To DST or not to DST. Currently the TZ DB has Tonga moving to DST on the 5th November. In 2016 they announced they were resuming DST, in the only reference to that I could quickly find <http://matangitonga.to/tag/daylight-saving?page=1>, does not state that this was for one year or whether the resumption was also for successive years. No reference for the change was added to the TZ DB. IATA it seems have taken it as being for a single year and so do not have DST starting this year. Given the discrepancy between IANA and IATA (besides the letter T), does any body have any authoritative/corroborating source for either position. /David
Lord Buddha wrote:
Given the discrepancy between IANA and IATA (besides the letter T), does any body have any authoritative/corroborating source for either position.
I don't have any information other than what you mentioned. It'd be helpful to have something authoritative. Quite possibly Tonga hasn't decided yet.
Indeed quite probably they have not decided yet. Waffling/Rambling... I know that is/has been the policy to try to find patterns/rules where none have been formally published by the authority for a timezone. Every/most years this causes airlines a hiccup for Tonga/Fiji. e.g. Actual flight time Auckland to Tonga is 2hrs 50mins. Up to November 5th, it is displayed as 2hrs 50mins. But from November 5th this is calculated/displayed as just 1hr 50mins using IANA TZ info. IATA who publish their own timezone updates to all IATA airlines based on authoritative communication from the governments concerned. IATA publishes that information as soon as announced, but IANA has not/does not until near the effective date unless some more important change forces a release. e.g. Fiji have changed when DST ends in 2018 and that has been published by IATA so all airline schedules are now based on that change, but no TZ DB is offically available yet. ( Have picked up your patch for that and used it to patch both JRE and JodaTime and that will go live next week hopefully for us ). Anyway, have emailed gov.to & tweeted to see if I can get an official position, but not too hopeful as, they just probably have not decided yet. On 19 September 2017 at 12:25, Paul Eggert <eggert@cs.ucla.edu> wrote:
Lord Buddha wrote:
Given the discrepancy between IANA and IATA (besides the letter T), does any body have any authoritative/corroborating source for either position.
I don't have any information other than what you mentioned. It'd be helpful to have something authoritative. Quite possibly Tonga hasn't decided yet.
On Sep 18, 2017, at 10:54 PM, Lord Buddha <lord.buddha@gmail.com<mailto:lord.buddha@gmail.com>> wrote: Indeed quite probably they have not decided yet. Waffling/Rambling... I know that is/has been the policy to try to find patterns/rules where none have been formally published by the authority for a timezone. Every/most years this causes airlines a hiccup for Tonga/Fiji. e.g. Actual flight time Auckland to Tonga is 2hrs 50mins. Up to November 5th, it is displayed as 2hrs 50mins. But from November 5th this is calculated/displayed as just 1hr 50mins using IANA TZ info. IATA who publish their own timezone updates to all IATA airlines based on authoritative communication from the governments concerned. IATA publishes that information as soon as announced, but IANA has not/does not until near the effective date unless some more important change forces a release. Is the issue the release schedule? It doesn't seem that way. It seems the issue is lack of authoritative information. Does IATA indeed have a source of authoritative information that isn't public? Or do they do what IANA does in the absence of authoritative information, i.e., guess? In other words, are we dealing simply with two different guesses that went in different directions? If IATA does indeed have authoritative information, it would be great if they would share that information (the source material, not just the conclusions). paul
On 09/19/2017 05:51 AM, Paul.Koning@dell.com wrote:
Does IATA indeed have a source of authoritative information that isn't public? Or do they do what IANA does in the absence of authoritative information, i.e., guess? In other words, are we dealing simply with two different guesses that went in different directions?
That's my guess, yes. (Couldn't resist....) I have privately emailed a question about this to the Tonga Consulate General in San Francisco.
Have got a reply back from the Tongan news site http://matangitonga.to/ They had a director that sits on the committee that decides this, but ultimately it goes the PM's office. They have said they want to write an article so that their readers know what is going on. This is a very good way of progressing this. As soon as they know we should find out, but from the papers view point, DST was not so popular. I will send them some info on why if they do, it is good for all to decide more than 2 weeks prior. Preferably the normal state should be a year in advance (for airlines). /David On 20 September 2017 at 04:34, Paul Eggert <eggert@cs.ucla.edu> wrote:
On 09/19/2017 05:51 AM, Paul.Koning@dell.com wrote:
Does IATA indeed have a source of authoritative information that isn't public? Or do they do what IANA does in the absence of authoritative information, i.e., guess? In other words, are we dealing simply with two different guesses that went in different directions?
That's my guess, yes.
(Couldn't resist....)
I have privately emailed a question about this to the Tonga Consulate General in San Francisco.
On Sep 20, 2017, at 12:48 AM, Lord Buddha <lord.buddha@gmail.com> wrote:
... I will send them some info on why if they do, it is good for all to decide more than 2 weeks prior. Preferably the normal state should be a year in advance (for airlines).
