Theory - proposal to delete the reference to population
References to population in tzcode2012b/Theory 1) "However, uninhabited ISO 3166 regions like Bouvet Island do not need locations, since local time is not defined there." 2) "Use UTC (with time zone abbreviation "zzz") for locations while uninhabited." ------- 1a) http://wikitravel.org/en/Bouvet_Island reports Olav Peak exists on Bouvet Island, without discussing a need for the peak. It is out of the scope of the tzdb as defined in the Theory file to decide on whether regions need locations. 1b) there is no definition of "uninhabited" German WP reports a team of people has been there after the 1970-01-01 cutoff point: http://de.wikipedia.org/wiki/Bouvetinsel#cite_note-6 "Vom 24. Dezember 1978 bis zum 8. März 1979 hielt sich ein norwegisches Forschungsteam auf der Insel auf und führte biologische und geologische Forschungsarbeiten durch." There are more sources supporting the claim that it had been populated. 1c) There is no evidence that local time is not defined if a population is absent. http://en.wikipedia.org/wiki/Yukon-Koyukuk_Census_Area,_Alaska is mostly uninhabited but the whole area is observing UTC-09:00 according to http://commons.wikimedia.org/wiki/File:Timezoneswest.PNG maybe covered by America/Anchorage Alaska Time but no further geographic description found HM as no zone, but http://www.timegenie.com/country.time/hm "To ensure accuracy of information, timegenie.com contacted the Australian Antarctic Division [http://www.antarctica.gov.au/] . On January 3, 2005, a senior official confirmed that Heard Island and McDonald Islands are UTC/GMT +5 hours." 2a) A location changes its UTC offset when people move out or in? http://en.wikipedia.org/wiki/Mount_Chamberlin_(Alaska) on UTC most of the year? A ship landing on an island, changes the UTC offset observed on that island? The time a meteorite lands on island depends on whether people have been there or not? 1+2) Violate tzcode2012b/Theory: ----- Scope of the tz database ----- The tz database attempts to record the history and predicted future of all computer-based clocks that track civil time. To represent this data, the world is partitioned into regions whose clocks all agree about time stamps that occur after the somewhat-arbitrary cutoff point of the POSIX Epoch (1970-01-01 00:00:00 UTC). 1+2) Are inconsistently applied. Uninhabited locations are treated differently, depending on whether they are located in countries that have been inhabited during all time between 1970-01-01 and 2012-05-12 or not. If Bouvet Island would be part of TF, it could be included in a zone but since it is covered by BV it is not. -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/
On 05/11/2012 06:37 PM, Tobias Conradi wrote:
It is out of the scope of the tzdb as defined in the Theory file ... 1b) there is no definition of "uninhabited" ... 1c) There is no evidence that local time is not defined if a population is absent.
The Theory file is not a formal definition; it's an attempt to write down common-sense rules about what's in the database. I suppose it might help to clarify it somewhat -- thus, for example, perhaps "uninhabited" might be clarified to have its common interpretation "no permanent inhabitants". I imagine it's possible to haggle over any such definition indefinitely, but we do have limited resources, and to be honest I'd rather leave it alone than haggle about it. Another way to think of it is that we have enough trouble worrying about locations containing permanent inhabitants, without also having to worry about transitory populations where data are even harder to come by and are more likely to be incorrect or in conflict. We're better off spending our limited resources in areas where the need is real and where we have real data.
German WP reports a team of people has been there after the 1970-01-01
There's a similar discussion in the English Wikipedia <http://en.wikipedia.org/wiki/Bouvet_Island#History>. But this was a temporary visit by a Norwegian research team, which doesn't count as permanent inhabitants.
On Sat, May 12, 2012 at 4:57 AM, Paul Eggert <eggert@cs.ucla.edu> wrote:
The Theory file is not a formal definition; it's an attempt to write down common-sense rules about what's in the database. I suppose it might help to clarify it somewhat -- thus, for example, perhaps "uninhabited" might be clarified to have its common interpretation "no permanent inhabitants". No inhabitant is permanent. But speaking of regions having population for more than X consecutive days is more robust. Even better delete any reference to population.
