[LEAPSECS] it's WP7A week in Geneva

[M. Warner Losh]

In message: <E44632D3-66D3-47D3-8B7A-7D8E7245D582 at noao.edu>
            Rob Seaman <seaman at noao.edu> writes:

: I'm sensitive (really) to the concerns expressed in the title of this  

: talk:

: 	Why leap seconds are difficult to get right for an equipment vendor �  

: Sam Stein, Symmetricom, Inc.

: 

: but the system engineering question here is backwards.  First,  

: discover the requirements.  Second, figure out how to meet them.  That  

: said, I hope the presentations will be posted online.


Sam is very much aware of the practical difficulties in getting leap
seconds right.  He was my boss, and as you all know, I've enumerated
the problems here many times.

While the requirements are "known" up front, the problems in meeting
them are much more non-trivial than people like Rob have said they
should be.  I've been there, done that.  They are hard when you have
to cover the edge cases that vender gear has to cover.  Especially
when you have "fast start" requirements coupled with "deep, cold
spares" that have to know what UTC time is faster than the gps almanac
can deliver it.  Eg, if you have to start up in < 5 minutes at the
right UTC time how do you do that when your GPS receiver has been off
for 2 years sitting on a shelf waiting to be put into serive?

Anyway, Sam will present a fair and balanced view of things.  He's a
good guy who is sensitive to both the benefits of leap seconds, as
well as to their costs.  He's in a position to know exactly what the
costs are as he's been there and done that for a lot longer than I
have...  Plus, he's a good speaker, so it should be an entertaining
talk...

Warner