Re: Request to Update tz Database for Manitoba Permanent Daylight Saving Time Change
Paul Eggert wrote:
If anyone has contacts with the Manitoba government that could help speed up that process, it'd be appreciated.
Although I am employed by the Government of Manitoba, I do not have any direct contact with the Premier’s or the Lieutenant Governor’s offices. I am also anxiously awaiting the OiC to make this change official. I have heard rumours that an Order will come out sometime this week, but nothing more solid than that. However, speaking ONLY as a citizen living in Manitoba, I see a problem with Tim Parenti’s commit cd70627 to eggert/tz. While CDT and EST are the same thing mathematically, they are vastly different things linguistically, legally, socially and politically. From <theory.html>, “Use abbreviations that are in common use among English-speakers […] CST/CDT/CWT/CPT Central [North America] […]”. CDT is the abbreviation in common use here in English; I have never heard anyone refer to local Manitoba time as EST during CDT. The published material from the government so far all refers to permanent Daylight Savings Time, and the expectation is that we’ll be on CDT for the foreseeable future, not EST. If the Order in Council (which I expect) explicitly refers to “Central Daylight Time”, then “EST” might also be legally inaccurate, but we’ll have to wait and see. If the 2023 Act gets Proclaimed instead, then it’ll be a new time zone “Manitoba Standard Time”, which also does not abbreviate to EST, it will abbreviate – with the force of law – to “MBT”. We’ll all find out which route the government will take shortly, and everyone hopes they do it sooner rather than later! -Adam Adam Thompson Consultant, Infrastructure Services [MERLIN] 100 - 135 Innovation Drive Winnipeg, MB R3T 6A8 (204) 977-6824 or 1-800-430-6404 (MB only) 💻https://www.merlin.mb.ca 🤳Follow MERLIN on Social Media
On Tue, 29 Sept 2026 at 10:24, Adam Thompson via tz <tz@iana.org> wrote:
While CDT and EST are the same thing mathematically, they are vastly different things linguistically, legally, socially and politically.
Our approach here mirrors that which we've taken with the other provinces and territories which have ended seasonal time changes this year, and reflects a long-standing norm to interpret year-round time standards as standard time. This fits legally as well, since the Act does not institute "permanent DST", but rather redefines a year-round standard time one hour ahead of the previous standard time. Additionally, EST matches existing and recent usage for year-round UTC−5 in North America — currently in Atikokan, Ontario as well as in Cancún and Panama, but also in Indiana until 2006. From <theory.html>, *“Use abbreviations that are in common use among
English-speakers […] CST/CDT/CWT/CPT Central [North America] […]”*. *CDT *is the abbreviation in common use here in English; I have never heard anyone refer to local Manitoba time as EST during CDT.
While that is a data point, seasonal usage of such nomenclature from before this change is not, itself, hard evidence of continued usage after the change. Specifically, the language you cite from theory.html exists to remind us of our archival role, and that we should follow *de facto* usage for abbreviations in a way that tends to lag, not lead. Amidst both changing and unchanging neighbors, though, predicting how future usage will shake out unavoidably results in circular logic either way, so we look in the meantime to both recent and nearby practice as laid out above, which is how we've landed on MST/CST/EST for now instead of PDT/MDT/CDT. There was similar discussion regarding British Columbia which has recently adopted a preferred abbreviation for official/governmental use, but for which it's also too early to tell what the general public will use: https://lists.iana.org/hyperkitty/list/tz@iana.org/thread/Y432QXUMA5SJDZTRLB... In short, it's very early days for all of these abbreviations. That said, the social difference between what is mathematically a zone shift and what is culturally less so is certainly acknowledged. It absolutely DOES strain the traditional nomenclature to extend "Mountain" to the west coast, "Central" into the mountains, and "Eastern" into the central plains. So, as described in the BC thread, we will continue to monitor the abbreviations that become adopted by the general public as the effects of these changes percolate throughout the affected regions and their neighbors.
it’ll be a new time zone “Manitoba Standard Time”, which *also* does not abbreviate to EST, it will abbreviate – with the force of law – to “MBT”.
Even if the new province-specific abbreviations do end up catching on, they would still create new problems for areas near provincial and zonal boundaries: Areas in southeastern BC which have chosen year-round UTC−6 remain covered by America/Edmonton, but should a location outside of Alberta get the abbreviation "ABT"? Much of the rest of BC is covered by America/Vancouver where the official guidance for UTC−7 is now "PCT", but this is identical to Creston and Fort Nelson's "MST" which continue to use that designation according to that same guidance, as well as to the "MST" we've used for Yukon since 2020 without issue. Likewise, using "MBT" for areas in northwestern Ontario that end up following Winnipeg's clocks could equally be a non-starter. So, since there is a lot we still don't know, this is why we haven't deviated from the traditional nomenclature for the time being. -- Tim Parenti
My 0.02: Using EST (static UTC-5) only slightly bothers me. What bothers me more is the tight deadline to get any software updates done and out there. If EST speeds up the TZ database update process and uses past practices, I'm for it. I'm new to the IANA mailing list and all this is new to me, but I'd assume the debate of whether to use "MBT" or some other convention can be opened up later after things are settled. I'd assume that there's not incredible pain in having multiple timezone definitions for the same UTC offset as well but happy to get educated.
james.epp--- wrote:
Using EST (static UTC-5) only slightly bothers me. What bothers me more is the tight deadline to get any software updates done and out there. [...]
AFAICT, the release of version 2026e was not delayed by arguments over the three-letter abbreviation for Manitoba’s new zone, despite how it may appear from the high volume of list traffic on this topic. Rather, it was delayed by the lack of evidence of signed legislation making the time change official. This is required, as stated before, and doubly so because of other recent “almost” changes that were rescinded before becoming official. -- Doug Ewell, CC, ALB | Lakewood, CO, US | ewellic.org
On 10/3/2026 2:17 PM, Doug Ewell via tz wrote:
AFAICT, the release of version 2026e was not delayed by arguments over the three-letter abbreviation for Manitoba’s new zone, despite how it may appear from the high volume of list traffic on this topic.
Rather, it was delayed by the lack of evidence of signed legislation making the time change official. This is required, as stated before, and doubly so because of other recent “almost” changes that were rescinded before becoming official.
Even that "delay" was a published change 12 days after the initial public announcement of intent on the radio (and then in a press release) and /_within a day of the actual effective date becoming public_/. I don't like using the word "delay" with that short of a time line.
James Bellaire wrote:
Rather, it was delayed by the lack of evidence of signed legislation [...]
Even that "delay" was a published change 12 days after the initial public announcement of intent on the radio (and then in a press release) and _within a day of the actual effective date becoming public._
I don't like using the word "delay" with that short of a time line.
Well, I didn’t say the interval (better word?) was the fault of the tzdb maintainers. Indeed, the responsiveness of getting that release out once the requirement was met was laudable. There seems to be a perception among the public that the moment a government official says “Yeah, it looks like we’ll probably do this,” tzdb will instantly be updated, not only on the IANA website but magically on everyone’s devices as well. The “delay” perceived by such public would be between the initial policy announcement on September 17 and the release of 2026e on September 30. So, well over 90% of that “delay” was on the Manitoba government. And, of course, none of that would have mattered if the time change were legislated to go into effect a year from now, which would have been the responsible thing to do, instead of a month from now. -- Doug Ewell, CC, ALB | Lakewood, CO, US | ewellic.org
On 10/3/2026 3:03 PM, Doug Ewell via tz wrote:
The “delay” perceived by such public would be between the initial policy announcement on September 17 and the release of 2026e on September 30.
Most of the public has no clue tzdata exists. They rely on the downstream programmers to keep application updates available. They don't see the process behind the curtain.
On Sat, 3 Oct 2026 at 15:20, James Bellaire via tz <tz@iana.org> wrote:
On 10/3/2026 3:03 PM, Doug Ewell via tz wrote:
The “delay” perceived by such public would be between the initial policy announcement on September 17 and the release of 2026e on September 30.
Most of the public has no clue tzdata exists. They rely on the downstream programmers to keep application updates available. They don't see the process behind the curtain.
