[db-wg] Where are we with NWIs?

Hi Nick

I am not interested in google searches for rules of parliamentary
committees. I am interested in RIPE specific documented rules. The
parliamentary committee situations you referred to are generally
formal, in person meetings where a single person has been appointed as
the president or chair. In those meetings people actually speak. The
chair does not have to go to extraordinary lengths to solicit
conversation, in the hope that the same, very small, group of people
will utter some words. It is completely out of context for the
situation on the DB-WG mailing list.

Your reference to the "Working Group Specific" section in ripe-692,
you have also taken out of context. That section refers to the
expectations of a chair in a RIPE Meeting. Then you say "By omission,
it clarifies that engaging in debate is not a RIPE-specific addition
to the normal list of duties of a chair, and by default, generally
accepted chair principles will apply." This is an extraordinary claim
even for you. The ripe-692 document is not an addition to some random
document found on the internet. 'Applying generally accepted
principles', that is an interesting comment from you Nick. The RIPE
Database Requirements Task Force, of which you were a member,
completely ignored all the generally accepted principles of business
analysis. You attempted to produce a hybrid business requirements and
technical requirements document for the database without defining the
purposes or identifying the major stakeholders and their needs. One of
the consequences of that is the mess we have had recently with geofeed
not being defined as a purpose.

You try to claim that in ripe-692 there is no specific reference to a
chair engaging in debate. In fact ripe-692 actually does have this
specific reference in section 2. Working Group Chairs and Co-Chairs.

"When participating in RIPE discussions, WG Chairs and co-chairs
should endeavour to make it clear whether they speak on behalf of
themselves, the organisations they work for, or the WG for which they
are co-chair."

You cannot get more specific than this. WG chairs CAN speak for
themselves in a discussion. This is not by omission, or by default, it
is clearly stated. I have frequently endeavoured to make it clear that
I speak personally as a former database engineer and analyst at the
RIPE NCC with a specific deep internal knowledge of how the database
works, operates and was designed.

You have ignored so many key points in your attempt to discredit me.
Firstly, unlike parliamentary committees, we have multiple, equal
ranking co-chairs. This allows for a separation of duties between the
co-chairs. William and I have always accepted that I will engage with
the community in discussions and William will determine consensus. So
I became the spokesperson trying to motivate people to discuss issues
and William sat in the background observing. This has worked quite
well and over the last 6 years we have managed to push many issues
over the finish line. As we have frequently endeavoured to make this
clear to the working group we have followed the specific rules written
into ripe-692. Now we have 3 co-chairs, I can continue to drive
discussion and we have 2 co-chairs to determine consensus. You even
acknowledged this option from your parliamentary documents "> Some of
these docs indicate ways of handling situations where a chair wants to
dive in, for example recusal from chair duty". This is exactly what we
do. When I engage in a discussion I don't take part in declaring
consensus.

You are also ignoring the clear fact that it is very hard to get
anyone to discuss many of the issues about the RIPE Database. The
subscribers to this mailing list are mostly observers. They follow
events but don't get involved. There is nothing wrong with being an
observer and keeping up with developments. But that leaves a very very
small core group of people who sometimes discuss issues. If we get a
handful of people in a discussion that is a rare success. As I pointed
out in my opening statement to this thread, quite often there is NO
discussion until I push it. You suggested in your previous email that
there are long standing, tried and tested, proven, well documented
ways of managing a working group mailing list and we absolutely must
not dare to change anything. Well it is clear that these old fashioned
ways of managing discussions frequently fail in this working group.
The old fashioned ways require a chair to sit back and say nothing,
community members discuss an issue, chair declares consensus or not.
Consensus absolutely requires discussion. NO discussion = NO
consensus!!! FACT!!! The community wall of silence is the killer of
consensus. It does not matter what the reasons are for people
remaining silent and just observing. That is what most people do.
FACT!!! Look back at the archives for proof if you need to. Having one
co-chair actively involved in discussions has proved to be effective
in getting enough discussion, sometimes, to achieve a consensus. Even
then some issues prove very hard to get community involvement. NWI-4
assigning an allocation was created 7 years ago based on an action
point from RIPE 70 in 2015. We still don't have any consensus on how
to do this. Getting discussions and consensus from this community is a
constant struggle for the chairs. Every option needs to be on the
table.

I am an 'active' co-chair. I strongly resent the implication from both
you and Randy that I am an 'activist' co-chair. An activist is "a
person who uses or supports strong actions (such as public protests)
to help make changes in politics or society". Completely inappropriate
language to use in this context.

