[ih] Cluster Addressing and CIDR

[Joe Touch]

Lloyd Wood wrote:
> On Tue, 14 Jan 2003, Joe Touch wrote:
> 
> 
>>Simone Molendini wrote:
>>
>>>>>So, IDs could be as valuable as RFCs.
>>>>>Then, why imposing a 6 month limit for IDs, when diskspace is so
>>>>>cheap?
>>>>
>>>>
>>>>The reason for ID disappearance has nothing to do with space.
>>>>
>>>>The IDs are deliberately ephemeral, intended to foster the open
>>>>exchange of partial ideas. Establishing them as archival from the
>>>>start imposes a hurdle that was percieved to inhibit this exchange.
>>>
> 
> Yes, but the process becomes increasingly formalised - originally RFCs
> were informal and barriers to publication were low. Now RFCs are
> formal and the draft process formalises the documents while rasising
> the barrier to publication as an RFC.
> 
> I've increasingly seen draft deltas go to a small off-WG audience
> before popping up when submitted as an 'official' draft just before a
> meet closing deadline; I'm not sure that this is good, since it limits
> on-list discussion before the meet.

I'm not clear that the two are related. There are few barriers to 
putting out an ID, except timing. Due to the manual processing 
requirements and the desire for a modicum of sanity checking, it's 
infeasible to handle the burst of submissions before each IETF meeting 
without including a buffer period.

This is the reason that many IDs submitted just before the deadline are 
concurrently posted to newsgroups; it allows open discussion while 
waiting for the burst to be processed. I agree that this isn't optimal - 
IMO, any idea that isn't sufficiently solid a few weeks before the IETF 
isn't sufficiently solid to warrant burning the $$ of people's time 
discussing it ;-)

> I'm worried that the process will even become more formal still, if:
> 
> http://ftp.ietf.org/iesg/iesg.2002-12-12
> IP     o Thomas to write (or cause to be written) a draft on "how to
>          get to Draft"

It's generally useful to help people design drafts in a way that reduces 
the repeated feedback from the RFC Editor. Most venues have some 
publication guidelines; it's not clear they're aiming at formalizing the 
process so much as providing editorial structure.

Joe