Agreed. Though the inherent lag is felt by the intermediate vendors downstream and contributes to the length of the overall data lifecycle observed by endusers. Where we've been aware of pending changes this year, Paul and I have tried to be proactive in reaching out to various jurisdictions and let them know what kind of information we need to most effectively and proactively document their changes — namely in RDEK, Northwest Territories, and Nunavik. The main reason we hadn't reached out to Manitoba any earlier was from the (apparently mistaken) assumption that their much longer public engagement survey period than others' equated to any sort of soft commitment that they wouldn't be rushing to make a change this year. For what it's worth, once Manitoba did make their policy announcement, I called and emailed the very next day; however, I got no direct response from those connections. (Admittedly, though, by that point, if they'd taken the time to respond to my inquiries with status updates, it may have only slowed things down.) -- Tim Parenti
<<On Sat, 3 Oct 2026 15:19:23 -0400, James Bellaire via tz <tz@iana.org> said:
On 10/3/2026 3:03 PM, Doug Ewell via tz wrote:
The “delay” perceived by such public would be between the initial policy announcement on September 17 and the release of 2026e on September 30.
Most of the public has no clue tzdata exists. They rely on the downstream programmers to keep application updates available. They don't see the process behind the curtain.
This seems as good a spot in this discussion to make this point: "Most of the public" includes nearly all politicians and government officials. Most people's mental model, as seen in politicians' messaging the past few months, is that if your clock is wrong, you don't install a software update, you "simply" change the clock. It's convenient if the clock "changes itself" for you, but if it doesn't, you do the same thing as your grandparents did. Hence all the communication has been "stop doing something you have always had to do" (i.e., "stop changing the clocks") rather than "change the time zone the clock thinks it's in". Only programmers and scientists have any notion of UTC offsets, let alone the sort of rules computers use to know when and how to change those offsets (i.e., what we do here). So the politicians don't think in terms of software vendors distributing updates, they think of *constituents no longer needing to reset the clock on the microwave oven*. -GAWollman
On Mon, 5 Oct 2026 at 13:04, Garrett Wollman via tz <tz@iana.org> wrote:
So the politicians don't think in terms of software vendors distributing updates, they think of *constituents no longer needing to reset the clock on the microwave oven*.
As we discuss on the other thread a potential "FAQ/primer for government officials" and ways to increase its visibility… https://lists.iana.org/hyperkitty/list/tz@iana.org/message/4LSU3NBLPI47QCRPI... Whatever we might end up putting together may well start with some version of: Q. Don't computers and phones mostly handle this themselves nowadays? A. Well, yes, but how do you suppose it is that they know what to do and when? ;) -- Tim Parenti
<<On Mon, 5 Oct 2026 18:28:46 -0400, Tim Parenti <tim@timtimeonline.com> said:
On Mon, 5 Oct 2026 at 13:04, Garrett Wollman via tz <tz@iana.org> wrote:
So the politicians don't think in terms of software vendors distributing updates, they think of *constituents no longer needing to reset the clock on the microwave oven*.
Whatever we might end up putting together may well start with some version of: Q. Don't computers and phones mostly handle this themselves nowadays? A. Well, yes, but how do you suppose it is that they know what to do and when? ;)
And many of them probably think "someone at Bell/Rogers/AT&T/Telstra/Vodafone pushes a button". I think, more to the point, you need to explain why it's important that computers and phones in Manitoba know what time it is in England. That's where the disconnect lies. To us it's a matter of course that computers need to communicate times in UTC *with other devices*, but that's not visible to casual end-users. -GAWollman
On 10/5/2026 6:28 PM, Tim Parenti via tz wrote:
On Mon, 5 Oct 2026 at 13:04, Garrett Wollman via tz <tz@iana.org> wrote:
So the politicians don't think in terms of software vendors distributing updates, they think of *constituents no longer needing to reset the clock on the microwave oven*.
As we discuss on the other thread a potential "FAQ/primer for government officials" and ways to increase its visibility… https://lists.iana.org/hyperkitty/list/tz@iana.org/message/4LSU3NBLPI47QCRPI...
Whatever we might end up putting together may well start with some version of: Q. Don't computers and phones mostly handle this themselves nowadays? A. Well, yes, but how do you suppose it is that they know what to do and when? ;)
I did a quick search for "how do I change the time zone for my province" and Google linked to the US DOT. (The results were better when I changed it to "my Canadian province" with a link to National Research Council Canada.) The general answer was "an act of government or legislation" ... not a reference to the mechanics behind propagating those changes to electronics. BTW: The US DOT website states in part: [1] /A typical time zone rulemaking to change one county normally takes between 6 months and 1 year to complete./ /If the decision is made to change the time zone boundary, DOT attempts to make the change effective at the next changeover to or from Daylight Saving Time, whichever is appropriate, to minimize confusion and disruption. / When official government policy is less than a year it is an uphill battle. [1] https://www.transportation.gov/regulations/procedure-moving-area-one-time-zo...
I think I've mentioned this before. Why not put a notice on the main IANA TzDb page in BIG RED LETTERS, something to the effect: ATTENTION lawmakers and timekeeping authorities: Changes to time zone rules do not update immediately. It is strongly recommended that any change to time zones is to take effect at least one year from the date the law or decision is officially finalized. This could link to a page with more detailed explanation: ... Unlike the impression that everything on the internet updates instantly the mechanisms for time zones and timekeeping must be updated and distributed to all the systems and applications. .... Cell phones, operating systems, databases... This takes time because each system update must be tested and distributed ... TzDb is the primary source... Maintainers of each system must incorporate the TzDb data and produce tested updates for each system ... CLDR ... and deploy those changes to all systems. ... To avoid difficulties in airline and train schedules, financial transactions, meetings, school hours ,,, and so on ... You all can write it better than my quick draft. But this seems to me the place where this sort of warning should appear. On 2026-10-03 03:03 PM, Doug Ewell via tz wrote:
James Bellaire wrote:
Rather, it was delayed by the lack of evidence of signed legislation [...] Even that "delay" was a published change 12 days after the initial public announcement of intent on the radio (and then in a press release) and _within a day of the actual effective date becoming public._
I don't like using the word "delay" with that short of a time line. Well, I didn’t say the interval (better word?) was the fault of the tzdb maintainers. Indeed, the responsiveness of getting that release out once the requirement was met was laudable.
There seems to be a perception among the public that the moment a government official says “Yeah, it looks like we’ll probably do this,” tzdb will instantly be updated, not only on the IANA website but magically on everyone’s devices as well.
The “delay” perceived by such public would be between the initial policy announcement on September 17 and the release of 2026e on September 30. So, well over 90% of that “delay” was on the Manitoba government. And, of course, none of that would have mattered if the time change were legislated to go into effect a year from now, which would have been the responsible thing to do, instead of a month from now.
-- Doug Ewell, CC, ALB | Lakewood, CO, US | ewellic.org
On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz <tz@iana.org> wrote:
Why not put a notice on the main IANA TzDb page in BIG RED LETTERS
Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.) We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative: https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE... Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful. That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with. -- Tim Parenti
We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to. Note that for Unicode (CLDR) there are the additional constraints imposed by names. We have one 'generic' name for time zones that don't have daylight savings, and three for ones that do; and in ~100 different languages, plus acronyms (where they are widely recognized). And it is harder for implementations to update names x languages than it is for the TZDB data we're based on. It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries. Mark On Sat, Oct 3, 2026 at 1:25 PM Tim Parenti via tz <tz@iana.org> wrote:
On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz <tz@iana.org> wrote:
Why not put a notice on the main IANA TzDb page in BIG RED LETTERS
Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.)
We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative:
https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE...
Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful.
That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with.
