[Top] [All Lists]

Re: Intent to revive "expires" header from draft-ietf-mailext-new-fields-15

2008-07-23 07:48:06

Paul Smith wrote:

what would you want the recipient MUA to do with these.

Whatever you like them to do, nothing is a valid option.

Some MUAs offer "priorities" (high-normal-bulk, or similar),
you could let your MUA treat "expired" mails as "bulk".

MUAs present "unread" mail in some order.  You could let
your MUA display "expired" mails last, or automatically
flag such mails as "read".  

Automatically moving them into another folder when they
arrive is nearly the same as automatically deleting them
as far as many users goes

Depends on the user.  I can't guarantee it, but I do check
spam folders (briefly), and rescue what's no spam.  Mostly
in an attempt to educate GMail.  With another ISP I check
the spam folder when its content is over 10 mails after
about a week.  IOW, I look at it, but only rarely.

GMail filters are not very sophisticated at the technical
side (= they are for ordinary users, not the folks on this
list :-), but in theory they could offer to "label" expired
mails as "expired".  

Whatever it was that "expired" back in 2007, likely I don't
need it anymore in 2008.  "Expires:" might help to sort and
trim old mail archives.

"confirm your subscription to list X in the next 3 days"
is still important if you receive it 4 days later. Now you
know you need to try again to get onto the list.

We're not in the same parallel universe.  If that was really
my subscription I'd have confirmed it immediately, therefore
it was somebody else claiming to be me while I was absent :-)

My problem is that I can't think of anything which is 
*absolutely* no use after a certain date

That is no big problem, you then simply don't use this info.
There are things in e-mail (not only certain header fields)
I never use, but hopefully others can make good use of them.

The attempt to revive "Expires:" is a long shot, and maybe
it won't get traction.  But based on Keith's semantics I
think it is relatively harmless as worst case, and as best
case potentially useful.


<Prev in Thread] Current Thread [Next in Thread>