Re: [Gendispatch] I-D Action: draft-thomson-gendispatch-no-expiry-00.txt

[Keith Moore <[e-mail]>]

On 11/6/19 1:32 AM, Martin Thomson wrote:

> On Wed, Nov 6, 2019, at 17:29, Keith Moore wrote:
>> I like the expiry date also.   At the very least it helps people see, to
>> a rough approximation, which internet-drafts are being maintained and
>> which are not.
>>
>> I also think that if one is searching for a particular (current) I-D,
>> the default search results shouldn't be flooded with drafts that are
>> arbitrarily old.
>>
>> How is the date of publication not sufficient for both of those use cases?

Maybe I don't understand the question.   As I see it, since we stopped 
actually making drafts inaccessible after six months, six months became 
just a convention for whether a draft is being actively maintained and 
for whether it's reported in default search results.   I think it's a 
useful convention for those purposes, but one could certainly implement 
it without actually having (for example) separate repositories.

On the other hand, if the only change you're proposing is to remove one 
line from the I-D header (and associated metadata) and change the 
boilerplate, I frankly don't think it's worth the expenditure of energy 
on the part of IETF, IESG, RFC editor, tools team, etc., that would be 
required to do that.   There doesn't seem to be an obvious benefit to 
justify doing that work, and it sort of looks like change for the sake 
of change.  (otherwise known as "action bias").   The -00 I-D also 
leaves unstated whether there's any significance at all to the six month 
figure and that seems like an undesirable ambiguity.

Keith