-- Tim Parenti
Hi I am a business consultant based in Victoria, BC, Canada. I became aware of an impending problem when our provincial government decided to ‘pull the switch’ and ditch Standard Time early this year. I was curious how it all worked and what had to change behind the scenes. My knowledge of the processes involving CLDR and tzData at the time was zero – like most people not ‘in the know’. I have learned a lot following this discussion over the last six months or so and the more I learn the more concerned I get about how the real world is going to cope. Microsoft has only recently pushed Windows Update 26H2, the first that contains the new BC time zone – and so far only to some streams. It will be late October before most users have the update. Nothing about it tells users they have anything else to do and IT managers seem oblivious to the problem in the main. Users have to manually change their time zone. Having 26H2 is not enough – you are still on Pacific Time US & Canada, which maps to America/lLosAngeles and will fall back in November. Microsoft also made available a paper pointing out that Outlook and Teams (and, by implication, lots of other similar non-MS software) sets events in UTC at the time the event is set, using the then-current local time zone. So, any meetings set up for after the next clock change are already wrong – and users do not even know that. This is orders of magnitude worse than Y2K – which was reasonably dealt with by IT staff and basically did not really involve end users at all. This is different. It most certainly DOES involve end users – and they are basically not aware of an impending issue. This already involves any future-dated time events that have been set while on the old time zone. That this involves most every timetable is quite scary. Yes airlines, for example, plan their slots based on UTC, but those map back to where they were expected to be when the timetable was set up. I don’t know what the solution is, but I know that the experts in IANA and those dealing with CLDR, as well as the suppliers like Microsoft etc, will all be held at fault if (when) the s*** hits the fan. Personally, I think a formal press release should be issued, ideally signed by all the above, drawing PUBLIC attention to the impending crisis. The fix will need to come from end user IT managers and database/application teams. So far, everyone seems to be sleepwalking into this without a care. Peter Aggus From: Mark Davis Ⓤ via tz <tz@iana.org> Sent: October 3, 2026 2:00 PM To: Tim Parenti <tim@timtimeonline.com> Cc: tz@iana.org Subject: [tz] Re: Suggestion: ATTENTION lawmakers and timekeeping authorities We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to. Note that for Unicode (CLDR) there are the additional constraints imposed by names. We have one 'generic' name for time zones that don't have daylight savings, and three for ones that do; and in ~100 different languages, plus acronyms (where they are widely recognized). And it is harder for implementations to update names x languages than it is for the TZDB data we're based on. It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries. Mark On Sat, Oct 3, 2026 at 1:25 PM Tim Parenti via tz <tz@iana.org <mailto:tz@iana.org> > wrote: On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz <tz@iana.org <mailto:tz@iana.org> > wrote: Why not put a notice on the main IANA TzDb page in BIG RED LETTERS Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.) We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative: https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE... Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful. That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with. -- Tim Parenti
Hi there, Respectfully, this is a technical group whose interests are making sure time is correctly calculated and thereby accurately reflected within technical platforms moving forward. Utilizing other knowledge sources or media organizations to voice concerns about vigilance in updating systems may be more a more effective way to communicate disruptive changes. Best, Jake On Saturday, 3. October 2026 23:43:50 (+02:00), peter--- via tz wrote: Hi I am a business consultant based in Victoria, BC, Canada. I became aware of an impending problem when our provincial government decided to ‘pull the switch’ and ditch Standard Time early this year. I was curious how it all worked and what had to change behind the scenes. My knowledge of the processes involving CLDR and tzData at the time was zero – like most people not ‘in the know’. I have learned a lot following this discussion over the last six months or so and the more I learn the more concerned I get about how the real world is going to cope. Microsoft has only recently pushed Windows Update 26H2, the first that contains the new BC time zone – and so far only to some streams. It will be late October before most users have the update. Nothing about it tells users they have anything else to do and IT managers seem oblivious to the problem in the main. Users have to manually change their time zone. Having 26H2 is not enough – you are still on Pacific Time US & Canada, which maps to America/lLosAngeles and will fall back in November. Microsoft also made available a paper pointing out that Outlook and Teams (and, by implication, lots of other similar non-MS software) sets events in UTC at the time the event is set, using the then-current local time zone. So, any meetings set up for after the next clock change are already wrong – and users do not even know that. This is orders of magnitude worse than Y2K – which was reasonably dealt with by IT staff and basically did not really involve end users at all. This is different. It most certainly DOES involve end users – and they are basically not aware of an impending issue. This already involves any future-dated time events that have been set while on the old time zone. That this involves most every timetable is quite scary. Yes airlines, for example, plan their slots based on UTC, but those map back to where they were expected to be when the timetable was set up. I don’t know what the solution is, but I know that the experts in IANA and those dealing with CLDR, as well as the suppliers like Microsoft etc, will all be held at fault if (when) the s*** hits the fan. Personally, I think a formal press release should be issued, ideally signed by all the above, drawing PUBLIC attention to the impending crisis. The fix will need to come from end user IT managers and database/application teams. So far, everyone seems to be sleepwalking into this without a care. Peter Aggus From: Mark Davis Ⓤ via tz <tz@iana.org> Sent: October 3, 2026 2:00 PM To: Tim Parenti <tim@timtimeonline.com> Cc: tz@iana.org Subject: [tz] Re: Suggestion: ATTENTION lawmakers and timekeeping authorities We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to. Note that for Unicode (CLDR) there are the additional constraints imposed by names. We have one 'generic' name for time zones that don't have daylight savings, and three for ones that do; and in ~100 different languages, plus acronyms (where they are widely recognized). And it is harder for implementations to update names x languages than it is for the TZDB data we're based on. It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries. Mark On Sat, Oct 3, 2026 at 1:25 PM Tim Parenti via tz <tz@iana.org <mailto:tz@iana.org> > wrote: On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz <tz@iana.org <mailto:tz@iana.org> > wrote: Why not put a notice on the main IANA TzDb page in BIG RED LETTERS Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.) We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative: https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE... Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful. That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with. -- Tim Parenti
I have a commission to write a news article about this, for The Conversation with a creative commons license, so they often get reprinted in other news venues. I was thinking of doing this post 1 Nov, but it sounds like something much sooner would be more useful? I can try to pull together the issues that have been raised in the discussion list. More is needed obviously, but this could be one part of it. Best, Michelle. --- Dr Michelle Bastian Senior Lecturer in Theory, Culture and Environment ESALA, Edinburgh College of Art, University of Edinburgh NB: I finish work at 3pm on Tuesdays and Thursdays Latest article: (with Miriam Jensen) The Temporal Deliberation Approach, in Planning Theory, https://journals.sagepub.com/doi/abs/10.1177/14730952261448513 From: Jake Zukowski via tz <tz@iana.org> Sent: 05 October 2026 17:20 To: tz@iana.org Subject: [tz] Re: Suggestion: ATTENTION lawmakers and timekeeping authorities Hi there, Respectfully, this is a technical group whose interests are making sure time is correctly calculated and thereby accurately reflected within technical platforms moving forward. Utilizing other knowledge sources or media organizations to voice concerns about vigilance in updating systems may be more a more effective way to communicate disruptive changes. Best, Jake On Saturday, 3. October 2026 23:43:50 (+02:00), peter--- via tz wrote: Hi I am a business consultant based in Victoria, BC, Canada. I became aware of an impending problem when our provincial government decided to ‘pull the switch’ and ditch Standard Time early this year. I was curious how it all worked and what had to change behind the scenes. My knowledge of the processes involving CLDR and tzData at the time was zero – like most people not ‘in the know’. I have learned a lot following this discussion over the last six months or so and the more I learn the more concerned I get about how the real world is going to cope. Microsoft has only recently pushed Windows Update 26H2, the first that contains the new BC time zone – and so far only to some streams. It will be late October before most users have the update. Nothing about it tells users they have anything else to do and IT managers seem oblivious to the problem in the main. Users have to manually change their time zone. Having 26H2 is not enough – you are still on Pacific Time US & Canada, which maps to America/lLosAngeles and will fall back in November. Microsoft also made available a paper pointing out that Outlook and Teams (and, by implication, lots of other similar non-MS software) sets events in UTC at the time the event is set, using the then-current local time zone. So, any meetings set up for after the next clock change are already wrong – and users do not even know that. This is orders of magnitude worse than Y2K – which was reasonably dealt with by IT staff and basically did not really involve end users at all. This is different. It most certainly DOES involve end users – and they are basically not aware of an impending issue. This already involves any future-dated time events that have been set while on the old time zone. That this involves most every timetable is quite scary. Yes airlines, for example, plan their slots based on UTC, but those map back to where they were expected to be when the timetable was set up. I don’t know what the solution is, but I know that the experts in IANA and those dealing with CLDR, as well as the suppliers like Microsoft etc, will all be held at fault if (when) the s*** hits the fan. Personally, I think a formal press release should be issued, ideally signed by all the above, drawing PUBLIC attention to the impending crisis. The fix will need to come from end user IT managers and database/application teams. So far, everyone seems to be sleepwalking into this without a care. Peter Aggus From: Mark Davis Ⓤ via tz <tz@iana.org<mailto:tz@iana.org>> Sent: October 3, 2026 2:00 PM To: Tim Parenti <tim@timtimeonline.com<mailto:tim@timtimeonline.com>> Cc: tz@iana.org<mailto:tz@iana.org> Subject: [tz] Re: Suggestion: ATTENTION lawmakers and timekeeping authorities We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to. Note that for Unicode (CLDR) there are the additional constraints imposed by names. We have one 'generic' name for time zones that don't have daylight savings, and three for ones that do; and in ~100 different languages, plus acronyms (where they are widely recognized). And it is harder for implementations to update names x languages than it is for the TZDB data we're based on. It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries. Mark On Sat, Oct 3, 2026 at 1:25 PM Tim Parenti via tz <tz@iana.org<mailto:tz@iana.org>> wrote: On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz <tz@iana.org<mailto:tz@iana.org>> wrote: Why not put a notice on the main IANA TzDb page in BIG RED LETTERS Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.) We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative: https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE... Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful. That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with. -- Tim Parenti The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. Is e buidheann carthannais a th’ ann an Oilthigh Dhùn Èideann, clàraichte an Alba, àireamh clàraidh SC005336.
