Re: [gpsd-dev] Planning a 3.11 snap release

[Eric S. Raymond]

Greg Troxel <address@hidden>:
> > GPSD is, in cold fact, that special.  It's used in life-critical navigation
> > and IFF systems.  This justifies extraordinary measures to get something
> > as basic as the time right.
> 
> You've made my point even more strongly than I could have :-)
> 
> Assuming your argument, what's the justification for fetching at build
> time instead of periodically at run time?

We don't know that GPSD will have Internet access. Remember the robot-sub
deployment case - I never forget it, it's a useful extremum.

If we're prepared to assume Internet access, we can also safely assume that 
NTP is available to condition the system clock from sources other than
GPSD itself. That case is easy; we don't need runtime fetching, NTP
does the equivalent for us.

>                                               Or having a separate
> leapseconds package that can be updated immediately when there's a new
> Bulletin C, independently of gpsd code status?

Again, imagine no Internet access.  Imagine you're a low-power SBC a thousand
miles from land...
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>

signature.asc
Description: Digital signature