Efficient board meetings, revised

[Ian Jackson]

David Graham - SPI Secretary writes ("Re: Efficient board meetings, revised"):
> I do not [like] the rules from 2004-10-15.iwj.1
...
> IRC discussions are an effective, efficient way to conduct meetings

This is, I think, the key point of disagreement between us.  I've
described my problems with IRC more fully in another message, so I
shan't repeat myself here.

> IRC logs, especially when read against a meeting's agenda are easy to
> parse into the discussions that took place, and provide an accurate and
> candidate representation of the views of those present.

I disagree with this.  Indeed, we've had discussions even during the
meeting about what people were referring to when they said something
and even in some cases about how they voted.

> Asking a question for clarification in a report is no reason to defer it
> for a month, nor is it reason to force closure and vote immediately.

Let me present a scenario, which is quite conservative in counting the
cost of IRC discssion.  Firstly, supposing the person with the
question asks the report author by email in advance:

Let us say that enquirer spends perhaps 1 minute writing that email,
and the report author (let us say) 3 minutes at their convenience to
read it and write a response (we'll assume that the question is
simple, the answer is known, and a 200-odd-word answer will do).  This
pair of mails will take the other interested parties perhaps a total
of 20 seconds each to read; let's say the response might take the
enquirer a full minute to read.  If we have 30 interested parties,
that works out as follows:
    1 minute of the enquirer's time to write the question
    3 minutes of the report author's time to write the answer
    1 minute of the enquirer's time to read the answer
   10 minutes (20s x 30 interested parties) others' reading time
  ---
   15 minutes total

If we do the same thing by IRC, then if we conduct the meeting in
full lockstep with everyone taking turns to speak and waiting for the
enquirer to read and ack the answer, it'll take 5 minutes of elapsed
time.  If the number of interested parties is the same, that's:
    5 minutes elapsed time
 * 30 interested parties
  ---
  150 minutes total

If the question or response are more complicated, or more discussion
is needed, or the report author needs to look up the answer, the
figures can easily get much, much worse.

Now, you may protest that the number of people on the mailing list is
much larger than the number of people who attend IRC meetings.  BUT
this is just a reflection of the comparitive inaccessibility of IRC
meetings !  People react to the waste of time by failing to
participate.

You may also say that we don't do things in lockstep like that.
Indeed, we don't.  We try to overlap the various activities somewhat.
This means that the enquirer may well be too busy reading and
considering the answer to pay proper attention to the next item, or
that the report's author may be too busy preparing their answer to
read the minor exchange that happens in the middle (which they maybe
had a comment on), etc.  With a synchronous low-bandwidth medium like
IRC, the only way to make sure people have enough time is to do it in
lockstep.

So, this is just a collosal waste of everybody's time, and totally
unnecessary.  After all, if someone wants to tell the board something
they can just email it to us and we don't need to formally accept it
or anything at all.  If they want to know that we've read it and are
happy for whatever it is to exist / carry on / be true / whatever,
then having the board members formally indicate their acceptance is
fine but there is _no_ reason to have everyone waiting around for it.

> > 3.  IRC discussions are less accessible to people who are not available
> >     at the time of the meeting than email.
> 
> Logs are distributed. Meetings are scheduled around when board members can
> attend. Some resolution discussion takes place privately on the board
> mailing list and is not even as open a process as the IRC meetings.

I think that we should usually discuss our board resolutions on
spi-general, or if applicable spi-private.  (I admit that I haven't
always been as careful about the To: headers of my messages as I
should.)

> provided the chairman of the meeting enforces procedural rules, such as
> not allowing digression, a far bigger enemy than topical discussion.

Have you asked the meeting chair by /msg to ask people to say on
topic ?  I agree that the digression is harmful.

> Resolutions are often posted days or weeks in advance and no replies are
> received until 3 hours before the meeting.

I agree that this is a problem.

> If you are not part of the solution, you are part of the problem.

I'd appreciate it if you'd respond to what I'm saying in good faith.
I agree that one of the problems has been the lack of timely
between-meeting participation and that I've been one of the culprits.
I'm trying to help fix that, by setting an explicit expectation for me
and other board members to meet.  I think that's more constructive
than finger-pointing or shouting at people on IRC.

> A 7 day minimum restriction means some issues, even if urgent, may go 5
> weeks without being voted on.

I think 5 weeks is fine if they're not urgent.  If they're urgent then
we can deal with them at the next meeting or by email; I explicitly
say that urgent things can be dealt with speedily.

> All items can and should be put on the agenda. The board has the right to
> defer any issue for further consideration.

The board can always decide to consider something immediately.
Perhaps I should explicitly mention that, although I thought it was
obvious from the discretion my resolution gives to the treasurer.

> The only thing stopping inter-meeting discussion of resolutions is board
> members not participating, and this resolution won't change that. People
> have to choose to participate in order to do so.

I think explicitly stating that they are supposed to participate, and
explicitly ensuring that the `it was only posted 48 hours in advance'
isn't an excuse because it's not true, and giving a period during
which their attention may be required, will probably help.

> > 14. The above rules may in any case be deviated from in case of
> >     necessity, such as external deadlines, emergencies etc.
> 
> These rules apply as long as they apply and if they don't apply well they
> won't apply because we don't need them to apply unless they apply, but if
> they do apply they do apply provided there are no reasons for them not to
> apply.

Don't be daft.  These rules are guidance to the secretary and
president (or whoever is writing the agenda or presiding over the
meeting).  I hope we can rely on then to interpret what's written in
the way it's intended, and if not we can formally or informally
clarify it.

> The meeting chair is already responsible for enforcing decorum and
> administering the meeting in an efficient manner.

Indeed, but it seems that I disagree with the secretary and/or meeting
chair about what's efficient and I hope the board will agree with me.
If the board agrees with me I trust the secretary and meeting chair to
respect the wishes of the board, just as I will respect the wishes of
the board if the vote goes against me.

> > 15. Guests will not be asked to state their name during the meeting.
...
> The practice of having guests state their name to the channel has been
> abolished independant of this resolution in favour of the secretary
> receiving the list over the course of the meeting by private messages.

That practice was abolished by 2003-09-01.iwj.1, which this new
resolution explicitly supercedes.  Therefore anything that was in
2003-09-01.iwj.1 which we want to continue should be in the new
resolution.

> Following the suggestion of a member, future minutes will include both the
> name and the on-line handle of all present, in the interests of
> understanding logs.

That's very sensible, thanks.

Ian.