Quoting Mark Davis (U) via tz on Saturday October 03, 2026:
We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to.
There is no doubt more we can do here, and I think there is merit in a dedicated straightforward resource that provides a step-by-step for policy makers. We've done a mix of presentations, engagements and articles on the topics, e.g. in 2023: Because of this propagation time, it is crucial that policymakers provide a year or more advance notice to IANA before changes to their time zone policies take effect. Without this lead time it is almost guaranteed there will be great confusion caused by the use of inconsistently distributed time zone data. With that said I think the key is identifying the right forums to share this work so it actually makes a useful impact. The nature of the work we do at IANA and broadly within ICANN tends to be more with national governments. With the recent changes in Canada it seems this is being more driven at a provincial level, and there is probably not a good nexus there to reach the right people. So, where can we spread the message to help it resonate? Is there any place where the message could help make an impactful difference to policymaking? kim
On Oct 4, 2026, at 9:57 AM, Kim Davies via tz <tz@iana.org> wrote:
With that said I think the key is identifying the right forums to share this work so it actually makes a useful impact. The nature of the work we do at IANA and broadly within ICANN tends to be more with national governments.
How many of the people in those national governments who are involved with timekeeping changes are aware that the IANA and ICANN are involved with the time zone database, how many think "they're the Internet people; time zones are something else, so they wouldn't have anything to do with them", and how many aren't even aware of the *existence* of the time zone database? Note that these people aren't necessarily bureaucrats and technicians in a timekeeping department of the government; legislators may be involved (to the extent of provoking a peeved head of government to publicly disclose a legislator's personal cell phone number to get people to call him and demand a change).
With the recent changes in Canada it seems this is being more driven at a provincial level, and there is probably not a good nexus there to reach the right people.
I think daylight saving time might also be under the control of states rather than solely of the federal government in Australia as well. The states have some more limited control in the US, as they can opt out of daylight saving time, as Arizona has done.
So, where can we spread the message to help it resonate? Is there any place where the message could help make an impactful difference to policymaking?
That's a good question (and one for which I have no good answer, as I'm not in the circles where the policymaking is done).
On 10/4/2026 2:15 PM, Guy Harris via tz wrote:
With the recent changes in Canada it seems this is being more driven at a provincial level, and there is probably not a good nexus there to reach the right people. I think daylight saving time might also be under the control of states rather than solely of the federal government in Australia as well. The states have some more limited control in the US, as they can opt out of daylight saving time, as Arizona has done.
Arizona, Hawaii, Puerto Rico and the pacific territories - all under US control and have decided to opt out of DST. (The territories are considered states for the purpose of the law.) States can opt out one or more of their time zones (if more than one zone in their state). No restriction on whether a state borders a time zone line (Ohio or Iowa could opt out of DST without further federal approval). Changing the dividing lines is decided by the US Secretary of Transportation. While Canadians would balk at that level of national control it helps provide a more coordinated definition of the time zones instead of allowing opt out and changing zones down to the community level. But all politics being local, it works for Canada.
On Oct 4, 2026, at 11:39 AM, James Bellaire <bellaire@tk.com> wrote:
While Canadians would balk at that level of national control it helps provide a more coordinated definition of the time zones instead of allowing opt out and changing zones down to the community level. But all politics being local, it works for Canada.
In theory, current information and communication technology might make it easier to deal with community-level control of timekeeping. In practice, however, I suspect the tzdb maintainers might throw up their hands if confronted with too much of that.
On 2026-10-04 13:04, Guy Harris via tz wrote:
On Oct 4, 2026, at 11:39 AM, James Bellaire wrote:
While Canadians would balk at that level of national control it helps provide a more coordinated definition of the time zones instead of allowing opt out and changing zones down to the community level. But all politics being local, it works for Canada.
In theory, current information and communication technology might make it easier to deal with community-level control of timekeeping.
In practice, however, I suspect the tzdb maintainers might throw up their hands if confronted with too much of that.
As is evident, control does not imply care, coordination, understanding of, or taking responsibility for, any of the disruption and negative impacts! -- Take care. Thanks, Brian Inglis Calgary, Alberta, Canada La perfection est atteinte Perfection is achieved non pas lorsqu'il n'y a plus rien à ajouter not when there is no more to add mais lorsqu'il n'y a plus rien à retrancher but when there is no more to cut -- Antoine de Saint-Exupéry
On 10/4/2026 3:04 PM, Guy Harris via tz wrote:
On Oct 4, 2026, at 11:39 AM, James Bellaire <bellaire@tk.com> wrote:
While Canadians would balk at that level of national control it helps provide a more coordinated definition of the time zones instead of allowing opt out and changing zones down to the community level. But all politics being local, it works for Canada. In theory, current information and communication technology might make it easier to deal with community-level control of timekeeping.
In practice, however, I suspect the tzdb maintainers might throw up their hands if confronted with too much of that.
I hope that it does not come to that point. There has been a lot of effort over the past 40 years to track every time change since 1970 and allow the option for tracking prior to 1970 for specific use cases. (40 years being the introduction of the Olsen database.) But, looking at the minor line shifts I wonder how many end users actually use the america/indiana/* , america/north_dakota/* , america/kentucky/* zones (and others created by a line change). If their software allows the choice of these options. If Golden and RDEK did shift to "Vancouver Time", how many people would look for a specific time zone for their community or would they simply change their clock settings to america/vancouver (or the alias provided by downstream software)? That is where one would need to look at the downstream programmers to see if they support each community level zone.
Date: Sun, 4 Oct 2026 11:15:06 -0700 From: Guy Harris via tz <tz@iana.org> Message-ID: <6E38E194-0AED-4977-A10B-40F6808523B9@sonic.net> | I think daylight saving time might also be under the control of | states rather than solely of the federal government in Australia | as well. Completely under the control of the states, the commonwealth govt has nothing to do with it at all, except that they can override the territories (NT & ACT, and a few remote ones I think) territory legislatures should they desire. When AU was created from the states, the commonwealth were granted specific powers (defence, international relations/trade, ...) and everything else was reserved to the states. The states can donate powers to the govt should they desire, but apart from making changes to the constitution (something rarely attempted and less often successful) that's the only way for the commonwealth to take over something. kre
On 10/4/2026 12:57 PM, Kim Davies via tz wrote:
Quoting Mark Davis (U) via tz on Saturday October 03, 2026:
We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to. There is no doubt more we can do here, and I think there is merit in a dedicated straightforward resource that provides a step-by-step for policy makers. We've done a mix of presentations, engagements and articles on the topics, e.g. in 2023:
Because of this propagation time, it is crucial that policymakers provide a year or more advance notice to IANA before changes to their time zone policies take effect. Without this lead time it is almost guaranteed there will be great confusion caused by the use of inconsistently distributed time zone data.
I understand the ask for a year's notice but the list of changes made where a year's notice was given is very short. In the US, the Cruz amendment to the current "Sunshine Protection Act" sets the effective date two years out from enactment (which will need to be modified if enacted in November or December before the bill dies). The 2021 version had an effective date of November 2023 when it passed the Senate in March 2022. So at least the senate committee has received the message on giving advance notice. The effective 2007 change in dates for DST was enacted in 2005. The effective 1987 change to the beginning date for DST was enacted in July 1986. I am not sure how well the public in Canada would have taken it if all of this year's changes were proclaimed effective March 2027 instead of immediately. The end users didn't want to wait a year for the change. Perhaps the presentation can include illustrations (without insulting specific policymakers) of where short notice caused a problem?