As I recall, IETF made that point rather explicitly some time ago, after the Turkey and/or Egypt DST debacle. It might not hurt to make it again. Perhaps have it issued over the signature of the chairman of the Internet Society, or something like that, to make it look more impressive to bureaucrats? paul
Paul.Koning@dell.com wrote:
As I recall, IETF made that point rather explicitly some time ago, after the Turkey and/or Egypt DST debacle.
Yes, the main tzdb page <https://www.iana.org/time-zones/repository/tz-link.html> now says, "If your government plans to change its time zone boundaries or daylight saving rules, let the mailing list know well in advance. With less than a year's notice there is a good chance that some computer-based clocks will operate incorrectly after the change, due to delays in propagating updates to software and data. The shorter the notice, the more likely clock problems will arise." It goes on to cite Matt Johnson's more-detailed blog post "On the Timing of Time Zone Changes" <http://codeofmatt.com/2016/04/23/on-the-timing-of-time-zone-changes/>. Perhaps the above info could be sent on the Tongan government, to give them an idea of the problems they'll face if they delay their decision further.
On 2017-09-20 09:55, Paul Eggert wrote:
Paul.Koning@dell.com wrote:
As I recall, IETF made that point rather explicitly some time ago, after the Turkey and/or Egypt DST debacle.
Yes, the main tzdb page <https://www.iana.org/time-zones/repository/tz-link.html> now says, "If your government plans to change its time zone boundaries or daylight saving rules, let the mailing list know well in advance. With less than a year's notice there is a good chance that some computer-based clocks will operate incorrectly after the change, due to delays in propagating updates to software and data. The shorter the notice, the more likely clock problems will arise." It goes on to cite Matt Johnson's more-detailed blog post "On the Timing of Time Zone Changes" <http://codeofmatt.com/2016/04/23/on-the-timing-of-time-zone-changes/>.
Perhaps the above info could be sent on the Tongan government, to give them an idea of the problems they'll face if they delay their decision further.
Please also consider adding an explicit request to add *tz@iana.org, as this is the major global coordinator and source of updates to systems and devices*, to their notification list which informs the media, airlines, IATA, and others of time zone and daylight saving changes. This information is somewhat buried in the recommendations at the end of Matt's post, insufficiently emphasized and explained IMHO (it is mentioned as "the tz discussion list" not ...global coordinator and source... which may not seem significant to bureacrats, politicians, or IT management). -- Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada
Lord Buddha wrote:
Have got a reply back from the Tongan news site http://matangitonga.to/
They had a director that sits on the committee that decides this, but ultimately it goes the PM's office. They have said they want to write an article so that their readers know what is going on. This is a very good way of progressing this.
As soon as they know we should find out, but from the papers view point, DST was not so popular.
Have you heard anything new about this? We will be issuing a new tzdb release soon because of recently-announced changes that take effect in a week or two in Northern Cyprus and Sudan, and it'd be nice to get Tonga squared away while we're at it. Last year, Tonga did not announce its decision until October 27 (just ten days' notice). If DST was not so popular last year, perhaps we should change tzdb to guess that Tonga will not use DST this year either. We can always change it back if this new guess is incorrect.
I have not heard anything. Was not so optimistic, but at least info was provided to the contact. In August government was disolved by the King. The current prime minister continued in office in care taker mode. It is easy to see that few decisions will be made until elections 16th November. I can't say what the right choice is. IATA still believe there will not no DST this year. For me, because of this difference, I end up with flight durations incorrectly calculated (according to IATA). On the JIRA ticket for this I just set to "blocked" as I have no clear correct answer. If IANA switched to no DST, it would make my flight times (after software deployments) correct, but at the same time, it is already pretty late to be changing this. Have informed the contact that currently DST will happen on computers/mobiles. Its really in their court to state early enough that this isn't what they want and do something about it. Will send you the email I have direct... Not really right to put in on this list. /David On 18 October 2017 at 22:09, Paul Eggert <eggert@cs.ucla.edu> wrote:
Lord Buddha wrote:
Have got a reply back from the Tongan news site http://matangitonga.to/
They had a director that sits on the committee that decides this, but ultimately it goes the PM's office. They have said they want to write an article so that their readers know what is going on. This is a very good way of progressing this.
As soon as they know we should find out, but from the papers view point, DST was not so popular.
Have you heard anything new about this? We will be issuing a new tzdb release soon because of recently-announced changes that take effect in a week or two in Northern Cyprus and Sudan, and it'd be nice to get Tonga squared away while we're at it.
Last year, Tonga did not announce its decision until October 27 (just ten days' notice). If DST was not so popular last year, perhaps we should change tzdb to guess that Tonga will not use DST this year either. We can always change it back if this new guess is incorrect.
On 10/18/2017 02:36 AM, Lord Buddha wrote:
I can't say what the right choice is.
Thanks for the background info. From what you write, it seems more likely that Tonga will not observe DST this year. So I'll shortly send out an email containing a patch along those lines, which I hope can make it into the next tzdb release.
participants (4)
-
Brian Inglis -
Lord Buddha -
Paul Eggert -
Paul.Koning@dell.com