Re: draft-ietf-dhc-dhcpv6-opt-timeconfig-02.txt

[Leslie Daigle <leslie@thinkingcat.com>]

CALSCH circa 1997???

If DHCP is supposed to be simple (yeah, yeah) then it
seems like this needs to be more crisp about what exactly
clients should expect of the message.  E.g., explicitly
that "this is the timezone info at the time the message
was sent."  Smart clients will take that under advisement
and update themselves by their own processes.  Dumb clients
will know they have to re-ping the DHCP server periodically.


Leslie.


Harald Tveit Alvestrand wrote:
> --On fredag, april 25, 2003 08:22:02 -0700 Bill Fenner 
> <fenner@research.att.com> wrote:
>
> > Indeed.  It's a hard problem to decide what is useful.  The name in the
> > Olson database is more useful if the node has the Olson database, since
> > it can know when historical daylight savings time was.  (Was it Daylight
> > Savings time on April 20, 1969?)  The POSIX rules allow a system without
> > the database to decide whether or not it's DST right now but doesn't say
> > anything about when the rule is applicable and when it might have 
> > changed.
> >
> > This almost argues that instead of either/or, the option should always
> > carry both and let the client decide which one it finds useful.
>
> argues for two options - one for Olsen, one for POSIX?
> and a third for "someone else's rules"???????
>
> Leslie - have we been here before? It feels so familiar.....







--

-------------------------------------------------------------------
"Reality:
    Yours to discover."
                               -- ThinkingCat
Leslie Daigle
leslie@thinkingcat.com
-------------------------------------------------------------------