With that said I think the key is identifying the right forums to share this work so it actually makes a useful impact. The nature of the work we do at IANA and broadly within ICANN tends to be more with national governments. With the recent changes in Canada it seems this is being more driven at a provincial level, and there is probably not a good nexus there to reach the right people.
So, where can we spread the message to help it resonate? Is there any place where the message could help make an impactful difference to policymaking?
Perhaps through the ITU? While there is no international approval required for a country to change their time zones, educating the right agencies in each country might help the decision makers in their countries. (Although with local politics the decision makers may still decide to ignore the technical hurdles.)
On 2026-10-04 12:27, James Bellaire via tz wrote:
On 10/4/2026 12:57 PM, Kim Davies via tz wrote:
Quoting Mark Davis (U) via tz on Saturday October 03, 2026:
We've discussed in Unicode about putting up a page with a similar warning. I would encourage an IANA web page on the topic. While the TZDB is unknown to many people, a page would provide a place for others to link to. There is no doubt more we can do here, and I think there is merit in a dedicated straightforward resource that provides a step-by-step for policy makers. We've done a mix of presentations, engagements and articles on the topics, e.g. in 2023:
Because of this propagation time, it is crucial that policymakers provide a year or more advance notice to IANA before changes to their time zone policies take effect. Without this lead time it is almost guaranteed there will be great confusion caused by the use of inconsistently distributed time zone data.
I understand the ask for a year's notice but the list of changes made where a year's notice was given is very short.
In the US, the Cruz amendment to the current "Sunshine Protection Act" sets the effective date two years out from enactment (which will need to be modified if enacted in November or December before the bill dies). The 2021 version had an effective date of November 2023 when it passed the Senate in March 2022. So at least the senate committee has received the message on giving advance notice. The effective 2007 change in dates for DST was enacted in 2005. The effective 1987 change to the beginning date for DST was enacted in July 1986.
I am not sure how well the public in Canada would have taken it if all of this year's changes were proclaimed effective March 2027 instead of immediately. The end users didn't want to wait a year for the change.
AFAICT those were premier, party, or government led initiatives, with their elections approaching, possibly wanting to appear show strengthm to their peers, voters, others out E, and someone down S! ;^p
Perhaps the presentation can include illustrations (without insulting specific policymakers) of where short notice caused a problem?
With that said I think the key is identifying the right forums to share this work so it actually makes a useful impact. The nature of the work we do at IANA and broadly within ICANN tends to be more with national governments. With the recent changes in Canada it seems this is being more driven at a provincial level, and there is probably not a good nexus there to reach the right people.
So, where can we spread the message to help it resonate? Is there any place where the message could help make an impactful difference to policymaking?
Perhaps through the ITU? While there is no international approval required for a country to change their time zones, educating the right agencies in each country might help the decision makers in their countries. (Although with local politics the decision makers may still decide to ignore the technical hurdles.)
Have they yet recognized that most of the world runs on tech they ignore, and it ignores them back? I am not sure how much they or the national telecomms or standards agencies would care about times or zones, except at levels where multiple zones are involved, nationally across North America, wide South American countries, someone at the EU level; tropical and Asian countries mostly don't use DST or have multiple zones. I think perhaps IATA is the closest linked to us (thru SSIM schedules, notice periods, standards, and notices) and best placed agency to reach the travel industry, transport utilities, regulators, and legislators, if we can get some cooperation, possibly also involving their government related counterpart ICAO, to document and illustrate the impacts, losses, and penalties when the passengers, staff, and crew devices, and business systems, are not on the same local times as airline and airport systems, with illustrations from short notice, high impact events like Egypt's fiasco, without mentioning anyone explicitly! -- Take care. Thanks, Brian Inglis Calgary, Alberta, Canada La perfection est atteinte Perfection is achieved non pas lorsqu'il n'y a plus rien à ajouter not when there is no more to add mais lorsqu'il n'y a plus rien à retrancher but when there is no more to cut -- Antoine de Saint-Exupéry
On 10/4/2026 3:29 PM, Brian Inglis via tz wrote:
On 2026-10-04 12:27, James Bellaire via tz wrote:
Perhaps through the ITU? While there is no international approval required for a country to change their time zones, educating the right agencies in each country might help the decision makers in their countries. (Although with local politics the decision makers may still decide to ignore the technical hurdles.)
I am not sure how much they or the national telecomms or standards agencies would care about times or zones, except at levels where multiple zones are involved, nationally across North America, wide South American countries, someone at the EU level; tropical and Asian countries mostly don't use DST or have multiple zones.
I am just considering a path to their policymakers. I expect there is some staff involvement. And while some staff and agencies may be surprised by an edict put forth by top leadership, one hopes that there is at least some rumor or discussion in advance or the edict (or a process where staff and agencies are involved in the mechanics of the change) that would allow for the people who know the challenges to help guide the process.
I think perhaps IATA is the closest linked to us (thru SSIM schedules, notice periods, standards, and notices) and best placed agency to reach the travel industry, transport utilities, regulators, and legislators, if we can get some cooperation, possibly also involving their government related counterpart ICAO, to document and illustrate the impacts, losses, and penalties when the passengers, staff, and crew devices, and business systems, are not on the same local times as airline and airport systems, with illustrations from short notice, high impact events like Egypt's fiasco, without mentioning anyone explicitly! Good suggestions. The more in country / local people who know the ramifications of changing time zones the better in affecting the process from the ground up in every jurisdiction.
On Sun, 4 Oct 2026 at 13:01, Brooks Harris <brooks@edlmax.com> wrote:
That said, putting it in red letters on the IANA page is not likely to accomplish much.
But is wouldn't hurt, would it?
If framed wrongly, or without care, it could. Ultimately, the goal is cooperation — "help us help you". Any sort of "tsk, tsk" with big red letters could cause our guidance to be perceived as scolding or adversarial, which would work against our actual goal. We can only really *advise* interested parties on how best to achieve a desired outcome with minimal disruption — primarily from a technical standpoint, but also to a lesser extent in other regards that ripple outward from there. We can also point to the potential risks and consequences of hastier decisions, so that decision-makers can better weigh them against the relative importance of other factors — not the least of which is the ever-fickle public opinion. That said, as Kim and Mark have discussed, I absolutely agree that we should make our official guidance more prominent and more replete — even if only so *I* can find it better when necessary! While we hear quite a bit from CLDR as some of our "closest neighbors" project-wise, it could help considerably if we were able to partner with IATA and ITU contacts on this as well. I'd imagine these sorts of last-minute changes can be rough on them, too, but they may also have some established processes that would be illustrative — e.g., beyond a certain point, the planes have already been scheduled, regardless of what local clocks may say! On Sun, 4 Oct 2026 at 14:28, James Bellaire via tz <tz@iana.org> wrote:
I am not sure how well the public in Canada would have taken it if all of this year's changes were proclaimed effective March 2027 instead of immediately. The end users didn't want to wait a year for the change.
Perhaps the presentation can include illustrations (without insulting specific policymakers) of where short notice caused a problem?
That's arguably the hardest part. One way to be neutral and not risk offending anyone is to simply publish observed timelines for the rollout of all changes — good, bad, and everything in between. But to be more effective, we would likely need to point to specific examples where short leadtimes have led to adverse impacts. The two strategies could perhaps work best in tandem.
Perhaps through the ITU? While there is no international approval required for a country to change their time zones, educating the right agencies in each country might help the decision makers in their countries. (Although with local politics the decision makers may still decide to ignore the technical hurdles.)
Yes, even if the best possible advice is made readily accessible through multiple channels, it will still occasionally fall on deaf ears. And even among those that *will* hear it, their tolerance for disruption in order to achieve a policy goal will vary considerably. It all comes with the territory. At the end of the day, it's just our job to document and translate — proactively when possible and prudent, and reactively (albeit measuredly) otherwise. On Sun, 4 Oct 2026 at 15:55, James Bellaire via tz <tz@iana.org> wrote:
I am just considering a path to their policymakers. I expect there is some staff involvement.
