[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Proposal to move all IETF Mail Lists to ietf.org



Dean Willis wrote:

On Jun 8, 2008, at 6:33 AM, David B. Nelson wrote:

Dean Willis writes...

However, I have seen cases recently where people need to establish
whether a particular John Doe was or was not subscribed to a given
IETF list on a given date (sometimes several years ago).

The "What did you know, and when did you know it?" line of inquiry?

And the "Did you have a duty to disclose the IPR held by your company that affected the standard being discussed on the mailing list that you were subscribed to?" line of inquiry.

Remind Where in the Note Well or bcp 79 it requires you to disclose IPR if you subscribe to a mailing list?

If you do not contribute to a document or participate in the discussion. then you do not have a disclosure requirement.

Knowing who is on the mailing list is irrelevant in that context as is for example knowing who reads the mailing list by looking in the archives.

This is more than the information that a subscribed-list archive
(as above) would have...

Do the current lists at ieft.org archive all of the add and remove request
messages, in addition of the content messages?

hard to say. I've asked at times if such information was available, and have received a variety of responses. So far, I've been able to find "participatory examples" when needed, so those message archives are very, very important!


... and it's the kind of thing IETF might be subpoenaed for.

Has that sort of thing ever happened to the IETF? You can't produce records
you don't maintain.  Unless you can quote some statutory requirement that
the IETF keep such records, it would be at our discretion.

Well, we might not be able to produce records, but we might have to spend some effort convincing the judge that we can't produce those records. This is easier if we have 1) a published policy on what is maintained, and 2) actually comply with that policy. My opinion is that it's generally to the IETF's benefit to have these records, but I'm sure that's not a universal truth.

Overall, this suggestion sounds reasonable in terms of neat organizational
housekeeping.  My question is who does the work to seamlessly migrate
well-functioning existing lists to the ietf.org servers?


As a WG chair, I've migrated several non-IETF lists to IETF in a fairly straightforward way:

1) Get the new list provisioned -- we have a process for this, and it works pretty well.
2) Warn the old list about cutover and freeze its membership
3) Add the subscribers from the old list to the new list (admin interface can do this)
4) Subscribe the new list to the old list
5) On cutover date, add an autoresponder to the old list to remind people who post to the wrong one
6) Change the list pointers on various web sites.
7) Wait a while, then kill off the old list.

I believe we were even successful in getting some of the old mail archives moved over for some of the lists, which is a nice touch.

From time to time, I've migrated a list without automatically moving the subscribers. This is a very lazy way to purge stale entries that don't happen to bounce, which happens more often than we probably think.

--
Dean