Meeting notes, 1/7/02

From: Tom Holub (tom@LS.Berkeley.EDU)
Date: Fri Jan 11 2002 - 13:20:36 PST

  • Next message: Hua-Pei Chen: "Re: Meeting notes, 1/7/02"

    (Next meeting, Monday, 1/14, at 11:00 AM in 261 Campbell)

    We began with a discussion of the CalNet decision to protect the privacy of
    the CalNet friendly name. Lucia wasn't at that meeting (and JC wasn't at
    ours), but her impression was that it won't be protected at the level of
    FERPA information, it just won't be published. When staff and faculty
    choose a friendly name, they'll be given a choice of all the names they
    currently own in the uclink/socrates name space, but they can choose a
    different name if they want to (if it's not taken). Students may not get
    a friendly CalNet ID. (Gordon) So we can't use the CalNet friendly name,
    if students won't have one. (Tom) We wouldn't be exclusively using the
    CalNet friendly name, it would just be one of the options. Students would
    probably pick their uclink name, or a new name.

    (Tim) What about the idea of using the left-hand-side of whatever address
    each person has listed in CalNet? Whether it's user@uclink or user@haas.
    That must be a preferred friendly name to some extent. We just need to
    figure out how to handle conflicts.

    (Jerry) What about namespace restrictions? The namespace isn't restricted
    to 8 characters, but what about periods, underscores, dashes, etc.? We
    should look into what is accepted in the namespace. We probably want to be
    as liberal as possible...what about Danielle_Hawn-Bluey (a real
    example--Tom)?

    At this point we were coming to something resembling a begrudging consensus.
    The proposal is:

    Someone goes to a web page to choose a Berkeley.EDU forwarding address.
    They are given a list of choices, including any names they own in the
    socrates/uclink/calnet namespace, plus the LHS of whatever they have listed
    as an address in CalNet. They can choose something else as long as it's
    not taken in the socrates/uclink/calnet namespace. (And if they do, it's
    reserved for all those other systems).

    We took a vote. Everyone's hands were raised, except mine, which was
    half-raised, and Janet's. Zane and Michael both agreed with me that
    the solution isn't what we'd prefer, but it seems necessary if we're to
    move forward. We appear to have made our first decision.

    We discussed briefly where this web page would exist. Certainly as part
    of the CalNet directory update, and possibly bilink too. Anywhere we can
    put a checkbox that says "also use this as my berkeley.edu address" seems
    good.

    The second topic of discussion was whether we make the system opt-in only,
    or whether we generate addresses. After the SurePay experience, people are
    wary of anything except opt-in. (We wouldn't be running our service on an
    unpatched IIS server with an unregistered 40-bit SSL certificate, but the
    point stands).

    (Tom) The availability of @University.EDU addresses is prevalent--if you
    look at the other UC's, only UCSC doesn't have at least some listed in their
    directory (they use @cats.ucsc.edu, a uclink analogue). UCLA has some @ucla,
    others with addresses like @mednet.ucla.edu and @geog.ucla.edu. The rest of
    the UC's all have primarily or solely @UCx.EDU addresses in their online
    directories. (Jerry) When I looked at uclink's mail logs, it didn't look
    like people at UC Riverside were using @ucr.edu addresses. [We fired up the
    computer projector to check it out] It looks like they're using
    auto-generated first.last@ucr.edu addresses in their directory, but maybe
    people are actually publishing different addresses.

    The tenor of the comments was aiming towards having the service be opt-in.
    We took another vote, this time with all but Tom and Janet raising their
    hands. [We're going to start a committee of our own.] But in terms of what
    this commitee is going to put it, it seems like we've now decided that
    the service should be opt-in for everyone.

    Having made those decisions, we wrapped up the meeting. Next week we'll
    begin talking about technical implementation. Our aim is to have a report
    ready for the February ITAC meeting.

    -- 
    Tom Holub (tom_holub@LS.Berkeley.EDU, 510-642-9069)
    College of Letters & Science
    249 Campbell Hall
    



    This archive was generated by hypermail 2b29 : Fri Jan 11 2002 - 13:20:39 PST