Even just having this information in a prominent, dedicated spot would help a lot for visibility if a staffer is asking a colleague, search engine, or LLM such things as "we want to do this; what do we need to consider first?" -- Tim Parenti
When it comes to examples... We have the bad example of Palestine from some years back where they did the announcement the day before it was supposed to happen. That was a lot chaos and eventually the decision was rescinded. There is also the good example of, of all places, Iran also from some years back where parliament decided to abolish DST but the guardian council (that is the priests) blocked the proposal and said they should have longer lead time. Have that proposal shown up again since? /MF
On 2026-10-03 14:00, Mark Davis Ⓤ via tz wrote:
It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries.
Agreed, and I am looking forward (if that is the right phrase...) to considerable confusion about naming and abbreviations next month. What are CLDR's plans for time zone names and abbreviations for America/Vancouver, America/Edmonton, America/Winnipeg, and America/Inuvik for the CLDR release planned for later this month? I looked for recent updates in the CLDR public repository and found only the following, which don't match my expectations as they seem to hardwire previous timekeeping practice, not the timekeeping that starts next month. Evidently I'm missing something. It'd be helpful to know the plans (at least for American English) as I imagine people will be asking on the tz list about this. <https://github.com/unicode-org/cldr/commit/016a645a70ce48e050dba145dbf838406...> (2026-09-29) changes America/Winnipeg to use stdOffset="-06" dstOffset="-05". <https://github.com/unicode-org/cldr/commit/7b1090ada749eb29d85d64bfb6c195267...> (2026-09-18) changes America/Inuvik to use stdOffset="-07" dstOffset="-06". <https://github.com/unicode-org/cldr/commit/d18f9652beadcec10ff40d71556e72826...> (2026-07-08) changes America/Edmonton to use stdOffset="-07" dstOffset="-06". <https://github.com/unicode-org/cldr/commit/c8babf10d858e413b977425efbff3b746...> (2026-03-07) changes America/Vancouver to use stdOffset="-08" dstOffset="-07".
CLDR's goal is to represent common usage, not the letter of the law, and we think the most common way these zones will be referred to this winter is their current (summer) names. These are names that users are familiar with, and it's how this whole process has been framed from the beginning: permanent daylight time. If new names become common this winter, we will adopt them. The changes linked above effectively ignore TZDB's decision to model these changes as new standard times. stdOffset="-06" dstOffset="-05" means: if the active offset (as returned by TZDB) is UTC-5, use the daylight name, even if TZDB says it's standard time (tm_isdst=false). These changes have been released in CLDR patches, which have rolled out alongside 2026b-e. My Android phone is already showing me PDT/MDT/CDT for Vancouver/Edmonton/Winnipeg this winter. On Tue, 6 Oct 2026 at 20:29, Paul Eggert via tz <tz@iana.org> wrote:
On 2026-10-03 14:00, Mark Davis Ⓤ via tz wrote:
It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries.
Agreed, and I am looking forward (if that is the right phrase...) to considerable confusion about naming and abbreviations next month.
What are CLDR's plans for time zone names and abbreviations for America/Vancouver, America/Edmonton, America/Winnipeg, and America/Inuvik for the CLDR release planned for later this month? I looked for recent updates in the CLDR public repository and found only the following, which don't match my expectations as they seem to hardwire previous timekeeping practice, not the timekeeping that starts next month.
Evidently I'm missing something. It'd be helpful to know the plans (at least for American English) as I imagine people will be asking on the tz list about this.
< https://github.com/unicode-org/cldr/commit/016a645a70ce48e050dba145dbf838406...>
(2026-09-29) changes America/Winnipeg to use stdOffset="-06" dstOffset="-05".
< https://github.com/unicode-org/cldr/commit/7b1090ada749eb29d85d64bfb6c195267...>
(2026-09-18) changes America/Inuvik to use stdOffset="-07" dstOffset="-06".
< https://github.com/unicode-org/cldr/commit/d18f9652beadcec10ff40d71556e72826...>
(2026-07-08) changes America/Edmonton to use stdOffset="-07" dstOffset="-06".
< https://github.com/unicode-org/cldr/commit/c8babf10d858e413b977425efbff3b746...>
(2026-03-07) changes America/Vancouver to use stdOffset="-08" dstOffset="-07".
Once the situation settles down in Canada we can deal with whatever conventions that they settle on for referring to their time zones (bearing in mind that the names have to be different from the US ones wherever they have different behavior — and that we have constraints on when we do a big sweep for language data.) Cf. https://www.unicode.org/cldr/charts/49/by_type/timezones.north_america.html (Luckily we only have acronyms in languages where people would be familiar with the acronyms.) On Tue, Oct 6, 2026 at 12:19 PM Robert Bastian <robertbastian@unicode.org> wrote:
CLDR's goal is to represent common usage, not the letter of the law, and we think the most common way these zones will be referred to this winter is their current (summer) names. These are names that users are familiar with, and it's how this whole process has been framed from the beginning: permanent daylight time. If new names become common this winter, we will adopt them.
The changes linked above effectively ignore TZDB's decision to model these changes as new standard times. stdOffset="-06" dstOffset="-05" means: if the active offset (as returned by TZDB) is UTC-5, use the daylight name, even if TZDB says it's standard time (tm_isdst=false).
These changes have been released in CLDR patches, which have rolled out alongside 2026b-e. My Android phone is already showing me PDT/MDT/CDT for Vancouver/Edmonton/Winnipeg this winter.
On Tue, 6 Oct 2026 at 20:29, Paul Eggert via tz <tz@iana.org> wrote:
On 2026-10-03 14:00, Mark Davis Ⓤ via tz wrote:
It would be terminally confusing for speakers of English, for example, to have "13 October 2026 14:30 Pacific Time" mean two different UTC times in different countries.
Agreed, and I am looking forward (if that is the right phrase...) to considerable confusion about naming and abbreviations next month.
What are CLDR's plans for time zone names and abbreviations for America/Vancouver, America/Edmonton, America/Winnipeg, and America/Inuvik for the CLDR release planned for later this month? I looked for recent updates in the CLDR public repository and found only the following, which don't match my expectations as they seem to hardwire previous timekeeping practice, not the timekeeping that starts next month.
Evidently I'm missing something. It'd be helpful to know the plans (at least for American English) as I imagine people will be asking on the tz list about this.
< https://github.com/unicode-org/cldr/commit/016a645a70ce48e050dba145dbf838406...>
(2026-09-29) changes America/Winnipeg to use stdOffset="-06" dstOffset="-05".
< https://github.com/unicode-org/cldr/commit/7b1090ada749eb29d85d64bfb6c195267...>
(2026-09-18) changes America/Inuvik to use stdOffset="-07" dstOffset="-06".
< https://github.com/unicode-org/cldr/commit/d18f9652beadcec10ff40d71556e72826...>
(2026-07-08) changes America/Edmonton to use stdOffset="-07" dstOffset="-06".
< https://github.com/unicode-org/cldr/commit/c8babf10d858e413b977425efbff3b746...>
(2026-03-07) changes America/Vancouver to use stdOffset="-08" dstOffset="-07".
On 10/6/2026 3:19 PM, Robert Bastian via tz wrote:
The changes linked above effectively ignore TZDB's decision to model these changes as new standard times. stdOffset="-06" dstOffset="-05" means: if the active offset (as returned by TZDB) is UTC-5, use the daylight name, even if TZDB says it's standard time (tm_isdst=false).
Thank you for the clarification. I read the changes in cldr as controlling the offset (overriding tzdata) instead of matching the offset provided by tzdata.
On 2026-10-06 12:19, Robert Bastian wrote:
The changes linked above effectively ignore TZDB's decision to model these changes as new standard times.
Thanks for explaining. Some more comments plus a couple of questions. As I understand it, after November 1 neither CLDR nor TZDB will use the English-language abbreviations suggested by the respective regional governments. For the new timekeeping we have: 0 1 2 3 -07 MST PDT PCT America/Vancouver -06 CST MDT ABT America/Edmonton -06 CST MDT ??? America/Inuvik -05 EST CDT MBT America/Winnipeg where column 0 is the UT offset, column 1 is the abbreviation that TZDB will use, column 2 is what CLDR will use, and column 3 is what the respective governments suggest (NWT does not suggest so I listed "???".) This disagreement is unfortunate but I don't see a simple solution other than maybe just going with column 0 in TZDB for now. Although columns 1 and 2 are both compatible with a lot of existing software and standards (e.g., Internet RFC 5322), neither column is likely to match what people will say; and although column 3 attempts to avoid that confusion it is at the expense of compatibility and it regionalizes abbreviations in a way that will likely have significant problems. I don't see anything in bleeding-edge CLDR mentioning "ABT" or "Alberta Time", or similarly for the other government-suggested abbreviations or names, so it looks like CLDR is also holding off on column 3 for now. Will this month's new CLDR stick with the name "Mountain Daylight Time" for Alberta's timekeeping, and will it allow "Mountain Time" as a generic name for this new timekeeping even though Alberta doesn't have alternatives for summer and winter? And what will CLDR do for "Mountain Standard Time" in Alberta - will that string be forbidden in Alberta, or never generated? Or am I misreading the tables again? Also, it seems the phrase "Pacific Time" is uniquely problematic here, as it exactly matches the legal name for -07 in BC whereas CLDR uses it as a generic name for both -08 and -07. We won't have this problem with "Manitoba Time" and "Alberta Time", though we will have other problems.
