[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: limits (Was: IETF ldapbis WG Last Call: draft-ietf-ldapbis-iana-04.txt)

At 12:17 PM 2001-11-30, Ryan Moats wrote:
>On Fri, Nov 30, 2001 at 11:42:36AM -0800, Kurt D. Zeilenga wrote:
>| At 10:07 AM 2001-11-30, Ryan Moats wrote:
>| >Because I've seen far too many cases where what was originally thought
>| >to be "enough space" ran out.
>| 
>| But these aren't limits.  They don't cause us to "run out" of anything.
>
>I think you are trying to have it both ways.  You are saying "anything
>longer MAY be refused as being too long" and then saying "these aren't
>limits".  They *are* a barrier to names (call it a guideline, call it
>a limit, it doesn't matter).

There is a huge difference between:
        Descriptors MUST be no longer than 48 characters.
than saying
        Registrations of descriptors longer than 48 characters
        MAY be refused.

The former is a limit or barrier.  The latter is guidance.

While we could strike the whole consideration and leave it
up to IANA to determine what "too large" to register, I think
that BCP 26 calls for us to provide sufficient guidance.

I've tried to make these choices in a manner which gives
requesters, reviewers, and the register wiggle room.  That's
why it's a MAY, not a MUST or a SHOULD.

It should be obvious that if the requester tries to register
something which "too long" per the guideline AND there are
no reasonable alternatives, that the register would exercise
appropriate judgement and not refuse the registration.

> What's so special of 16 over 32?

Basically, I choose 16 as, IMO, 8 seemed is too small and
32 is excessive [taking everything into consideration].

But, with this said, 16 isn't "so" special.  I could likely
live with 16 or 24 or 32.  I suggest we have Bob do a straw
poll at IETF#52 and take it from there.