NavList:

There are many, many ways to get accurate UT these days, and everyone should have at least two independent sources.

• I: The shortwave broadcasts from WWV remain as a time-honored standard with "all the bugs worked out". Anyone interested in celestial navigation should probably obtain a cheap shortwave radio just so you can listen to those time ticks for occasional confirmation to validate other, more convenient sources
• II: There is a direct link to the USNO time right on the NavList main web page: select the "DATA" menu, then select "USNO Universal Time". I added this during the big overhaul of the site layout back in June.
• III: There are numerous Internet time web sites, including NIST servers and the 'time.is' site that Rommel mentioned nearly all of which use the latency (delay) accounting trick which Lu Abel mentioned in one post. This standard for measuring and then removing the round-trip travel time during synchronization is an essential component of the Internet "Network Time Protocol" which the great majority of Internet time web sites use. You can trust the time displayed by the USNO time web site and those other time web sites within less than one second error and normally less than 0.2 seconds error. Reliable, high-speed Internet access does help, but these time server web sites today are generally trustworthy even with less reliable Internet service.
• IV: Modern smartphones have several different, independent methods for displaying accurate UT, again reliable within a fraction of a second. Since nearly all smartphones have GPS receivers built-in, they have direct access to highly accurate GPS time. While older devices sometimes could not "keep up" to display the time correctly synchronized that problem has mostly disappeared. For Android users, I recommend downloading an app called "GPS Test" which has many useful features including a very nice UTC display. Smartphones also access the mobile network time, and this is usually the source of the time displayed by the standard clock functions on the smartphone. And finally smartphones have easy access to numerous apps which can display network time not really different from the time server web sites in III (above). In my experience with new smartphones in the past two years, these three independent methods of acquiring UT all agree within a fraction of a second.
• V: Stand-alone GPS receivers also have access to highly accurate GPS time. Modern GPS devices should no longer suffer from slow displays which were a problem ten years ago, but this is a declining prdoct category, and I feel that each device and its display should be independently tested with some other source for UT.

Of course, with software solutions, particularly on rapidly-evolving platforms like smartphones, one has to be wary of third-party apps and other products. They are blackboxes (then again, so are all clocks). What will they do if they do lose their Internet connection? What happens when they are left running for an extended period of time? Do they frequently re-synchronize? So in general, we need some procedure for testing and validating these time sources. When we have genuinely independent sources, like GPS time on a smartphone, and Internet time on a separate computer, just like multiple chronometers 150 years ago, they serve as checks on each other. If two such independent time sources differ by less than a quarter of a second from each other, then there is a high probability that they are both within half a second of the correct Universal Time. And for occasional more certain validation of the time, fire up the shortwave and watch the digits click over on your Internet time source or your smartphone's GPS software as the beeps and clicks come in from WWV. You may be amazed by how well they agree. 

Bottom line: Yes, buy a shortwave radio. Yes, trust Internet time servers. Yes, trust GPS time displayed on modern devices (smartphones). Trust but verify... by comparing multiple sources.

Frank Reed
Conanicut Island  USA