[rfc-i] Substantial revision

[Joe Touch]

On 5/29/2012 10:04 AM, Joe Hildebrand wrote:
> On 5/29/12 10:43 AM, "Joe Touch"<touch at isi.edu>  wrote:
>
>> Change control can be useful for IDs (using the author's source, which
>> might support revision), but very few RFCs are ever "bis'd",
>
> Really?  That's very surprising to me.  Do you have any stats to back that
> up?  I would have guessed that something like 20% of them eventually get
> bis'd, which is significant.

There are 116 current -bis IDs.

There are 107 current IDs based on the Word template.

Both out of approx. 2600 current IDs.

Counting -bis from RFCs is a lot harder, since the term 'bis' is dropped.

>> and benefit from a clean-slate revision more than mere incremental editing.
>
> Perhaps those folks need to release their first version more quickly, then.

IMO, an RFC that is *intended* to be useful for less than 5 years is a 
waste of time. Better that we have fewer RFCs that have the long view.

>> I.e.,
>> revision support may be useful for author source, but isn't relevant for
>> either the submission or output formats.
>
> That's great if the author and the author's input is still around.  One
> benefit of the submitted format is likely to be that it is archived in a
> stable location.

I was considering two cases:

	- during the ID process, when revisions are frequent
		author source should definitely be available then

	- when an RFC is revised
		by that point, a clean sheet ought to be both
		preferred and encouraged; the utility of the original
		source in that process is very low AFAICT

Joe