ietf-mailsig
[Top] [All Lists]

Re: Why we really don't require requirements

2004-10-02 04:32:45

On Sep 30, 2004, at 2:53 PM, ned(_dot_)freed(_at_)mrochek(_dot_)com wrote:

I don't know enough
about CRISP to comment, but if GEOPRIV is an example of a success due to having used a requirements document, well, all I can say is it is at best a very anomalous case. IMO the whole GEOPRIV effort was mismanaged dating back to its origins in the SPATIAL effort. (I also think almost all of the blame for this rests with the IESG and not the group itself, and since I was an IESG member at the time I include myself as a responsible party. But that's an
entirely separate discussion not relevant to the matters at hand.)

I'm not sure what the point is here... that the IESG is at fault? That certainly doesn't speak against the point I made: GEOPRIV spent a long time in fundamental disagreement and finally quit wondering around after they put their requirements in writing.

On Oct 1, 2004, at 3:29 PM, ned(_dot_)freed(_at_)mrochek(_dot_)com wrote:

IMO what's missing from the charter is something that's really hard to add: The major failing we have seen over and over and over again in this space is that we've let the practically undeployable "best" be the mortal enemy of the very
deployable "just barely good enough".

Deployability is what matters. Period. A solution which offers only a small increase in security (e.g., it can with some trouble be sppofed, or replays are hard but possible, or whatever) but which is deployable is going to be a million times more valuable than any of the schems we've devised in the past,
even though those schems offer much much better security.

I would agree with this. And I think this was the issue we had at the last BoF. This point was not clearly articulated. Instead we spent too much time arguing about phishing prevention being a social or technical issue.

The trick is going to be coming up with something that is both deployable and
which we can get approved. I'm honestly not optimistic that there's an
intersection between the two, even at this late date and after so many
failures.

Just my opinion, but I think the goal is attainable.

On Sep 30, 2004, at 8:57 PM, John Levine wrote:

It certainly seems logical that one would start by defining
requirements and then refine them into a spec and eventually into
working code.  This is basically the waterfall model of software
design that was popular in the 1960s.  It's fallen our of favor,
because it's proven many times to be a one-way ticket to disaster.
The short reason is because if we knew what we were going to build, we
would have built it already.

Requirements definition is found in more than just the waterfall software development methodology. Are you seriously trying to suggest that requirements for software development is just a 35-year-old, faded and debunked fad?

On Oct 1, 2004, at 9:56 PM, John Levine wrote:

This is exactly the kind of war by proxy that I was referring to.

You are the one that said this was an obvious but debatable requirement. I'm not sure how it can be both, but the fact that it is being debated means that it is not obvious to everyone. I suspect this isn't the only requirement that isn't obvious.

If you don't want a "war by proxy", then be prepared to repeat these conversations over and over again as this group goes through each proposal. Or do you suggest we simply give up/down votes for each proposal without discussion? Because that is the only way this group will avoid these conversations.

I look forward to multiple and repeated discussions about the various merits of key management systems, either putting keys directly into DNS or just using DNS as a pointer, re-using S/MIME or PGP vs a new mail encryption scheme, the identity from which the key is to be derived, preservation of forwarding semantics, etc..

I'd like an automobile that costs nothing, protects the occupants from
injury in any possible type of accident, gets a thousand miles to the
gallon, and has no effect on the environment, but making those my
requirements isn't going to help me buy a car.

So we should trust the car salesman instead of doing our homework? You have obviously had much different car purchasing experiences than I.

-andy


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