RE: UTC as basis for time legislation
... But other than that, if you say GMT you mean the same time as if you say UTC.
Uh... you like waving red flags in front of bulls, don't you? Or maybe walking up to a bee hive and whacking it with a short stick is your idea of a good time. :)
Ouch. My apologies to the hornets. I did some reading and I think I now understand some of the subtleties involved. Let me restate the point I should have made and hopefully not disturb another beehive. The tzdata uses an approximation of the subtle details of timekeeping. With those approximations, GMT and UTC end up being described both as having a constant 0 offset. So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes. Is that correct? paul
On 2011 Sep 23, at 18:44, <Paul_Koning@Dell.com> wrote:
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes.
Is that correct?
It is correct under ITU-R TF.460-6. It will not remain true if the draft revision to that document is approved in its current form at the Radiocommunication Assembly next January. Alternatively, the emeritus director of the BIPM has suggested that the ITU-R should relinquish the definition of UTC to the CGPM, and the result of that option is even less clear. -- Steve Allen <sla@ucolick.org> WGS-84 (GPS) UCO/Lick Observatory Natural Sciences II, Room 165 Lat +36.99855 University of California Voice: +1 831 459 3046 Lng -122.06015 Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m
On Sat, Sep 24, 2011 at 3:53 AM, Steve Allen <sla@ucolick.org> wrote:
On 2011 Sep 23, at 18:44, <Paul_Koning@Dell.com> wrote:
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes.
Is that correct?
It is correct under ITU-R TF.460-6.
http://www.itu.int/dms_pubrec/itu-r/rec/tf/R-REC-TF.460-6-200202-I!!PDF-E.pd... has no reference to GMT. Also I don't understand why tzdata approximation is involved. I would assume that legal GMT in the UK in the year 2011 is exactly the same as UTC. I could not find a law for that. Maybe an older legal definition would make GMT the same as UT0. ITU-R TF.460-6: "UT0 is the mean solar time of the prime meridian obtained from direct astronomical observation;" Can anyone give a UK law relating GMT to UTC? -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/tobias_conradi
Tobias Conradi <tobias.conradi@gmail.com> wrote:
Also I don't understand why tzdata approximation is involved.
The approximation is not in tzdata it is in the definition of civil time, which as Clive illustrated is often not clearly defined.
I would assume that legal GMT in the UK in the year 2011 is exactly the same as UTC. I could not find a law for that.
No, the law here says legal time is GMT, but most official time signals are UTC. There have been a few unsuccessful attempts to deal with the mismatch between de jure and de facto UK time, e.g. http://www.publications.parliament.uk/pa/ld199798/ldhansrd/vo970611/text/706... Tony. -- f.anthony.n.finch <dot@dotat.at> http://dotat.at/ Shannon, Rockall, Malin: Southwest 4 or 5, occasionally 6 in Rockall and Malin, backing south 5 to 7 later, perhaps gale 8 later. Moderate or rough. Showers, rain later. Moderate or good, occasionally poor later.
On Mon, Sep 26, 2011 at 12:29 PM, Tony Finch <dot@dotat.at> wrote:
Tobias Conradi <tobias.conradi@gmail.com> wrote:
I would assume that legal GMT in the UK in the year 2011 is exactly the same as UTC. I could not find a law for that.
No, the law here says legal time is GMT, but most official time signals are UTC. There have been a few unsuccessful attempts to deal with the mismatch between de jure and de facto UK time, e.g. http://www.publications.parliament.uk/pa/ld199798/ldhansrd/vo970611/text/706...
So there is one flavor of GMT that is UTC and there is legal GMT which is something different. 2011i/tzdata equates GMT for 2011 with UTC. I would now reply to the former text: On Sat, Sep 24, 2011 at 3:53 AM, Steve Allen <sla@ucolick.org> wrote:
On 2011 Sep 23, at 18:44, <Paul_Koning@Dell.com> wrote:
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes.
Is that correct?
It is correct under ITU-R TF.460-6. To my understanding the statement of country A is ambiguous. Whether there is approximation in tzdata depends on how one interprets the the statement of country A. This has nothing to do with ITU-R TF.460-6.
t -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/tobias_conradi
Tobias Conradi said:
No, the law here says legal time is GMT, but most official time signals are UTC. There have been a few unsuccessful attempts to deal with the mismatch between de jure and de facto UK time, e.g. http://www.publications.parliament.uk/pa/ld199798/ldhansrd/vo970611/text/706...
