Efficient board meetings, revised

Ian Jackson ijackson at chiark.greenend.org.uk
Sat Oct 16 22:10:56 UTC 2004


Bruce Perens writes ("Re: Efficient board meetings, revised"):
> 1. I */HATE /*IRC. Most folks type so slowly, it's like pushing my head 
> slowly through gelatine to get anything done. Actually, the SPI board is 
> better than most in this regard. IRC with mere mortals rather than 80 
> WPM-typing hackers is pure torture.

This is my problem with IRC board meetings for anything but the formal
confirmation of decisions etc.  For serious discussion, where it
actually matters, I think IRC is very poor.  The problems stem from
the real-time, text-mediated nature of the medium.

I'd guess that most of us can read (with an ordinary level of
attention to detail) at perhaps 500 WPM and type at maybe 50 WPM, very
roughly.  That means that we can read perhaps somewhere around 5..20
times faster than we can type, even if we don't have to think about
what we're saying or do any research.

This means that if we're having a conversation in an IRC meeting, and
we're mainly waiting for one person to speak, all of us are reading
about 10% of the time; 90% of the time is spent just waiting.

This leads to a number of `workarounds':

Sometimes people try to do something else in the resulting small gaps.
This makes the problem better for them, but of course worse for
everyone else as their small lapses of attention can be noticed and
cause delays etc.

Often, people abbreviate what they say, or hurry to try to get it out
faster.  This makes their utterances harder to follow, sometimes
ambiguous, and less well thought out.

Sometimes we have several conversations at once, on different topics.
This makes the discussion confusing; it also means that people who are
reading more closely, following references, and/or perhaps having
difficulty with the somewhat clipped English (see above) will lag
behind and either lose track of the conversation or have to explicitly
call for a pause (which is socially awkward).

A problem related to those is that, in an IRC conversation, everyone
has to be participating at the same rate because it's fully
synchronous.  There is no easy facility for someone who wants to think
about a question more deeply, or do some background reading, to do
so.

To illustrate the difference, imagine if we were trying to resolve
this very disagreement in an IRC meeting.  How could we find the time
to type out long, thought-out arguments such as those we're seeing in
this thread, let alone find the time to sleep on them, consider the
best way to put them and all the relevant facts, etc. ?

Some of these problems exist in in-person meetings too, but they are
less severe.  Most serious in-person meetings of organisation boards
/are/ backed up by comprehensive documentation and writing up of
proposals in advance - including explicit rules about how much time
committee members are expected to have to consider the issues.  And,
when the meeting is actually taking place, the full range of in-person
cues (body language, mid-sentence pauses, ums, interruptions, catching
of eyes, etc.) is available to provide much higher-bandwidth output
from each individual than most of us can achieve via a keyboard.

I would like to see SPI's meetings operate more like the meat-life
meetings of serious organisations: we should do our preparing of
position papers, discussings of our arguments, and attempts to
persuade, drafting of resolutions, etc., offline.  Only the most
urgent or contentious issues should remain for undigested discussion
during the meeting - and even for contentious issues, ideally, the
meeting provides the venue for the formal settlement of the dispute
one way or the other, via a vote.

I also find IRC is fine for more informal settings.  If what you
want is to quickly get a high-interaction-rate discussion amongst a
few people, and exact questions of what was agreed and making sure
everyone is carried along by the discussion are less important, then
IRC can be a very good alternative to phone calls.  So I would be very
happy (for example) to schedule a more interactive flame-fest^W
discussion with David Graham and others, so that we can try to
understand each other better and maybe shortcut some of these
arguments.  But, I think that for formal decisionmaking IRC should be
used not because it is a good medium for discussion, but because it
(a) meets the requirements that we have a _meeting_ as opposed to an
exchange of memos and (b) has some value in being able to chivvy
people to turn up and/or chastise them if they come unprepared or
without sufficient participation in advance.

Ian.




More information about the Spi-general mailing list