Re: [LEAPSECS] Rubber seconds

From: Ed Davies <ls_at_EDAVIES.NILDRAM.CO.UK>
Date: Fri, 05 Aug 2005 15:43:01 +0100

Poul-Henning Kamp, replying to Markus Kuhn, wrote:
>>>Also, your UTS proposal is a total non-starter: Rubber seconds is not
>>>a usable solution.

I replied:
>>Whether rubber seconds are usable or not depends on
>>what problem you intend them to be a solution to. If
> you just want a monotonic timestamp stream to be able
> to give ordered labels to events (e.g., file system
> transactions) then they are pretty good.
>
> Similarly, making sure that scheduled events don't
> fall between the cracks of clock updates is a lot
> easier if you steer the clock via rate changes rather
> than make jumpy phase changes.

Poul-Henning Kamp replied, quoting only the first
sentence of my first paragraph:
> Rubber seconds are _never_ usable, just like
> rubber meters, rubber kilos and rubber unit charges
> are never usable.

Why do you write this when I've just given a examples
of cases where rubber seconds are usable? It might
help if you could explain what you think is wrong with
my examples.

Please note that I'm not suggesting that rubber seconds
are good in all cases: just that there are some
circumstances where they are acceptable and, maybe,
preferable to having non-monotonic times or times like
14:59:60.

Remember that this is a list for the discussion of the
future of leap seconds. Any answer which simply assumes
that leap seconds should go away is rather missing the
point. Personally, I think that a switch to TAI (or at
least frozen UTC) is, on balance, probably a good idea
but that before such a decision is made there should
be a proper discussion of what the real issues with
keeping leap seconds are (and also, of course, what the
real issues with discarding them are - though that's
not so relevant to rubber seconds).

Ed Davies.
Received on Fri Aug 05 2005 - 07:44:37 PDT