See also https://www.unicode.org/reports/tr35/dev/tr35-dates.html#Using_Time_Zone_Nam... As to “Alberta time”, outside of where local abbreviations are available, the generic location name may be used. Examples at https://st.unicode.org/cldr-apps/v#r_zones/da// (These are examples, not production data). “ Edmonton-tid” which I think is “Edmonton time” -s Enviado desde mi iPad
El oct 6, 2026, a la(s) 6:44 p.m., Paul Eggert via tz <tz@iana.org> escribió:
On 2026-10-06 12:19, Robert Bastian wrote:
The changes linked above effectively ignore TZDB's decision to model these changes as new standard times.
Thanks for explaining. Some more comments plus a couple of questions.
As I understand it, after November 1 neither CLDR nor TZDB will use the English-language abbreviations suggested by the respective regional governments. For the new timekeeping we have:
0 1 2 3 -07 MST PDT PCT America/Vancouver -06 CST MDT ABT America/Edmonton -06 CST MDT ??? America/Inuvik -05 EST CDT MBT America/Winnipeg
where column 0 is the UT offset, column 1 is the abbreviation that TZDB will use, column 2 is what CLDR will use, and column 3 is what the respective governments suggest (NWT does not suggest so I listed "???".)
This disagreement is unfortunate but I don't see a simple solution other than maybe just going with column 0 in TZDB for now. Although columns 1 and 2 are both compatible with a lot of existing software and standards (e.g., Internet RFC 5322), neither column is likely to match what people will say; and although column 3 attempts to avoid that confusion it is at the expense of compatibility and it regionalizes abbreviations in a way that will likely have significant problems.
I don't see anything in bleeding-edge CLDR mentioning "ABT" or "Alberta Time", or similarly for the other government-suggested abbreviations or names, so it looks like CLDR is also holding off on column 3 for now. Will this month's new CLDR stick with the name "Mountain Daylight Time" for Alberta's timekeeping, and will it allow "Mountain Time" as a generic name for this new timekeeping even though Alberta doesn't have alternatives for summer and winter? And what will CLDR do for "Mountain Standard Time" in Alberta - will that string be forbidden in Alberta, or never generated? Or am I misreading the tables again?
Also, it seems the phrase "Pacific Time" is uniquely problematic here, as it exactly matches the legal name for -07 in BC whereas CLDR uses it as a generic name for both -08 and -07. We won't have this problem with "Manitoba Time" and "Alberta Time", though we will have other problems.
Will this month's new CLDR stick with the name "Mountain Daylight Time" for Alberta's timekeeping, and will it allow "Mountain Time" as a generic name for this new timekeeping even though Alberta doesn't have alternatives for summer and winter?
This is a problem that already exists with America/Phoenix. I think ICU uses "Mountain Standard Time" for Arizona's generic time, which makes sense but is not what CLDR specifies. It would make sense to do the same for permanent-DST zones, and to spec that.
And what will CLDR do for "Mountain Standard Time" in Alberta - will that string be forbidden in Alberta, or never generated? Or am I misreading the tables again?
I think this question is about parsing, which I don't know the answer to. Parsing localized time zone names is generally not a good idea. On Wed, 7 Oct 2026 at 04:52, Steven R. Loomis <srl295@gmail.com> wrote:
See also https://www.unicode.org/reports/tr35/dev/tr35-dates.html#Using_Time_Zone_Nam...
As to “Alberta time”, outside of where local abbreviations are available, the generic location name may be used. Examples at https://st.unicode.org/cldr-apps/v#r_zones/da// (These are examples, not production data). “ Edmonton-tid” which I think is “Edmonton time”
-s
Enviado desde mi iPad
El oct 6, 2026, a la(s) 6:44 p.m., Paul Eggert via tz <tz@iana.org> escribió:
On 2026-10-06 12:19, Robert Bastian wrote:
The changes linked above effectively ignore TZDB's decision to model these
changes as new standard times.
Thanks for explaining. Some more comments plus a couple of questions.
As I understand it, after November 1 neither CLDR nor TZDB will use the English-language abbreviations suggested by the respective regional governments. For the new timekeeping we have:
0 1 2 3 -07 MST PDT PCT America/Vancouver -06 CST MDT ABT America/Edmonton -06 CST MDT ??? America/Inuvik -05 EST CDT MBT America/Winnipeg
where column 0 is the UT offset, column 1 is the abbreviation that TZDB will use, column 2 is what CLDR will use, and column 3 is what the respective governments suggest (NWT does not suggest so I listed "???".)
This disagreement is unfortunate but I don't see a simple solution other than maybe just going with column 0 in TZDB for now. Although columns 1 and 2 are both compatible with a lot of existing software and standards (e.g., Internet RFC 5322), neither column is likely to match what people will say; and although column 3 attempts to avoid that confusion it is at the expense of compatibility and it regionalizes abbreviations in a way that will likely have significant problems.
I don't see anything in bleeding-edge CLDR mentioning "ABT" or "Alberta Time", or similarly for the other government-suggested abbreviations or names, so it looks like CLDR is also holding off on column 3 for now. Will this month's new CLDR stick with the name "Mountain Daylight Time" for Alberta's timekeeping, and will it allow "Mountain Time" as a generic name for this new timekeeping even though Alberta doesn't have alternatives for summer and winter? And what will CLDR do for "Mountain Standard Time" in Alberta - will that string be forbidden in Alberta, or never generated? Or am I misreading the tables again?
Also, it seems the phrase "Pacific Time" is uniquely problematic here, as it exactly matches the legal name for -07 in BC whereas CLDR uses it as a generic name for both -08 and -07. We won't have this problem with "Manitoba Time" and "Alberta Time", though we will have other problems.
My earlier table about abbreviation use in western Canada inadvertently omitted Yukon. Here's my attempt at fixing that: 1 2 3 4 -07 MST ? YST America/Whitehorse -07 MST PDT PCT America/Vancouver -06 CST MDT ABT America/Edmonton -06 CST MDT ? America/Inuvik -05 EST CDT MBT America/Winnipeg where column 0 is the UT offset, column 1 is the abbreviation that TZDB uses for timestamps after November 1, column 2 is CLDR uses, and column 3 is what the respective governments suggest. The "?" for Yukon is because CLDR's common/main/en.xml doesn't list an abbreviation for the Yukon metazone; it lists only "Yukon Time" as the long name. As before, the "?" for NWT is because NWT doesn't suggest an abbreviation as far as I know.
To be fair, the common usage in Canada even during the summer isn't necessarily to use xDT to denote the time zone. The average person often doesn't know when DST turns on or off so just uses xST year-round. It's quite common for someone to say "5pm EST" in the middle of July while intending UTC-4 (albeit imprecise and inaccurate). Others get around this by simply using xT year-round. This seems more like a conscious effort to get more people to actually start calling daylight time xDT, which is currently uncommon. Displaying a time in both xST and xDT during the winter seems to be very uncommon and likely to cause even more confusion (e.g. 3 p.m. PST, 4 p.m. PDT, 4 p.m. MST). Framing this like it is the current practice seems odd as most people do not think about the second letter in the abbreviation. With the exception of BC's Pacific Time which is genuinely challenging, using the official "Alberta Time" and "Manitoba Standard Time" would probably be more unambiguous and intuitive to the average person.
Hi Tim,
That said, putting it in red letters on the IANA page is not likely to accomplish much.
