[Ietf-calsify] Section 3.8.4.7: Recommended syntax for the UID property

Martin Kiff mgk at webfeet.co.uk
Thu Feb 22 22:51:48 PST 2007


Cyrus Daboo wrote:

> Hi Martin,
>
> --On February 22, 2007 4:39:48 PM +0100 Martin Kiff 
> <mgk at webfeet.co.uk> wrote:
>
>> If you put an .ics file on the web, you've published a lot of email-like
>> addresses; be prepared for them to be picked up by Spam-harvesters. I
>> think the spec should suggest some other format.
>
>
> The reality is a lot of clients tend to use a "GUID". We should 
> probably add that as an alternative recommended format.
>
I think it would be better - or I've seen a URL referring to just that 
single event which keeps the transparency (GUID's would be opaque? There 
would be no way by eye to confirm that they are unique?) I think in DNS 
data the contacts are user-part.domain-part

> As to spam harvesting, its only the domain portion that is a valid 
> "network address", and frankly that is likely to have much more 
> exposure via email, so I would not be more concerned about having that 
> in .ics data than the current state of affairs.
>
Let's say, it is something I've seen in practice... messages bouncing 
off the spam defences sent to addresses taken from .ics files.

No, they wouldn't (shouldn't!) be real addresses but they load your ISP 
connection and add stress to your defences and it's a bit tough if you 
have a 'catch all' mailbox for your domain.

Regards, Martin


More information about the Ietf-calsify mailing list