Re: Straw-man charter for http-bis

[Julian Reschke <[e-mail]>]

Keith Moore wrote:
>>> My recommendation would be for the group to construct a list of errata
>>> and get consensus on that list.  Each erratum should mention the
>>> specific sections and text of RFC 2616 that it applies to, what the
>>> problem is, and what changes are needed to fix the problem.
>>> ...
>> Yes, that's what we have been (slowly) doing over the last months. See
>> <http://www.w3.org/Protocols/HTTP/1.1/rfc2616bis/issues/>.
> I understand, and this is a useful head start.  But you don't have a
> working group yet, and therefore you don't have working group consensus
> yet.  The item isn't completed until an open, chartered working group
> has actually had time to go over the document.   It's not acceptable to
> skip this step.

Right now things that are marked closed are in this state because the 
proposed resolutions got consensus on [e-mail]. I know that's 
not a WG consensus, but it's the closest thing to that that we currently 
have.

Of course, should an HTTP WG be formed, all these proposed resolutions 
can be challenged again (actually, that can happen even in the absence 
of a WG :-).

>>> My guess is that if the group sees its task as making a good
>>> errata-and-fix list for 2616,  it will stay focused and finish in a
>>> reasonable amount of time.  If at that point it is seen as appropriate
>>> to actually update 2616, this will be a straightforward task which won't
>>> take a lot of additional time.  (I do not propose that this task be
>>> delegated to the RFC editor - the RFC editor function needs to stay
>>> separate.)
>> I personally think that this should be a by-product of collecting and
>> resolving the errata.
> I disagree, because I think the temptation to needlessly edit text will
> be too large.

I guess that depends on who's doing the editing. Speaking for myself, I 
have tried to make sure that changes are kept as minimal as possible, 
and that each and every change is linked to the issue it's supposed to 
resolve.

Best regards, Julian