Re: [ga] @xxxxxxxxx

[crispin@xxxxxxxxx]

Hi Danny 

It seems we both agree on the desirability of obscuring email addresses
in web archives.  The address obscuring algorithm you reference is built
into Mhonarc, and ICANN has been using it for its list archives since at
least November of last year, with no complaints (see, for example,
http://forum.icann.org/reform-comments/general/msg00145.html). 

I understand that in some cases the Mhonarc algorithm may be
inconvenient.  For several reasons, however, I don't agree with your
desire to change that algorithm.  While I don't expect to convince you,
I thought I would give some of those reasons:

    1) The address obscuring algorithm is part of Mhonarc.  While it
    would certainly be possible to customize Mhonarc to do something
    else, the long term cost of any customization has to be considered. 
    The choice isn't just between two end results, one has to consider
    the effort required to achieve those results, and in the ICANN
    environment it would take some work to do what you suggest. 

    2) Use of any algorithm that leaves an easily human decipherable
    address is based on the assumption that address scrapers will remain
    stupid.  That is not at all a good assumption, in my opinion.  The
    Mhonarc algorithm of completly overwriting the address is extremely
    resistant any future developments by spammers. 

    3) There is a good privacy argument that we should be very resistant
    to publishing someone's email address without very good reason.  In
    fact, one could argue that the real issue is whether we should go
    back and obscure all the email addresses in old archives.  Had I the
    resources to spare, I'd be more inclined to spend them there.

    4) I believe that you have seriously overstated the magnititude of
    the problem you have identified.  Any person who is a member of a
    list has complete access to all the addresses ever published; the
    primary purpose of the list archives is to provide a record of
    public discussion, not to provide a convenient directory for 
    participants email addresses.

    5) And there are essentially trivial workarounds for anyone who really 
    feels that their address must be visible in the archives -- they could
    simply use your trick, or give a URL.

As in most things, there are tradeoffs.  In my opinion the tradeoffs in
this case come down very strongly on the side I have described.

Regards
Kent Crispin


On Tue, Jul 22, 2003 at 12:24:32PM -0400, DannyYounger@xxxxxx wrote:
> Dear Kent,
> 
> Recently Matt Hooker posted a letter to the IPC list in response to a query 
> by IPC president Jane Mutimear asking for details on instances of registrars 
> failing to comply with their obligations under the registrar accreditation 
> agreement.
> 
> Matt's letter posted at http://ipc.songbird.com/members-archive/msg00147.html 
> contained the following appeal:  "If you would like to join us in lawsuits 
> against these criminals, e-mail us at StrategicInvestCom@xxxxxxxxx and let us 
> know. We need partners in litigation to defray the up front costs of the 
> litigation."
> 
> Kent, unfortunately owing to the way you have this mailing list configured 
> (as well as the GA archives), all email addresses are converted to @xxxxxxxxxx.  
> How would anyone be able to get in contact with Matt if they wished to do so? 
>  This is not helpful.  
> 
> Please allow us to view the e-mail addresses.
> 
> Thank you,
> Danny Younger
> dannyyounger[at]cs.com

-- 
Kent Crispin                               "Be good, and you will be
Technical Systems Manager, ICANN            lonesome."
crispin@xxxxxxxxx,kent@xxxxxxxxxxxx                    -- Mark Twain
               p: +1 310 823 9358  f: +1 310 823 8649