Re: Request for review of adjtimex(2) man page

From: Laurent Georget
Date: Wed Jan 07 2015 - 08:37:40 EST


Hello,


>I've added a large number
> of FIXMEs in the draft below, and would be happy if anyone can supply
> some content to fill any of the gaps.

Isn't the man page going to be very long and hard to read if we explain all
the NTP internals in adjtimex.2? I think we could create a second man page
(for instance ntp.7) in section 7 to provide extensive explanations and refer
to it from adjtimex.2. Does that seem okay?

> .BR STA_PLL
> Enable phase-locked loop (PLL) updates (read-write) via
> .\" FIXME Any pointer to further information about what this means?
> .\" (It was not immediately obvious from a scan of the RFC, whether
> .\" this is described in the RFC.)
> .BR ADJ_OFFSET .

> .BR STA_FLL
> Select frequency-locked loop (FLL) mode (read-write).
> .\" FIXME Any pointer to further information about what this means?
> .\" (It was not immediately obvious from a scan of the RFC, whether
> .\" this is described in the RFC.)

In this particular case, for example, PLL and FLL are the two modes by which
the clock offset can be controlled. Roughly speaking, PLL is phase locking
(the clock time is adjusted to the source time) and FLL is frequency locking
(the clock frequency is adjusted to the source frequency). Is that level of
details be sufficient for adjtimex.2?

Valuable resources that go beyond the content of the RFC are publications and
presentations from Pr. David Mills (http://www.eecis.udel.edu/~mills/) like
http://www.eecis.udel.edu/~mills/database/brief/clock/clock.pdf.

Cheers,

Laurent

Attachment: signature.asc
Description: OpenPGP digital signature