Subject: Re: identd...
To: Peter Seebach <seebs@plethora.net>
From: Andrew Brown <atatat@atatdot.net>
List: current-users
Date: 12/13/2000 15:55:43
>>Well, it does matter.
>
>Okay, I see your point.  I was thinking in terms of "if I don't care, why
>should I worry".  My initial response was going to be "if I care, I can
>just put the information out...", but then I do have the privacy concern,
>so something like encryption or timestamps is necessary, and if I want to
>use timestamps, I *do* need to keep the records.
>
>That said, I still don't feel I necessarily have an obligation to anyone
>*outside* my systems to keep track of the timestamps longer than I want to
>for my own purposes... but I can see why it would be good to keep them.

the timestamp thing was something i popped in since it seemd (to me,
at least) a good way to get rather unique "cookies" for each query.  i
doubt very much (but would be delighted if someone could
authoritatively comment either positively or negatively about this)
that two calls to gettimeofday(2) could ever return the same struct
timeval.

to take the cookie thing a little further, have your identd simply
hand out a strictly increasing value (0, then 1, then 2, etc), keep
track of what it handed out last, and what each number maps to (either
via logging or some other simple method).

my point was only that they be unique.

-- 
|-----< "CODE WARRIOR" >-----|
codewarrior@daemon.org             * "ah!  i see you have the internet
twofsonet@graffiti.com (Andrew Brown)                that goes *ping*!"
andrew@crossbar.com       * "information is power -- share the wealth."