Re: [802SEC] reflector observations
Bill -
I couldn't agree more. As I pointed out a while back, the first thing we
need, rather than the rules themselves, is a set of "guiding principles"
from which we can derive any rules we deem necessary, and against which the
validity of a rule can be checked if we later find it to be questionable.
Regards,
Tony
At 10:05 03/07/2003 -0700, Bill Quackenbush wrote:
>All,
>
>I disagree only slightly with Bob. I would delete his first sentence.
>I agree with the rest of what he has to say.
>
>There is nothing wrong with the SEC writing the rules. The problem is
>the way we go about it. And in particular, the habit of
>confusing/substituting "wordsmithing" for making policy decisions. The
>generation of the rules text needs to follow the policy decisions, not
>what policy can we discover in/ infer from text that has been hacked
>together through multiple acts of wordsmithing. If we wrote standards
>the way we write our rules, we would have long ago been out of business.
>
>Thanks,
>
>wlq
>
>"Grow, Bob" wrote:
> >
> > Matt:
> >
> > The major problem with our rules and our rules change process is that
> we (the SEC) are writing them. We have been maintaining them for some
> time through a cumbersome process where the focus is on the specific
> words rather than what a rules change is trying to accomplish. There
> have been so many authors that there isn't a consistent style which is
> important for understanding the intent of the rules.
> >
> > I have expressed before that I believe we would be better off with a
> total rewrite of our rules. A single author would be able to reorganize
> them into a better format (Bill made a proposal on this), that author
> would use the current rules as the basis. A third party author would not
> be inventing new rules for us, but simply capturing the intent of the SEC
> and producing the language that we would still vote to accept. If our
> SEC discussions were focused more on what we are trying to accomplish
> rather than the specific words, I believe rules changes would be
> accomplished much more efficiently.
> >
> > A total rewrite is I believe the best way to fix the inconsistencies
> with superior rules, inconsistent naming and abbreviation, etc.
> >
> > --Bob Grow
> >
> > -----Original Message-----
> > From: mjsherman@research.att.com [mailto:mjsherman@research.att.com]
> > Sent: Wednesday, July 02, 2003 6:04 AM
> > To: millardo@dominetsystems.com; stds-802-sec@ieee.org
> > Subject: RE: [802SEC] reflector observations
> >
> > Howard,
> >
> > I have mixed feelings about this. First the good part - you'll put me
> > out of a job. Probably a good thing since I'll simply focus my efforts
> > on other EC issues. Next the bad part - we develop our rules by
> > consensus. I'm not sure we want someone to write the rules for us,
> > which sounds kind of like what you are proposing. I wouldn't mind
> > having someone else to run the ballots and collect the feedback. But
> > I'm not sure we should turn over the process of writing the rules to an
> > outside party.
> >
> > Another question for you is - Are you proposing changing the current
> > process we use to develop rules? In my mind, the rules need a complete
> > rewrite. Again, I'm not sure we would want to turn that over to an
> > outside party. Yet it is too big a task for any one of us to take on
> > all at once, perhaps even with outside help.
> >
> > I think a large part of the problem is that we have become more aware of
> > the short comings in our rules, and of the rules in general. On one
> > recent P&P revision (the one to use Executive Committee rather than
> > Sponsor Executive Committee in the rules) someone abstained citing as
> > one of their reasons the fact that the use of acronyms vs spelling the
> > acronyms out was inconsistent throughout the document for many acronyms.
> > I'm not sure how critical all these issues are, but in the past I think
> > we may have ignored these more minor issues rather than fixed them.
> >
> > I try to initiate 2 rules changes a cycle (I'm running 3 right now).
> > Other EC members run other P&P revisions as well. This cycle we
> > promised 2 revisions by other EC members but only delivered one (for a
> > total of 4). If we actually manage to complete an average of 3
> > revisions per cycle, that is a total of 12 revisions per year. This
> > means our rules are in an almost constant state of change. Note that I
> > maintain a list of pending P&P revisions that individuals have noted
> > they would like to make. The list grows at a rate which is faster than
> > I take things off. As a result, I currently have a backlog of about 12
> > P&P revisions that need to be run. At our current rate of progress it
> > will take us about 1.5 years to get through them all. But as I noted,
> > the list grows fast than it shrinks, so I can't say honestly we will
> > ever get through them all. Also, I recently started to note conflicts
> > in rules above and below our P&P, that need to be resolved outside our
> > P&P that I'd like to see addressed. What I am really saying is that
> > there are degrees of perfection and clarity. Part of the problem is
> > that we seem to have moved to a state where we desire greater perfection
> > in our rules then we used to. All of these issues have been around for
> > years, but we are only now trying to address them. However, we may also
> > now be expecting too much perfection in our rules, and maybe we need to
> > back off a little.
> >
> > Just a few rambling thoughts to get the conversation going.
> >
> > Mat
> >
> > PS - Yes I could have removed you from the "To" line so you only got one
> > copy, and yes I could have gone down and deleted the rest of the e-mail
> > trail. But it is easier to just push respond to all and waste a few
> > kilobytes - especially with our reflectors new found performance.
> >
> > Matthew Sherman
> > Vice Chair, IEEE 802
> > Technology Consultant
> > Communications Technology Research
> > AT&T Labs - Shannon Laboratory
> > Room B255, Building 103
> > 180 Park Avenue
> > P.O. Box 971
> > Florham Park, NJ 07932-0971
> > Phone: +1 (973) 236-6925
> > Fax: +1 (973) 360-5877
> > EMAIL: mjsherman@att.com
> >
> > -----Original Message-----
> > From: Howard Frazier [mailto:millardo@dominetsystems.com]
> > Sent: Wednesday, July 02, 2003 12:45 AM
> > To: stds-802-sec@ieee.org
> > Subject: [802SEC] reflector observations
> >
> > Folks,
> >
> > A few observations:
> >
> > 1) We have exchanged more messages on the subject of
> > wireless projectors than we exchanged on the subject
> > of whether 802.11g was ready for RevCom. In fact,
> > we have exchanged more messages on this subject than
> > anything else we have discussed in recent memory.
> >
> > 2) Most of the messages on the subject have been sent
> > both to the SEC reflector, and to a list of individual
> > addresses. I am not sure of how many of you
> > are getting duplicates, but I have gotten every one
> > of the messages on this thread twice.
> >
> > 3) Most of the messages on the thread have quoted
> > the earlier messages. The most recent messages are
> > about 9 kibibytes in length, but add only a few hundred
> > bytes of new information.
> >
> > The point being that we don't really need to send
> > duplicate copies of redundant information.
> >
> > And everyone seems to have overlooked the fact
> > that the SEC reflector is operating MUCH FASTER
> > now. I notice that the duplicate copies are received
> > back to back, with one addressed to me individually
> > (thus bypassing the reflector entirely) and one coming
> > via the reflector, with less than 1 minute difference
> > in delivery time. Pretty good, eh?
> >
> > Since we seem to have so much time to devote to the
> > question of whether we should spend an extra $400 to
> > buy a projector with a wireless interface (vs one without),
> > may I ask that we spend a few cycles discussing the
> > suggestion I tossed out earlier, which was for us to
> > consider retaining a paid staffer to help with P&P
> > changes and project tracking?
> >
> > Howard Frazier
Regards,
Tony