Bot Version: v1.3.2.1 (old)
Page: https://en.wikipedia.org/wiki/Global_Positioning_System
Diff: https://en.wikipedia.org/w/index.php?title=Global_Positioning_System&type=revision&diff=781479255&oldid=781013068

Original Text:

As of January 2017, GPS time is 18 seconds ahead of UTC because of the leap second added to UTC on December 31, 2016.<ref>{{cite web|title=Notice Advisory to Navstar Users (NANU) 2016069|accessdate=January 2, 2017|url=http://www.navcen.uscg.gov/?pageName=currentNanus&format=txt|archive-url=https://gps.afspc.af.mil/gps/archive/2016/nanus/2016069.nnu|archive-date=November 30, 2016|publisher=GPS Operations Center}}</ref>

Changed To:

As of January 2017, GPS time is 18 seconds ahead of UTC because of the leap second added to UTC on December 31, 2016.<ref>{{cite web|title=Notice Advisory to Navstar Users (NANU) 2016069 |accessdate=January 2, 2017 |url=http://www.navcen.uscg.gov/?pageName=currentNanus&format=txt |archive-url=https://web.archive.org/web/20131909223500/http://www.navcen.uscg.gov/?pageName=currentNanus&format=txt |archive-date=May 21, 2017 |publisher=GPS Operations Center |deadurl=yes |df= }}</ref>

Context: The currentNanus announcement file hosted by www.navcen.uscg.gov was cited as the authoritative source, but as the announcement file is updated over time, https://gps.afspc.af.mil/gps/archive/2016/nanus/2016069.nnu was provided as an archive-url of the announcement. (I'll admit that it was a bad idea for the human editor (me) to cite an announcement like this, but I still think InternetArchiveBot's behavior could be improved).

I feel that the InternetArchiveBot's actions were inappropriate because:

• The bot shouldn't have changed the archive-url from gps.afspc.af.mil to web.archive.org without good reason (it's possible that it incorrectly assumed https://gps.afspc.af.mil to be invalid because of the DoD CA used to sign ".mil" domains)
• It chose a very old archive from web.archive.org despite more recent archives being available: https://web.archive.org/web/*/http://www.navcen.uscg.gov/?pageName=currentNanus&format=txt (although in this instance, web.archive.org doesn't have any snapshots of the currentNanus text file taken at the correct time to capture the announcement)

Coincidentally, it also incorrectly flagged http://www.navcen.uscg.gov/?pageName=currentNanus&format=txt as a deadurl (is it possible that it is actually referring to the gps.afspc.af.mil archive-url that it removed?)