Okay. I was just under the impression UTC and UCT are the same and thus have little reason to be distinct. It just seemed odd they were different, especially when they're used synonymously, and knowing tzdata is so broadly used for basically everything, I could reasonably see a ton of people relying on the non-aliasing behavior. The citations seem interesting, so it is definitely worth a mountain and three fourths of testing if you were to make any changes to `Etc/UCT`. ----- Isiah Meadows contact@isiahmeadows.com www.isiahmeadows.com On Tue, Mar 5, 2019 at 4:45 PM Paul Eggert <eggert@cs.ucla.edu> wrote:
On 3/5/19 12:00 PM, Isiah Meadows wrote:
So basically just back-compat?
That's one reason it's there, yes.
My impression is that "UCT" is used on occasion as seen in the citations below, although it's by far less common than "UTC". The question is whether a shell command like "TZ=Etc/UCT date" should generate the abbreviation "UCT" (perpetuating the "typo") or "UTC" ("correcting" it). I am quoting the words "typo" and "correcting" so as not to prejudge the matter.
Some sample uses of "UCT":
Gray JE, Machlan HE, Allan DW. The effect of humidity on commercial cesium beam atomic clocks. Proc 42nd Ann Freq Control Symp 1988, 514-518. https://dx.doi.org/10.1109/FREQ.1988.27649 "The implications of such changes are quite significant in the generation of International Atomic Time (TAI) and of Universal Coordinated Time (UCT)." The authors were from the Time & Frequency Division, NIST.
Kouba J. Relativistic time transformations in GPS. GPS Solutions (2002) 5: 1. https://doi.org/10.1007/PL00012907 "It is interesting to note that TT, TAI, and UTC (the Universal Coordinated Time) all have nominally the same ..." The authorwas from the Geodetic Survey Division, Natural Resources Canada.
Messerschmit DG. Relativistic Timekeeping, Motion, and Gravity in Distributed Systems. Proc IEEE, vol. 105, no. 8, pp. 1511-1573, Aug. 2017. https://dx.doi.org/10.1109/JPROC.2017.2717980 "Universal coordinated time (UTC) is the common civil definition of time." The author is an emitus professor in the EECS dept, UC Berkeley.
Freeman RL. Standard time and frequency. Reference Manual for Telecommunications Engineering, R. L. Freeman (Ed.) (2002). https://doi.org/10.1002/0471208051.fre035 "Another necessity, in some cases, is to be able to derive universal coordinated time (UCT) or a comparable time scale." Freeman's a widely-published telecom author.