Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802SEC] New Model for IEEE Standards Maintenance

Geoff
While I agree with your points in theory, I notice that it is predicated with "supposed to have", "should be", and "should have".
I agree that this is what is supposed to happen. However, I think that the reality is that it is now what actually does happen.
I think that there should be some sort of a comment period to review the edition. If there are actual mistakes, then we might fast-track a corrigendum PAR based on the edition to enable a proper voting process.
Ideally, this is the exception and most editions simply go together without any problems. However, I am always worried about any protocol that only works if there are no mistakes. IMHO: It should work efficiently when there are no mistakes/problems, but there should always be an error recovery method. 

James Gilb

On 02/04/2011 04:56 PM, Geoff Thompson wrote:

Adrian-

I'm afraid that I have to take an opposing view on this.
My experience has been that the time to present these changes to
existing text is at the time when the amendment is initially put up for
approval. That is the point in time when the most eyes examine the draft
and that is exactly what is needed for this sort of exercise.

Kicking this portion of the problem down the road to be resolved at
revision time only sets things up for a number of problems:
1) inconsistencies/conflicts will produce WG overhead in errata and
requests for interpretation if it isn't done right the first time
2) Editing for revisions is work that is more difficult to get funded
than editing for a new project with the carrot of new products
3) Close scrutiny in voting for revisions is work that is more difficult
to get accomplished than the equivalent effort for a new project with
the carrot of new products.

It is a challenge to handle multiple edits to the same area when more
than one project is in Sponsor Ballot but I believe there are extensions
to editing tools that can display this and that the proper text
reconciliation work should be done as early as possible when the task
groups are still at the height of their participation.

Amendments are supposed to have clean enough editng instructions that
generating an edition is an exercise that only takes expertise in
operating editing tools. No expertise in the subject matter should be
required. No ambiguities should be encountered. Those should have been
seen and resolved during project balloting.

Having standing conflict in a published standard is at worst bait for a
lawsuit.

Best regards,
Geoff

On 2/2/11 10:32 PM, Stephens, Adrian P wrote:

Hello James,

I think I'd go further than that. IMHO, any complex amendment
is likely to contain errors that make a strict interpretation of its
editing instructions impossible. (e.g. I'm just about to roll in a
400-page amendment #8.)

When a revision is active, the editor can highlight these
inconsistencies/conflicts
either to the group responsible for making changes, or to the
balloters by comments
in the draft revision.

I'm not sure how this should be treated in the case of an Edition.
If the editor is aware of a conflict, should he highlight this, or
silently
ignore it?

This is separate from the question of whether the editor correctly
interpreted
the instructions in the amendment.


Best Regards,
Adrian P STEPHENS
Tel: +44 1954 204 609 (office)
Tel: +44 792 008 4900 (mobile)
Skype: adrian_stephens
----------------------------------------------
Intel Corporation (UK) Limited
Registered No. 1134945 (England)
Registered Office: Pipers Way, Swindon SN3 1RJ
VAT No: 860 2173 47

-----Original Message-----
From: owner-stds-802-sec@ieee.org [mailto:owner-stds-802-sec@ieee.org]
On Behalf Of James P. K. Gilb
Sent: Thursday, February 03, 2011 12:54 AM
To: Grow, Bob
Cc: tony@JEFFREE.CO.UK; STDS-802-SEC@LISTSERV.IEEE.ORG
Subject: Re: [802SEC] New Model for IEEE Standards Maintenance

All

The only issue I have with editions is that they should still have some
technical review to make sure that no mistakes were made.

Although we try hard, when you have 3 amendments modifying the same
location in a base standard, it may take some technical expertise to
make sure everything came out right.

It could be as simple as the Sponsor assigning a group of reviewers to
assist the editors. Then it could be a simple approval by the Sponsor
that based on the opinion of the reviewers, the edition is OK for
publication.

James Gilb

On 02/02/2011 09:01 AM, Grow, Bob wrote:

Tony:

I have been working with publication staff on publishing all
amendments, corrigenda and errata as editions. Publication staff has
been generally favorable to this, the only identified negative is
product pricing (why do I have to pay for all 3500 pages of 802.3
when I already have it and only want the new 50 pages).

If you or others are interested, I can provide details on what I've
proposed. I'm certainly willing to accept support, and constructive
criticism as well for any IEEE-SA participants.

Your proposal to do editions at least every three mitigates the
pricing issue because a revision presents the same situation.

--Bob


-----Original Message-----
From: ***** IEEE 802 Executive Committee List *****
[mailto:STDS-802-SEC@ieee.org] On Behalf Of Tony Jeffree
Sent: Wednesday, February 02, 2011 4:03 AM
To: STDS-802-SEC@LISTSERV.IEEE.ORG
Subject: [802SEC] New Model for IEEE Standards Maintenance

Reflecting on this new model, I have a couple of observations.

As I said on the conference call, I don't believe that the changes as
outlined are of any great benefit (or dis-benefit for that matter) to
802,
which seems to be a wasted opportunity when I believe that a simple
change
COULD be made that would actually be of benefit.

I have never understood the point of the 3-year revision rule -
apparently
it is OK to have a gozillion amendments approved in years 1-3 after a
revision, and all is OK for those 3 years, but suddenly at the end of
year
3, it is not-OK anymore. That makes no sense to me whatever, and will
make
even less sense once the revision cycle moves to 10 years.

What would make far more sense to me would be to lose the 3-year
revision
rule, and instead, impose a requirement to produce an Edition when
there are
N amendments (where N probably equals 3) that haven't previously been
incorporated into a revision or an edition. That would materially
improve my
situation in 802.1, as it would remove an arbitrary requirement to
revise
after 3 years when an editorial roll-up would be entirely sufficient
to the
needs both of the readership and the WG. Producing editions on a regular
basis is in any case something that I try to do with 802.1Q already.

Regards,
Tony

----------
This email is sent from the 802 Executive Committee email reflector.
This list is maintained by Listserv.

----------
This email is sent from the 802 Executive Committee email reflector.
This list is maintained by Listserv.



----------
This email is sent from the 802 Executive Committee email reflector.
This list is maintained by Listserv.

----------
This email is sent from the 802 Executive Committee email reflector.
This list is maintained by Listserv.





----------
This email is sent from the 802 Executive Committee email reflector.  This list is maintained by Listserv.