Multiple UTC zones
Does anyone know if there is a concrete reason for why Etc/GMT, Etc/UTC and Etc/UCT are three separate zones? I would think that Etc/UTC would be the only zone entry and the others would be links. Thanks.-Matt
I believe the abbreviations displayed are different; each may therefore have its own purpose. -- Tim Parenti On 24 June 2014 16:25, Matt Johnson <mj1856@hotmail.com> wrote:
Does anyone know if there is a concrete reason for why Etc/GMT, Etc/UTC and Etc/UCT are three separate zones? I would think that Etc/UTC would be the only zone entry and the others would be links.
Thanks. -Matt
On 06/24/2014 01:25 PM, Matt Johnson wrote:
Does anyone know if there is a concrete reason for why Etc/GMT, Etc/UTC and Etc/UCT are three separate zones?
They use different time zone abbreviations, and so cannot be links. For example: $ TZ=Etc/GMT date; TZ=Etc/UCT date; TZ=Etc/UTC date; date -u Tue Jun 24 20:31:12 GMT 2014 Tue Jun 24 20:31:12 UCT 2014 Tue Jun 24 20:31:12 UTC 2014 Tue Jun 24 20:31:12 GMT 2014 The last line of output depends on your 'date' implementation; GNU 'date', for example, says "UTC", but tz 'date' says "GMT".
Come to think of it, the tz implementation of 'date -u' should say "UTC" instead of "GMT", and in general the tz code and documentation should prefer UT or UTC to GMT whenever this would improve technical accuracy. Although "GMT" is the traditional time zone abbreviation output of 'date -u', POSIX has allowed "UTC" ever since IEEE Std 1003.1-1992. Outputting "UTC" is more technically correct, certainly for time stamps since 1961, and arguably even before that if one interprets "UTC" proleptically. Also, outputting "UTC" is now a quite-common behavior, since it's the standard behavior in GNU/Linux. So I'll look into proposing a patch to the tz code to have it support this behavior. There are probably a few other places in the code that should also prefer "UT" or "UTC" to "GMT".
I think there are other places affected by this also. Where I noticed was in the CLDR, as discussed here:http://stackoverflow.com/a/24395573/634824
Date: Tue, 24 Jun 2014 14:02:10 -0700 From: eggert@cs.ucla.edu To: tz@iana.org CC: mj1856@hotmail.com Subject: 'date -u' should say "UTC", not "GMT"
Come to think of it, the tz implementation of 'date -u' should say "UTC" instead of "GMT", and in general the tz code and documentation should prefer UT or UTC to GMT whenever this would improve technical accuracy. Although "GMT" is the traditional time zone abbreviation output of 'date -u', POSIX has allowed "UTC" ever since IEEE Std 1003.1-1992. Outputting "UTC" is more technically correct, certainly for time stamps since 1961, and arguably even before that if one interprets "UTC" proleptically. Also, outputting "UTC" is now a quite-common behavior, since it's the standard behavior in GNU/Linux. So I'll look into proposing a patch to the tz code to have it support this behavior. There are probably a few other places in the code that should also prefer "UT" or "UTC" to "GMT".
On Jun 24, 2014, at 5:02 PM, Paul Eggert <eggert@cs.ucla.edu> wrote:
Come to think of it, the tz implementation of 'date -u' should say "UTC" instead of "GMT", and in general the tz code and documentation should prefer UT or UTC to GMT whenever this would improve technical accuracy. Although "GMT" is the traditional time zone abbreviation output of 'date -u', POSIX has allowed "UTC" ever since IEEE Std 1003.1-1992. Outputting "UTC" is more technically correct, certainly for time stamps since 1961, and arguably even before that if one interprets "UTC" proleptically. Also, outputting "UTC" is now a quite-common behavior, since it's the standard behavior in GNU/Linux. So I'll look into proposing a patch to the tz code to have it support this behavior. There are probably a few other places in the code that should also prefer "UT" or "UTC" to "GMT".
UTC, yes. UT? I know of UT1 and UT2, but neither are the same as UTC and neither is applicable here. paul
On 06/24/2014 02:10 PM, Paul_Koning@Dell.com wrote:
UTC, yes. UT?
UT a.k.a. Universal Time" is deliberately ambiguous: it refers to all the UTx variants together. Technically, GMT can refer to either a civil day (day starts at midnight) or an astronomical day (day starts at noon), and UT avoids this twelve-hour ambiguity by standardizing on the civil day. Until 1952 the US Naval Observatory attempted to resolve the ambiguity by using the term "Greenwich Civil Time" (GCT) for civil-day GMT but that terminology did not catch on widely (it wasn't used in Britain), whereas UT has.