I imagine it's possible to haggle over any such definition indefinitely, but we do have limited resources, and to be honest I'd rather leave it alone than haggle about it. Deleting the definition is saving even more time. No question to that part of the Theory file anymore.
Another way to think of it is that we have enough trouble worrying about locations containing permanent inhabitants, without also having to worry about transitory populations where data are even harder to come by and are more likely to be incorrect or in conflict. Is there conflicting information for HM or BV, the two ISO 3166-1 codes which seem to be currently the only ones that are not mapped to any zone?
We're better off spending our limited resources in areas where the need is real and where we have real data. Do you think the need reported for a zone for HM during the last hours is not real?
Do you think the data reported for HM is not real data? -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/
On 05/11/2012 10:44 PM, Tobias Conradi wrote:
Do you think the need reported for a zone for HM during the last hours is not real?
Yes, that's correct. There does not seem to be a real need here. We haven't seen any need from permanent residents of Heard Island and McDonald Islands, or any real data about such residents.
On Sat, May 12, 2012 at 9:48 AM, Paul Eggert <eggert@cs.ucla.edu> wrote:
On 05/11/2012 10:44 PM, Tobias Conradi wrote:
Do you think the need reported for a zone for HM during the last hours is not real?
Yes, that's correct. Thanks for the confirmation.
We haven't seen any need from permanent residents of Heard Island and McDonald Islands, or any real data about such residents. And needs that don't come from permanent residents are not real?
If two ships enter territorial waters of HM would the Olson-Eggert-IANA time zone database give them advice for how to set their clocks? tzcode2012b/Theory says: ----- Scope of the tz database ----- The tz database attempts to record the history and predicted future of all computer-based clocks that track civil time. Not recording the offsets for HM ridicules the first sentence of the Scope section. -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/
Could I join the chorus of people asking for a separate mailing list for release announcements? Because I have absolutely no interest in conducting the historical research to catalog what time zone was observed by the 1971 French expedition to Heard Island, or engaging in a debate over whether something that is called a 'location' could ever also be called a 'place', or enumerating every inch of ground gained and lost in a civil war between factions that observed different time zones. I care about the actual setting of the clock in places, (locations, sites, venues, locales, and other spatial designations) where my software has a user base. I don't want to assert that an autistic compulsion to know the time everywhere for all of history is not a legitimate need; that is a political question beyond my pay grade, and ties into the broader question of "neurodiversity" - whether autism, schizophrenia, and other disintegrative conditions ought to be labeled as "disorders"T or merely "differences." Nevertheless, I'd like the ability to post letters without needing to enter into a long discussion of fine details of philately with a stamp collector. Moreover, the need to have a bright-line classification of all the details of time zone observance for all of history is bound to fail for reasons already stated: the world simply is a complex and ambiguous place, and the answers to the "edge cases" will be at best equivocal.
On Sat, May 12, 2012 at 3:37 AM, Tobias Conradi <tobias.conradi@gmail.com> wrote:
References to population in tzcode2012b/Theory 1) "However, uninhabited ISO 3166 regions like Bouvet Island do not need locations, since local time is not defined there."
According to https://github.com/valodzka/tzcode/commit/6ebeff7617961b81da63377669138e44a3... https://github.com/valodzka/tzcode/blame/6ebeff7617961b81da63377669138e44a3f... the above provision exists since 2005-08-29 tzcode2005m. The mailing list archive for the related month is at http://mm.icann.org/pipermail/tz/2005-August/thread.html Under "Request from CLDR committee" there is http://mm.icann.org/pipermail/tz/2005-August/013149.html ----------- Paul Eggert eggert at CS.UCLA.EDU Thu Aug 18 21:52:49 UTC 2005 .... I'll insert the word "inhabited" before "country"; that's simpler. ------------ -- Tobias Conradi Rheinsberger Str. 18 10115 Berlin Germany http://tobiasconradi.com/
participants (3)
-
Kevin Kenny -
Paul Eggert -
Tobias Conradi