. "Break nothing that works now" is not a viable plan for introducing change. (I don't think that's what Dave was voting for but the above one-liner could be misinterpreted that way)

gregc

--
Greg Connor <gconnor(_at_)nekodojo(_dot_)org>


<Prev in Thread] Current Thread [Next in Thread>