So there is one flavor of GMT that is UTC and there is legal GMT which is something different.
Yes and no. Legal time in the UK is Greenwich mean time (or that plus one hour). That means mean solar time at Greenwich, which is equal to UT1 within a few milliseconds (it depends on exactly what you mean by "mean" and exactly where "Greenwich" is). The easiest-to-find sources of time in the UK are all UTC. To actually get your hands on GMT you need a source of DUT and to do the subtraction. That does *not* mean that "there is one flavor of GMT that is UTC". I've not found a reported law case where the difference is explored, let alone where it mattered. The closest I got was a case hinging on a time difference of 8 seconds between two events, but there was no mention of which of GMT or UTC applied.
2011i/tzdata equates GMT for 2011 with UTC.
In effect, yes. It takes the view that the difference is too small to matter *in this context*.
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements.
To my understanding the statement of country A is ambiguous. Whether there is approximation in tzdata depends on how one interprets the the statement of country A.
No. The statement of country A is *not* ambigous (except at the millisecond level as mentioned above). The statements are technically and de jure different. Whether country A actually meant UTC when it wrote GMT is a separate question that you'd have to ask country A.
This has nothing to do with ITU-R TF.460-6.
But the proposed amendment to it would mean that "too small to matter in this context" will cease to be true eventually. -- 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 Mon, Sep 26, 2011 at 4:13 PM, Clive D.W. Feather <clive@davros.org> wrote:
Tobias Conradi said:
No, the law here says legal time is GMT, but most official time signals are UTC. There have been a few unsuccessful attempts to deal with the mismatch between de jure and de facto UK time, e.g. http://www.publications.parliament.uk/pa/ld199798/ldhansrd/vo970611/text/706...
So there is one flavor of GMT that is UTC and there is legal GMT which is something different.
Yes and no. Which part you refer to by "no"?
2011i/tzdata equates GMT for 2011 with UTC.
In effect, yes. It takes the view that the difference is too small to matter *in this context*. I personally would rather not guess what view it has, but stick to what it does.
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements.
To my understanding the statement of country A is ambiguous. Whether there is approximation in tzdata depends on how one interprets the the statement of country A.
No. The statement of country A is *not* ambigous (except at the millisecond level as mentioned above). Since there are different types of GMT it is. There is the legal GMT which is ambiguous in itself and the BBC-GMT.
The statements are technically and de jure different. Yes, the first is ambiguous.
Whether country A actually meant UTC when it wrote GMT is a separate question that you'd have to ask country A.
This has nothing to do with ITU-R TF.460-6.
But the proposed amendment to it would mean that "too small to matter in this context" will cease to be true eventually. GMT is not mentioned in the document http://www.itu.int/dms_pubrec/itu-r/rec/tf/R-REC-TF.460-6-200202-I!!PDF-E.pd...
Also, since there is a time span when GMT was legally defined different from UTC, I think an ITU document cannot change that retroactively. In the linked document I also see no change of definition of UTC. -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com
Steve Allen said:
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes. Is that correct?
It is correct at present. tzdata doesn't bother with the difference.
It is correct under ITU-R TF.460-6. It will not remain true if the draft revision to that document is approved in its current form at the Radiocommunication Assembly next January.
If this happens, then GMT and UTC will start to drift apart. The point at which the tzdata files will take notice of this has not yet been decided (or discussed, I think). -- 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
Though unless I'm misunderstanding the issue, the existence of the right/ part of the database suggests that zoneinfo can accomodate without a code change, and that'll buy time to externalize the new needed knowledge into something like djb's /etc/leapsecs.dat, yes?
Bennett Todd <bet@rahul.net> wrote:
Though unless I'm misunderstanding the issue, the existence of the right/ part of the database suggests that zoneinfo can accomodate without a code change, and that'll buy time to externalize the new needed knowledge into something like djb's /etc/leapsecs.dat, yes?
No, because the "right" TAI-based tables are incompatible with POSIX. POSIX time is essentially civil time and if the basis of civil time changes then POSIX will follow. If GMT and UTC become separated so that we can no longer fudge the definition of civil time, the disambiguation will have to be resolved by the various civil authorities. Tony. -- f.anthony.n.finch <dot@dotat.at> http://dotat.at/ Viking: West 5 to 7, occasionally gale 8 at first, backing south 5 or 6. Rough becoming moderate. Squally showers, rain later. Moderate or good.
