[Air-L] More on the change

[jeremy hunsinger]

>
>
> My worry is about the .01% (including an infamous example from a few  
> years
> ago) who might use such an episode as the occasion to attempt to  
> inflict
> real damage on the organization when discovering that, contrary to  
> Jeremy's
> assurances, we are unable to persuade / bribe Dreamhost into deleting
> problematic postings.

you don't do it that way.  as i noted, you send them a service request  
for the tar file.gz file of the archives of the list in question, you  
edit that file, and you ask them to replace that file.

however, you shouldn't just do that for anyone at any time.   it isn't  
a weekly thing, it is a... if there is a legal notice from a lawyer  
and it has been approved by our lawyer.   or... if there is a clear  
issue like 'oops i posted my ssn'.   takedown is not something we  
should do without significant cause.   I think it has been done once  
under my watch and it was something equivalent to 'i posted my ssn'.    
It was done at that time because that was acceptable under the  
virginia tech system.   AoIR as far as I know has no policy on  
takedown.  I've just more or less proposed one.  If we had a clear  
one, then i suspect this issue would be resolved.

In regards to the Lawyer, we have plenty on the list who might be  
willing to donate some consultation to a 503c non-profit.  Some things  
concerning the law... you need to consult a lawyer about.  Michael  
Zimmer for instance pointed out that we might be exempt from certain  
legal actions like the ones the exec seems to be worried about.  I'm  
sure you'd like a second opinion on that.   On the other hand...  I  
don't think you should worry about phantoms.  If you have a real  
threat, please do deal with it.   If this is only a hypothetical and  
protecting from a hypothetical, I have to think that the likelihood is  
going to be pretty small that this hypothetical could not be avoided  
without the action taken.   Acting on hypotheticals is hugely  
problematic... especially when the action taken seems to be contrary  
to every other best practice for list management.   I'm on 100+  
academic/professional lists, this is the only list with this  
configuration that isn't an announce only list that i'm on.  There is  
a defense in keeping with the best practices of list management....



> (Yes, it would be a good idea to explore alternative hostings - but  
> this is
> not anything that is going to happen anytime soon.  Please remember  
> that
> your Executive Committee, including our systems officers, are  
> volunteers,
> whose work for AoIR receives no compensation either from AoIR or  
> from their
> home institutions, e.g., in the form of release time.  I don't mean  
> this to
> sound huffy or snarky, but we pretty well have our hands full as it  
> is.  To
> be sure, if the membership want to add such exploration to our  
> agenda, we
> can do that - but it will have to wait until other current matters and
> responsibilities are discharged.  Realistically, this should be an  
> agenda
> item for the new Executive Committee that will be constituted in  
> October,
> following this summer's/winter's elections - more on that soon.)
> To be sure, to paraphrase Jeremy, we could get a lawyer.  But that's
> precisely the problem we're trying to avoid. That is, part of the  
> concern
> some of us have on the Executive Committee is that going down that
> particular path could become extremely expensive extremely quickly,  
> and
> thereby severely undermine what we're otherwise able to do as an
> organization.
>
>> From within this framework, I don't see this as a paternalistic  
>> stance to
> protect 99.99% of our list members from themselves.  I see this as
> anticipating a possible problem, one that, worst-case, could be quite
> serious indeed - and then trying to prevent such a problem if at all
> possible.
>
> Yes, this precise scenario has not happened yet.  Perhaps the  
> chances of it
> happening are vanishingly small.  This is something we're trying to  
> explore
> further, precisely in order to determine how genuinely possible the
> worst-case scenario is.  Please give us some time to work through  
> all of
> that.
> But we didn't just dream this up in order to give ourselves  
> additional work
> (o.k., that was snarky - sorry!). A couple of episodes within recent  
> memory
> and experience (i.e., very unhappy list members who had every good  
> reason in
> the world to want to have a posting removed, which we were unable to  
> do -
> and threatened suit from another list member over a similar issue),  
> while
> separate, have nonetheless suggested to at least a couple of us on  
> the EC
> that this is a possibility we should explore. And, should it turn  
> out to be
> a realistic possibility - one that could, worst-case, inflict  
> serious damage
> on the organization - we are obviously charged with and responsible  
> for
> doing our best to forestall and prevent such damage.
> Again, I don't see that as paternalism - I regret if it has come  
> across that
> way.
>
> In the meantime, I would then ask the membership to further  
> consider: how
> large/small of a possibility does this sort of scenario have to be to
> justify the change in setting?
> I have in mind here a version of Pascal's wager - part of which  
> argues that
> the more significant the consequences of a particular possibility  
> may be,
> the smaller that possibility will be before we take pre-emptive  
> action.
> For example, I generally take an umbrella and drive a car (instead of
> bicycling) if the chance of rain is greater than 30%.  If I end up  
> getting
> wet, it's unpleasant, but not fatal.
> By contrast, the chances of my getting killed while driving are  
> much, much
> smaller - as they must be, in order for me to, in effect, bet that I  
> won't
> die if I drive to work. Yes, the chances of such a worst-case  
> scenario are
> quite small - perhaps vanishingly small.  But the consequences could  
> be
> severe.
> Between these two: the chances of my being killed or seriously hurt if
> involved in a driving accident while not wearing a seatbelt are  
> somewhat
> higher, though still quite small.  But the difference is enough to  
> persuade
> me to wear the seatbelt, despite the minor inconvenience of doing so.
> The analogy I'm trying to offer, then, is that the changed setting is
> something like wearing a seatbelt while driving.  Of course, we cannot
> eliminate all risks.  But given the significantly awful consequences  
> that
> might follow from certain events, however improbable, it seems  
> prudent to do
> what we can to avoid those consequences, even if at the cost of some
> inconvenience.
>
> So, again, how should we bet?  Should we remain with the previous  
> setting,
> and take our chances, because we're betting that the likelihood of  
> such a
> worst-case scenario is so low, and the deficits of changing the  
> setting -
> putting on the seatbelt - offset our improving our chances of avoiding
> potentially awful consequences,
> or do we decide that it's better to avoid the greater risk in favor  
> of a
> somewhat smaller one (i.e., by changing the setting - putting on the
> seatbelt), because the consequences could be egregious, even if the
> possibility of those consequences coming to pass may be very small?
>
> Now that I've jumped in with both feet - just to be clear: I don't  
> have a
> dog in this fight (another reason I've not responded to some of the
> postings).
> Especially because this is a technically-savvy group of scholars and
> researchers who desire and enjoy communicating with one another, I'm  
> just
> not convinced (yet) that a change in setting will in fact result in  
> a severe
> downturn of communication on the list.  But I personally do not care  
> which
> way we decide to go.  I - and the EC - are simply trying to make  
> sure that
> we are doing what we can to be aware of possible problems,  
> especially as
> these are suggested to us by experience, and then pursue possible  
> solutions.
>
> Thanks again to all for the multiple and most interesting comments,
> arguments, observations, and data!
>
> cordially,
> - c.
>
>
> _______________________________________________
> The Air-L at listserv.aoir.org mailing list
> is provided by the Association of Internet Researchers http://aoir.org
> Subscribe, change options or unsubscribe at: http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
>
> Join the Association of Internet Researchers:
> http://www.aoir.org/