[ntp:questions] NTP & PPS, part 2 ;)

[Harlan Stenn]

While the following is not as applicable to the debian packaging
example, it's still on-point.

I'll also add that it's penny-wise and pound-foolish for us to take
patches that fix something for 1 specific case and have it break things
for anybody else.

The people who submit patches that are readily included in the
distribution are folks who make sure they have *portable* solutions.

If we get a patch that is clearly non-portable, it sits there until
somebody else (a volunteer) can look at it and figure out how to
implement the fix in a portable way.  Sometimes it helps seeing what
somebody comes up with.  Sometimes it's not all that helpful.

If the issue is one that affects a lot of people, it will get attention
sooner than an issue that affects fewer people.

Remember, we have a *lot* of open issues out there and each volunteer
chooses the bugs they work on, and how much time they choose to spend on
NTP.  I'm currently paid to put in 15-20 hours/week on NTP development,
and I'm working on the highest-priority bugs that are blocking the
release of 4.2.8.  I put in *many* more hours as a volunteer on NTP, and
many more still for NTF.

Put another way, NTP is successful because it approaches timekeeping
from a "wide scope".  It is running on many tens of millions of devices
that cover a wide range of operating conditions.  The volunteers who
actively work on the project have a wide-scope attitude, too.  They may
not have started with a wide-scope attitude, but their "usefulness" to
the project (and to NTP's constituency) are directly related to how well
they can embrace this wider-scope attitude.  We do have some volunteers
who have a narrow focus.  They work in limited areas.

H