[LEAPSECS] Bloomberg announced its smear

[Stephen Colebourne]

On 24 September 2016 at 21:52, Warner Losh <imp at bsdimp.com> wrote:
> I tend to view smearing as a repudiation of the concept of leap
> seconds. Let's do this kludge to keep the crazy world of nutty
> software written by people that don't know or don't care about leap
> seconds hobbling along at the expense of a bit of accuracy in time to
> try to shoe-horn the POSIX time_t without leap seconds standard with
> the reality of leap seconds. Better to not have them in the first
> place, but that ship sailed 40 yaers ago.

"crazy world of nutty software"?

The harsh truth for time nerds is that the vast majority of people and
developers in the world do not care about leap seconds. At all. They
have no desire to specify business rules around them, or think about
them in any way. The vast majority of applications do not care if the
length of a second is slightly longer (as it is with smearing). What
is baked into a gazillion pieces of code is that there are always 60
seconds in a minute, and calling those applications nutty, or buggy,
or wrong isn't helping - after all, the vast majority of them perform
a useful job for their authors.

None of the above denies the fact that there are some people and
applications that care about leap seconds. But they are by far in the
minority (yet almost certainly over-represented on this list).

It really is time for the time nerd community to hold their noses and
accept that there should be some agreed way to map between time as it
really is (variable length days with leap seconds) and time as the
vast majority want to view it (86400 subdivisions per day). As I said
before, I don't care what the smearing standard is, but we really
really need one.

And of course once agreed, the OS should provide APIs for both smeared
and leaping UTC.

Stephen