[ih] Cluster Addressing and CIDR

[Andrew Russell]

>At 18:21 -0700 1/14/03, Andrew Russell wrote:
>>>My experience is that formality creeps in primarily as the process 
>>>is abused.  The more it is abused the more necessary it is to make 
>>>rules about things where it could be assumed that good and fair 
>>>behavior would prevail.  As the stakes increase, that becomes less 
>>>the case. The only way for it not to happen is to work on things 
>>>that few people care about!  Either because they don't know it is 
>>>important or because it isn't!
>>>
>>>Take care,
>>>John
>>
>>From my research into the early institution-building of Internet 
>>standards (ICCB, IAB, IETF), it seems that another reason for 
>>building in formality is to allow open participation. As 
>>participation
>
>I don't see how formality allows open participation.  Generally, the 
>minimal number of rules is best and then only to ensure fair 
>participation.  My experience has been that the number of rules and 
>the formality of the process increases either when fairness is 
>abused or one group attempts to maintain control of the process. 
>The first is a case where without written rules some try to use the 
>fact that "there is no rule that says I can't" to abuse a fair and 
>reasonable process.

In this case I would think that the creation of rules is a healthy 
development for preserving a fair and reasonable process.

>Initially there were very few if any written rules.  My 
>understanding is that the rules in place came about when it became 
>clear that the IETF/IESG/IAB etc.  needed to be able to ensure that 
>a process was followed that would not subject it to law suits or 
>claims of anti-trust behavior.  The process was and always has been 
>about as open as you can get without formality.  Frankly, I think 
>the current process is so open that it provides the perfect disguise 
>for manipulation by anyone with the resources to play the game.

More formal structures allow a body to move away from a small 
"council of elders," thus making leadership somewhat accountable to 
the participants. Comparatively, the IETF has a minimum of rules and 
institutional structures; but those rules keep the process 
consistent, and therefore make the IETF more reliable as an 
institution. Without formal rules, a small group of people could make 
decisions that don't have broad support (such as was alleged of the 
IAB with CLNP in 1992). It's a classic constitution-building question 
- how much authority and structure is needed to preserve freedom? Is 
it too easy to manipulate the system?

Cheers,

Andy