[ale]OT It begins...

[Robert Reese]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Jim

*********** REPLY SEPARATOR  ***********
On 1/27/2004 at 1:26 AM Jim Popovitch wrote:

>There is no feasible way to do what you are asking for, short of
>removing the sender's email address entirely.

True.  But that is entirely feasible.


>  Even then... it is
>ineffective because the body of the email may contain email
>addresses from a reply or forward, and the diversity of email
>clients makes it  

Not exactly, since forwards are almost always editable to remove
email address of others.  While you won't get 100% compliance, you
will find that those that are 'clued in' will happily remove other
people's email addresses from their forwarded messages.  In fact,
given the technical ability of those on this list, it would seem
trivial for us to adjust our email client's forwarding scheme to not
include email addresses.


>near impossible to determine where a client may inject the original
>posters email address (headers included).  

First, headers could be eliminated altogether in an archive if you
wish.  Most people probably would like at least the message IDs to
remain, however.  Neither of these presents much of a technical
problem, as each line in the headers begins with a title.  Filtering
these prior to posting to the archive should be relatively easy,
especially since it is the product of the mailing list engine and
therefore should be fairly uniform.  The individual client additions,
IIRC, should all begin with "X", so filtering those out ought to also
be easy enough.


>Given the use of email addresses in sigs, RFCs, announcements, and
>code samples (not to mention emails containing valid addresses that
>posters are trying to share, i.e. Congressmen, etc), it would
>require human
>intervention (i.e. a moderator, also not perfect) to scan each email
>and decide what email addresses should remain. 

For email addresses in the body, the human intervention is already
there; it's proper position in the sequence is just prior to sending
the message.  It shouldn't happen in the processing phase. ;c)   As
for 'sigs' that is up to the individual that inserts the signature. 
If they wish to share their unadulterated email address within the
body of an email, that is their choice; they won't get upset if it
ends up in another email (or shouldn't, anyway).

Email addresses being shared are often done so with the shared
addressee's knowledge that it is exposed.  Government email addresses
are already assumed to be exposed, by their nature.


>If you don't want your address public, get a diff email account for
>public lists.

I do.  Doesn't mean I shouldn't work to protect it as long as
possible, however.  ;c)  I'd rather it take four years, not four
weeks, to get a dozen messages a week.  ;c)


It can be done.  I've seen it over and over.

Cheers,
Robert Reese~

-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0.3

iQA/AwUBQBafzbw8BOWncaQMEQJbwQCgv3opaCN23R7j24PL6Y4gyMgoRgEAoIA9
Xazoda8HGp6Ek4Pj/mRj7IJf
=R22G
-----END PGP SIGNATURE-----


Type: DH/DSS 4096/1024 AES-256
Key ID: 0xA771A40C
Fingerprint: CAE2 81CA A7CD 6681 341C  E3A9 BC3C 04E5 A771 A40C