<html><head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body>
Eric Raymond's <a moz-do-not-send="true" href="https://www.ntpsec.org/white-papers/stratum-1-microserver-howto/">HOWTO
for a stratum 1 server</a> uses <i>gpsd</i> to provide the GPS
data to ntpsec.
<div class="moz-forward-container">
<p>It explains the decision with this:</p>
<p><i>If you are already familiar with ntpd and wonder why this
recipe uses gpsd through SHM rather than ntpd’s native
refclock 20 GPS driver, the answer is this: when refclock 20
is configured to use 1PPS, it mixes in-band time data with
1PPS in a way that causes it to behave badly, and possibly get
rejected as a falseticker, when 1PPS is only occasionally
available.</i></p>
<p>The history notes on the page indicate it was first published
late in 2016, and last updated in the middle of 2018. There has
been a lot of water under the bridge, and many changes to ntpsec
since then.<br>
</p>
<p>My question is this: Four years on, is this still a valid
reason to prefer <i>gpsd</i> and <i>refclock shm</i> over <i>refclock
nmea</i> and <i>refclock pps</i> ?</p>
<p>In my mind, using the <i>refclock shm</i> and also running <i>gpsd</i>
is more complication than relying on the two <i>refclock</i>
already available. But if the latter is prone to unexpected
failures, then there are good reasons to accept the complication
of the former.<br>
</p>
<p>(Appologies if this is a duplicate message. I have not received
my copy of the message I sent earlier this moth, nor has it
appeared in the web-archives, nor have I received any reply to
my query to the users-owner email address.)<br>
</p>
<pre class="moz-signature" cols="72">--
--
Do things because you should, not just because you can.
John Thurston 907-465-8591
<a class="moz-txt-link-abbreviated moz-txt-link-freetext" href="mailto:John.Thurston@alaska.gov" moz-do-not-send="true">John.Thurston@alaska.gov</a>
Department of Administration
State of Alaska</pre>
</div>
</body>
</html>