[LEAPSECS] Coming of age in the solar system
M. Warner Losh
imp at bsdimp.com
Sun Sep 5 08:50:27 EDT 2010
In message: <1283669678.9574.171.camel at localhost>
Paul Sheer <p at 2038bug.com> writes:
: At the very least, UTC-SLS is an obvious solution
: that fixes the problem for 99% of the remaining 1%
: of companies that are actually effected by the problem.
Except, as previously noted, these half solutions cause more problems
than they are worth and actively get in the way of doing things right
for systems that care and matter. It leads to infrastructures that
need to be replaced or circumvented in order to come close to doing
the right thing. It leads to large errors that didn't matter 5 or 10
or 15 years ago to many people, but increasingly do matter in today's
highly connected economy. 10 years ago, financial markets would have
been happy with SLS, but today a 500ms error in time can cost people
millions (and have). Many other market segments are likely to follow.
The more entrenched these half solutions become, the more difficult it
is to do things properly. This suggests that you'll have a fragmented
effect on time: some folks will follow the strict SI ticking, while
others will slew on this day and you get divergence in what the right
: Hmmmm *perhaps* *there* *are* *some* *talented*
: *programmers* *out* *there* *that* *will* *implement*
: *this* one* *day*.
Maybe the talented programmers have implemented that, and found that
it was such a dumb idea they didn't want to share that code with the
rest of the world? Or were smart enough to understand the systemic
effects of it and decided to pass since it has all the problems of
having to know where there are leap seconds and makes a number of
assumptions about what kind of errors are acceptable to everybody
which might prove to be unwise.
More information about the LEAPSECS