Paul Eggert said:
UTC, yes. UT? UT a.k.a. Universal Time" is deliberately ambiguous: it refers to all the UTx variants together.
Are you sure of that? It has always been my understanding that UT refers to either UT0 or UT1, but *never* UTC. -- Clive D.W. Feather | If you lie to the compiler, Email: clive@davros.org | it will get its revenge. Web: http://www.davros.org | - Henry Spencer Mobile: +44 7973 377646
On 06/24/2014 02:34 PM, Clive D.W. Feather wrote:
Paul Eggert said:
UTC, yes. UT? UT a.k.a. Universal Time" is deliberately ambiguous: it refers to all the UTx variants together. Are you sure of that?
It depends on who you talk to, but that's the sense I intended. You can see similar usages in the Collins English dictionary and in Wikipedia: http://en.wikipedia.org/wiki/Universal_Time#Versions http://dictionary.reference.com/browse/universal-time I take your point that some sources are stricter, and say that UT must be UT0 or UT1 (or maybe UT1R or UT2...), so I suppose it's yet another term that is controversial to some extent.
Paul Eggert <eggert@cs.ucla.edu> wrote:
Technically, GMT can refer to either a civil day (day starts at midnight) or an astronomical day (day starts at noon), and UT avoids this twelve-hour ambiguity by standardizing on the civil day. Until 1952 the US Naval Observatory attempted to resolve the ambiguity by using the term "Greenwich Civil Time" (GCT) for civil-day GMT but that terminology did not catch on widely (it wasn't used in Britain), whereas UT has.
Britain (the RGO) switched from the astronomical to the civil meaning of GMT in 1925. http://www.apparent-wind.com/gmt-explained.html Tony. -- f.anthony.n.finch <dot@dotat.at> http://dotat.at/ Irish Sea: Southeast 4 or 5. Slight. Occasional rain. Moderate or good.
On Tue 2014-06-24T21:10:25 +0000, Paul_Koning@dell.com hath writ:
UTC, yes. UT? I know of UT1 and UT2, but neither are the same as UTC and neither is applicable here.
The name that was in contemporary use during the 1960s is another tl;dr story as seen in http://www.ucolick.org/~sla/leapsecs/timescales.html#UTC "UTC" was not used in print until 1965, was not found in broadcast time signals nor any recommendation until 1974, and all around that range of dates different agencies employed different terminologies. -- Steve Allen <sla@ucolick.org> WGS-84 (GPS) UCO/Lick Observatory--ISB Natural Sciences II, Room 165 Lat +36.99855 1156 High Street Voice: +1 831 459 3046 Lng -122.06015 Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m
Steve Allen <sla@ucolick.org> wrote:
Regarding that page you might want to add a note about the 1972 change to UTC in the time signals from the RGO. They still referred to their official time scale after this change as GMT. So it is wrong to say that GMT does not now and has never had leap seconds. http://www.lib.cam.ac.uk/deptserv/manuscripts/RGO_history/rgo_home_ch5.html Tony. -- f.anthony.n.finch <dot@dotat.at> http://dotat.at/ Shannon: Cyclonic 4 or 5, increasing 6 at times. Slight or moderate. Rain or showers. Moderate or good.
On Tue 2014-06-24T14:02:10 -0700, Paul Eggert hath writ:
So I'll look into proposing a patch to the tz code to have it support this behavior. There are probably a few other places in the code that should also prefer "UT" or "UTC" to "GMT".
That's a good convention for tz to adopt. At the moment there are probably more bureaucrats stressing over aspects of this than ever in history. No answer can conform to the legal systems of the world because the answer depends on who you ask, which country you live in, which language you speak in that country. It gets very confusing and is definitely tl;dr, but see preprint 11-662 at http://www.cacr.caltech.edu/futureofutc/2011/preprints/index.html and preprint 13-505 at http://www.cacr.caltech.edu/futureofutc/preprints/index.html for some investigations into how confused the answer gets. -- Steve Allen <sla@ucolick.org> WGS-84 (GPS) UCO/Lick Observatory--ISB Natural Sciences II, Room 165 Lat +36.99855 1156 High Street Voice: +1 831 459 3046 Lng -122.06015 Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m
participants (7)
-
Clive D.W. Feather -
Matt Johnson -
Paul Eggert -
Paul_Koning＠dell.com -
Steve Allen -
Tim Parenti -
Tony Finch