You also referred to me "taking a specific point of view". Yes,
sometimes I do this for a number of reasons. One example was in the
discussion about the "country:" attribute. The discussion was slow and
in-precise. I put forward the idea of introducing a new "geo-country:"
attribute with a clear geolocation definition to replace "country:"
with. This was one of my crazy ideas that I knew would never happen.
But it had the effect of focussing people's minds to agree that
"geofeed:" is the way to go with the eventual deprecation of the
"country:" attribute. As Gert said, everyone is free to reject any
idea I put forward and agree on something else. That is exactly what
people did in this case and that was a result. In other situations I
may use my deep internal knowledge of the engineering design of the
database to suggest a certain path over another. For example
introducing a new status value 'ALLOCATED-ASSIGNED' rather than the
combined primary key 'inetnum:status'. Over 7 years there has been
short bursts of discussion on this. Currently the feeling is to reject
my suggestion of the status value in favour of the combined primary
key. When I make suggestions people are free to accept them, debate
them or reject them. I am not an 'activist' hitting people over the
head with a stick to force them to agree with me. But if we go down
the path of a combined primary key we cannot accept consensus based on
the usual handful of people who control or influence most database
decisions. This is a fundamental change to the query interface of the
database that will impact millions of queries per day by hundreds of
thousands of users. This will need much wider consultation including
opinions from many of the bigger users of the database. Sometimes by
using my internal knowledge of the database I can advise the community
to take a simpler path to achieve an acceptable (maybe compromise)
result. The alternative is often to allow the community to take
another path, wait for the impact analysis, realise the impact of the
change is too high, then start again on a simpler path. It takes so
much time and effort to get the community to discuss an issue, this
process may add another year to resolving the issue.

I accept that there are old fashioned, tried and tested, well
documented ways of doing things. But society changes over time. The
co-chairs must also evolve their methods of working to handle the
situation they are faced with today.

cheers
denis
co-chair DB-WG




On Fri, 21 Jul 2023 at 11:39, Nick Hilliard <nick at foobar.org> wrote:
>
> denis walker wrote on 21/07/2023 06:22:
> > On Thu, 20 Jul 2023 at 16:41, Nick Hilliard via db-wg <db-wg at ripe.net> wrote:
> >> the job of a chair is to ensure that the business of the forum is done,
> >> in an orderly and efficient way.
> >
> > Where is this stated?
>
> Denis,
>
> It's stated in pretty much every formal description of a chair's duties
> ever written. A good reference point would be Robert's Rules:
>
> http://www.rulesonline.com/rror-10.htm
>
> See the paragraph starting "His duties are generally as follows:".
>
> >>   From this point of view, an "active"
> >> chair is good, and to be welcomed.  What isn't ok is when a chair
> >> decides to take a specific point of view and actively engages on an
> >> issue, on one side or another.
> >
> > Where is it stated that a chair can't take a specific point of view
> > during a discussion? Maybe a chair takes a point of view in order to
> > elicit further details from those involved in a discussion on that
> > point.
>
> An online search for "committee chair duties" or "what are the roles and
> responsibilities of a chairperson" will provide plenty of references.
> None of them include getting involved in a discussion, and many of them
> are quite explicit that the chair should not to get involved. RR's are
> diplomatic but firm about this:
>
> > The chairman of a committee usually has the most to say in reference to questions before the committee; but the chairman of an ordinary deliberative assembly, especially a large one, should, of all the members, have the least to say upon the merits of pending questions.
>
> (http://www.rulesonline.com/rror-10.htm)
>
> Or Cushing's Manual:
>
> > "It is a general rule in all deliberative assemblies, that the presiding officer shall not participate in the debate or other proceedings, in any other capacity than as such officer. He is only allowed, therefore, to state matters of fact within his knowledge; to inform the assembly on points of order or the course of proceeding when called upon for that purpose, or when he finds it necessary to do so; and, on appeals from his decision on questions of order, to address the assembly in debate.
>
> (https://www.gutenberg.org/files/60757/60757-h/60757-h.htm, section 202)
>
> There are plenty of other references too, and it's not relevant that
> RR's / Cushing's Manual refer to parliament / large assemblies - the
> principle is generally accepted to apply to chairs in general,
> regardless of the context of the forum they serve.
>
> Some of these docs indicate ways of handling situations where a chair
> wants to dive in, for example recusal from chair duty, or resignation in
> situations where that would be more appropriate.
>
> To use current parlance, "active chair": good, "activist chair": not so
> much.
>
> > There are a couple of documents on chairs responsibilities:
> > https://www.ripe.net/publications/docs/ripe-692
> > https://www.ripe.net/publications/docs/ripe-765
> > https://www.ripe.net/publications/docs/ripe-764
> >
> > Please tell me which part of any of these docs I am not complying with.
>
> The "Working Group Specific" section in ripe-692 lists the duties of the
> chair.  By omission, it clarifies that engaging in debate is not a
> RIPE-specific addition to the normal list of duties of a chair, and by
> default, generally accepted chair principles will apply. I don't
> especially see a good reason to update the WG Chair documents to make
> this more explicitly, but if it's not abundantly clear already, maybe
> something should be put in there to acknowledge it?
>
> Nick