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes. Is that correct?
It is correct at present. tzdata doesn't bother with the difference.
It is correct under ITU-R TF.460-6. It will not remain true if the draft revision to that document is approved in its current form at the Radiocommunication Assembly next January.
If this happens, then GMT and UTC will start to drift apart. The point at which the tzdata files will take notice of this has not yet been decided (or discussed, I think).
Because GMT is roughly UT1 while UTC is atomic clock time, and without leap seconds the two drift apart -- right? tzdata gives offsets in units of minutes (no support for fractions of a minute -- see for example the comment on Amsterdam Mean Time in the "europe" file). So I guess we'd be good for half a century or so. Or perhaps we'll see the introduction of leap minutes (or leap hours) to replace the former leap seconds... Or "GMT" might end up being legally redefined as meaning UTC rather than (roughly) UT1. paul
On 2011 Sep 26, at 07:18, <Paul_Koning@Dell.com> wrote:
Because GMT is roughly UT1 while UTC is atomic clock time, and without leap seconds the two drift apart -- right?
Right, and that would create opportunity for legal tests of the sort which have never before been explored. Islamic countries in particular might find the new UTC to be a problem, but the official objections to change have been from UK, China, Canada.
tzdata gives offsets in units of minutes (no support for fractions of a minute -- see for example the comment on Amsterdam Mean Time in the "europe" file). So I guess we'd be good for half a century or so. Or perhaps we'll see the introduction of leap minutes (or leap hours) to replace the former leap seconds... Or "GMT" might end up being legally redefined as meaning UTC rather than (roughly) UT1.
tzdata and tzcode in their current form can already keep a zone aligned with GMT/UT1 to the nearest second. That's shown here http://www.ucolick.org/~sla/leapsecs/right+gps.html although the cumbersomeness of the current representation might trigger desire to streamline the rules. -- Steve Allen <sla@ucolick.org> WGS-84 (GPS) UCO/Lick Observatory Natural Sciences II, Room 165 Lat +36.99855 University of California Voice: +1 831 459 3046 Lng -122.06015 Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m
On Mon, Sep 26, 2011 at 4:18 PM, <Paul_Koning@dell.com> wrote:
So if country A says its legal time is an hour ahead of GMT and B says they are an hour ahead of UTC, those are technically different statements. However, to the approximation of tzdata, the outcome is the same: both country A and B would be shown in the tzdata file as having an offset of 60 minutes. Is that correct?
It is correct at present. tzdata doesn't bother with the difference.
It is correct under ITU-R TF.460-6. It will not remain true if the draft revision to that document is approved in its current form at the Radiocommunication Assembly next January.
If this happens, then GMT and UTC will start to drift apart. The point at which the tzdata files will take notice of this has not yet been decided (or discussed, I think).
Depending on GMT definition, to my understanding they are already apart.
Because GMT is roughly UT1 while UTC is atomic clock time, and without leap seconds the two drift apart -- right?
To my understanding that is the reason that legal GMT and UTC did already diverge.
tzdata gives offsets in units of minutes (no support for fractions of a minute -- see for example the comment on Amsterdam Mean Time in the "europe" file). The file reads:
# Amsterdam Mean Time was +00:19:32.13 exactly, but the .13 is omitted # below because the current format requires GMTOFF to be an integer. # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Europe/Amsterdam 0:19:32 - LMT 1835 0:19:32 Neth %s 1937 Jul 1 -------- That means full seconds. -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/
On Mon, Sep 26, 2011 at 4:33 PM, Tobias Conradi <tobias.conradi@gmail.com> wrote:
On Mon, Sep 26, 2011 at 4:18 PM, <Paul_Koning@dell.com> wrote:
Because GMT is roughly UT1 while UTC is atomic clock time, and without leap seconds the two drift apart -- right?
To my understanding that is the reason that legal GMT and UTC did already diverge. I was referring to "GMT is roughly UT1 while UTC is atomic clock time".
For the leap second http://tf.nist.gov/pubs/bulletin/leapsecond.htm " Its purpose is to keep the UTC time scale within ±0.9 s of the UT1 astronomical time scale" -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com
participants (6)
-
Bennett Todd -
Clive D.W. Feather -
Paul_Koning＠Dell.com -
Steve Allen -
Tobias Conradi -
Tony Finch