what is a problem

Marshall Rose mrose@dbc.mtview.ca.us
Wed, 13 Nov 2002 15:32:10 -0800


> A problem, for example, from an AD perspective, may be
> that ADs have the responsibility for making sure that stuff
> that goes through IETF process produces implementable,
> reasonable standards.  And yet they only have 2 
> checkpoints in a work item's history:  WG chartering
> time, and document publication time.  In between
> those 2 points, much may happen to take things in
> a completely different direction.  (Yes, AD's
> can/should be involved through the course of a WG's
> life, but honestly, independently of time commitments,
> that's indirect persuasion).  

well, if that's how the ADs are running their areas now, then i'm
surprised there aren't more problems. you're correct that there are only
two formal checkpoints. however, the responsible AD (or a delegate)
should be providing constant feedback to the WG... while there may be
consternation, there shouldn't be any surprise.
    
    
> A problem, from an IETF participant perspective, may be 
> that IETF participants can carry a given work item so 
> far (e.g., through WG process) and then the work item 
> enters a state that the peopple who have worked on it 
> can no longer determine what is happening to it, whether 
> it is progressing, whether or what issues there are.  
> Although they are collectively responsible for completing 
> the work item, they have no way to read what is going on, 
> or when it will re-enter a more determinate state.  And, of 
> late, this state has typically lasted from weeks to months 
> to years. This is an *IETF* problem, because it means that 
> there is a huge chunk of indeterminism in the middle of 
> trying to get work done.
> 
> An IESG charter and transparency might be implemented
> as part of the solution to these problems -- but I think
> our discussion for now had better focus on getting
> agreement on what the actual work-stoppage, root
> cause problems are: things that, if they were magically
> addressed, would mean the IETF was functioning well.

i refer you to the PACT I-D that was written by geoff and myself.
    
    PACT = Predictability, Accountability, Competency, Timeliness...
    
/mtr