[LEAPSECS] Do lawyers care (know) about leap seconds?

[Warner Losh]

On Oct 1, 2014, at 6:02 AM, Greg Hennessy <greg.hennessy at cox.net> wrote:

>> But the basic point still remains: If you have to sugar coat the actual standard
>> with a fake standard to paper-over people’s inability to deal with the actual
>> standard, this suggests that you have the wrong actual standard.
> 
> I would agree that we have the wrong actual standard. We've had leap
> seconds since 1972, but POSIX still mandates we ignore the leap seconds
> in places. It would be nice if the standards and the practices match.
> Some people want to change the standards, and others want to change
> the practices.

Another example of well-intended people sugar coating the UTC standard and
causing problems. It is precisely well-intended efforts like this which has made
it effectively impossible to implement the UTC standard pedantically correctly. UTC
gets a number of things right, as a purely academic standard. And it is quite useful
for those purposes. However, as a real-world standard, experience of the last 42
years of implementation suggests that it is hard to actually, really implement.

On the other hand, the POSIX standard is easy to implement and generally
hard to get wrong. As a programming standard, it has made things easy. It
is this experience that leads many to desire that harmonization between the
two go POSIX (time w/o leap seconds) instead of UTC, which gets the agreement
better, but had weird warts to make synchronization appear invisible (but the
:60 and time time-slip for those not implementing it exposes this invisibility
as being illusionary).

Warner

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 842 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://pairlist6.pair.net/pipermail/leapsecs/attachments/20141001/d8faec6d/attachment.pgp>