TC Membership Rules - Take 2

["Hal Lockhart" <hlockhar@bea.com>]

I raised this issue before, but the thread went off in other directions and
I did not have the cycles to pursue the matter. I was also waiting until the
new TC process was published to see which points had been clarified.

Let me start by disposing with a couple of points raised last time.

1. Some argued that the process should allow for chair's discretion. I
agree, however this should only be necessary in rare, borderline cases. This
note is about bread and butter decisions that come up every week, not once
or twice a year. If chairs are forced to constantly make these decisions on
their own, it will mean different TCs use different procedures and create an
impression of arbitrary use of power by chairs. Neither of these is
desirable for an open standards organization. The secretaries and chairs
need to know at the begining of every meeting whether or not they have
quorum. It should be possible to determine this by a well defined algorithm
95% of the time.

2. Some have suggested that attendance calculations can always begin with a
meeting that was attended, because that will always be true of a member.
This is only true of people who join at the initial meeting. Under the old
rules people could join the TC without ever attending. Under the new rules,
they cannot join without attending, but we are now required to enforce the
same rules during the probationary period, so the same condition (never
attending) can and does occur.

I suggest that the following points be clarified and that the "official"
interpretation of the process be made a part of the TC guidelines. The two
relevant sections of the TC Process are:

----------------------------------------------------------------------------
---------------
2. d.

During the probationary period the Prospective Member must follow the same
attendance rules as a TC Member, as defined in the section "Termination of
TC Membership"; if attendance requirements are not met the Prospective
Member will not become a Member at the end of the probationary period and
must re-apply for membership.

2. f.

* A member shall be warned by mail from the chair of the TC upon his first
failure to be present during the roll call at the beginning of the meeting
of two out of every three successive meetings of the TC. Membership shall be
terminated if the member fails to be present during the roll call of the
next meeting following transmittal of the warning or if the member
consistently fails to be present during the roll call of two of every three
meetings thereafter.
----------------------------------------------------------------------------
---------------

These leave a lot of questions open.

A) "A member shall be warned by mail from the chair of the TC upon his first
failure to be present..."

The choice of the word "first" seems deliberate, but ambigious. Does a
person ever have to be notified a second time? I am not talking about the
individual that misses one or two meetings a month. We can get rid of them
with the "consistently fails" clause. I am talking about a member who misses
two meetings, gets warned, attends every meeting for six months and then
misses two in a row. Do I have to warn them? Should I drop them immediately
without warning? Are they entitled to a warning per month? per year? per TC
lifetime?

B) Now let's consider the actual calculation. Warning is required in two
cases: (P=Present, N=Not Present) N-N and N-P-N. The first half of the rule
says if they miss the next meeting, they are out. But what if they attend?
We have N-N-P and P-N-P (considering only the window of the last three
meetings. In neither case do we warn them, because even though N-N-P is 2
out of 3, it is not the "first failure."

Again, if they attend the next meeting, there is no problem, so let's assume
they don't. We have: N-P-N and N-P-N. Since they both missed 2 out of three
again, it seems like I should throw them out, but I don't see how I can
under the rules given. They did not miss a meeting immediately after being
warned and I would not consider that a total history of ...P-P-P-P-N-N-P-N
or ...P-P-P-P-P-N-P-N meets the criteria of "consistently fails". So I guess
they stay in limbo for an unspecified period until a pattern develops.

The other thing that bothers me is that P-N-P-N-N gets you thrown out, but
P-N-N-P-N does not. This seems arbitrary and unfair.

C) During probation, we are supposed to "follow the same rules." I presume
that means that if they do N-N or N-P-N then we warn them that they may not
become become members. If they miss the next meeting, we immediately drop
them back to observers and do not continue calculating their attendance
pattern. If they attend, then we fall into the "consistently fails" case,
except that it is unlikely that this condition will be satisfied before
their probation is over. Presumably the intent is that the probationary
period and membership period be considered a continous sequence for the
purposes of calculating membership elibility. In other words, we do not
"reset the count" when someone becomes a member.

However, if the TC meets only monthly, then the person will become a voting
member before they can possibly fail the attendance criterion. In this case,
it is even more important to consider all meetings as a continous sequence.
Does everyone agree with this interpretation?

D) I don't want to start a thread on Kavi bugs, but the current Kavi
calculations about membership don't seem to follow any consistent algorithm
as far as I can tell. Ideally, Kavi should follow the algorithm determined
by the outcome of this thread. If this is not possible, Kavi should NOT send
out any messages about membership transitions, because they just create
confusion.

Hal