But is wouldn't hurt, would it? The main TzDb page is the entry point to all of the TzDb complexity. I'm sure most people don't care. But we'd like the lawmakers to at least have a chance of being informed. We could promote the TzDb page reference more effectively, couldn't we? I looked around (again) for references to TzDb. Searching Google for "time zone" one of the results is the Wikipedia article: Time zone https://en.wikipedia.org/wiki/Time_zone There's no mention of TzDb! We could add a paragraph and an external link. That might help, for example. There a Wikipedia article tz database https://en.wikipedia.org/wiki/Tz_database We could add the warning there. There are many sites showing time zones and tools, such as timeanddate.com. If we had a warning like this on the main page we could ask them to also reference it, I bet they probably would. Little by little that warning would find more exposure. ( I'm not expecting it to go viral. ) We just need a reasonably well written title and page and give it a try. Thanks, -Brooks On 2026-10-03 04:24 PM, Tim Parenti wrote:
On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz<tz@iana.org> wrote:
Why not put a notice on the main IANA TzDb page in BIG RED LETTERS
Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.)
We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative: https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE...
Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful.
That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with.
-- Tim Parenti
For what it's worth, as a regular wikipedia editor I've just added some text to https://en.wikipedia.org/wiki/Time_zone#Computer_systems and https://en.wikipedia.org/wiki/Tz_database#Maintenance. -- John Le 2026-10-04 19:00, Brooks Harris via tz a écrit :
Hi Tim,
That said, putting it in red letters on the IANA page is not likely to accomplish much.
But is wouldn't hurt, would it?
The main TzDb page is the entry point to all of the TzDb complexity. I'm sure most people don't care. But we'd like the lawmakers to at least have a chance of being informed. We could promote the TzDb page reference more effectively, couldn't we?
I looked around (again) for references to TzDb. Searching Google for "time zone" one of the results is the Wikipedia article:
Time zone https://en.wikipedia.org/wiki/Time_zone
There's no mention of TzDb! We could add a paragraph and an external link. That might help, for example.
There a Wikipedia article tz database https://en.wikipedia.org/wiki/Tz_database We could add the warning there.
There are many sites showing time zones and tools, such as timeanddate.com. If we had a warning like this on the main page we could ask them to also reference it, I bet they probably would. Little by little that warning would find more exposure. ( I'm not expecting it to go viral. )
We just need a reasonably well written title and page and give it a try.
Thanks, -Brooks
On 2026-10-03 04:24 PM, Tim Parenti wrote:
On Sat, 3 Oct 2026 at 15:58, Brooks Harris via tz <tz@iana.org> wrote:
Why not put a notice on the main IANA TzDb page in BIG RED LETTERS
Although we do have some basic guidance in tz-link.html, it is admittedly a bit hidden: https://data.iana.org/time-zones/data/tz-link.html#coordinating (So hidden that I keep accidentally looking for it in theory.html instead.)
We could probably stand to beef that up with some harder data and/or make it more visible in the distribution. For that purpose, the timeline that Christian Cruz posted about the America/Coyhaique zone's (not-so-recent) creation may be illustrative: https://lists.iana.org/hyperkitty/list/tz@iana.org/message/5I4VYFK5QQHKELZIE...
Chronicling aspects of the fuller data lifecycle on a site like my https://tzdata-meta.timtimeonline.com/ could also be helpful.
That said, putting it in red letters on the IANA page is not likely to accomplish much. As others have pointed out, governments and the general public aren't always aware of our project to begin with.
-- Tim Parenti
That's a good start, thank you. I think it's places like this where a warning to policy makers should be. Lawmakers and their assistants are more likely to find it. I think an explanation page should reside in the main TzDb page. I think Kin Davie's paragraph is a good start for a more complete explanation. Because of this propagation time, it is crucial that policymakers provide a year or more advance notice to IANA before changes to their time zone policies take effect. Without this lead time it is almost guaranteed there will be great confusion caused by the use of inconsistently distributed time zone data. -Brooks On 2026-10-05 05:37 AM, john--- via tz wrote:
For what it's worth, as a regular wikipedia editor I've just added some text to https://en.wikipedia.org/wiki/Time_zone#Computer_systems and https://en.wikipedia.org/wiki/Tz_database#Maintenance.
Date: Tue, 29 Sep 2026 16:18:12 -0400 From: Tim Parenti via tz <tz@iana.org> Message-ID: <CAFpi07y8RsMYX5LVbk8HuGGMEsWEkxp8VUVhmSGOwnu8Ha4NvA@mail.gmail.com> | Specifically, the language you cite from theory.html exists to | remind us of our archival role, and that we should follow *de facto* usage | for abbreviations in a way that tends to lag, not lead. This is a nice objective, but it ignores reality - unless some other source for the abbreviations shows up (and CLDR doesn't really impact enough places to make a huge difference) what tzdata supplies decides what will be the common de facto usage - so the policy is effectively for tzdata to follow what tzdata has already done. I'd suggest a better approach, when a clear demonstrated desire to make the abbreviation be a particular string exists, that tzdata should assist in that endeavour, rather than fight against it, by installing the abbreviation that has been requested (or legislated, or is clearly derivable from a legislated long zone name). If it turns out after doing that that something different becomes popular, which I suspect would be unlikely, then it can be changed. Consider for a minute what would happen should the US State Dept demand (or US Congress, or the California legislature) decide that California will also use UTC-7 year round, and legislate that the abbreviation shall be CAT (CA Time). Even perhaps impose a penalty for anyone using anything different. What would tzdata do then, just claim "that doesn't apply to us, we're going to use MST instead". Really? What if Donald Trump decided to de-fund UCLA, and impose tariffs of 200% on anything UCLA wants to purchase ... would tzdata just say "who are you to tell us what to do?" Would that fly? So, I'd suggest PCT (for BC) and if it is what they want, MBT for Manitoba, and whatever the Albertan's picked (ABT?) for Alberta. If the NWT have picked a name for their zone, use it as well. What harm can it possibly do? These strings are largely just noise, but people like their noise to be of a particular timbre. Regions near borders using the zone of the other side of the border, which actually need a zone of their own, rather than just using the normal one for the "other side", should just use the other side's abbreviation, whatever that is, that is how they will think of it. "Even though we're on the Alberta side of the border, we're using PCT because we interact with British Columbians more than other Albertans". kre
On 2026-09-29 15:39, Robert Elz via tz wrote:
Consider for a minute what would happen should the US State Dept demand (or US Congress, or the California legislature) decide that California will also use UTC-7 year round, and legislate that the abbreviation shall be CAT (CA Time).
Even more, consider what would happen if every state and province, or indeed every city, legislated its own time zone abbreviation and insisted that we use it. What would we do then? What we'd likely do is switch to numeric abbreviations. There are good reasons in TZDB itself to avoid a multiplicity of alphabetic abbreviations for what is in practice the same timekeeping. And this isn't just us: there's a similar issue, for example, for Canadian national broadcasts, which will almost surely not list PCT, YST, WNUT, NWTT, ABT, CST, MBT, EST separately, as these would be eight confusing abbreviations for what are only three distinct time zones. Until/unless popular usage agrees on time zone abbreviations, numeric abbreviations are likely the best way to go. Besides, abbreviation choice is more of a localization than a timekeeping issue, and it's better for CLDR (which deals with localization) to deal with it than for us to deal with it.
CLDR doesn't really impact enough places to make a huge difference
Although that was true decades ago, nowadays surely many more users deal with time zone names and abbreviations via CLDR than directly via TZDB.
On Tue, Sep 29, 2026, 4:10 PM Paul Eggert via tz <tz@iana.org> wrote:
CLDR doesn't really impact enough places to make a huge difference
Although that was true decades ago, nowadays surely many more users deal with time zone names and abbreviations via CLDR than directly via TZDB.
Sometimes I really wish as a normal user I could see TZDB codes instead of CLDR names. I was recently trying to find out if an Android phone has the new timezone data for Alberta or not. I decided to use it to create a calendar event after November 1 in Alberta time and compare it with US Pacific Time to find out what the phone thinks Alberta's time on that day is. I was given this list of Canadian time zones to pick from. I think many people would agree it's very hard for a normal user to figure out what they should choose: Mountain Standard - Dawson Creek GMT-7 Mountain Standard Time GMT-7 Pacific Daylight Time GMT-7 Yokun Time GMT-7 Central Standard - Saskatchewan GMT-6 Central Standard Time GMT-6 .... (followed by two GMT-5 choices, two GMT-4 choices, and a GMT-3:30) I really wish I could see America/Edmonton in front of one of them. That way, I could be certain which one it thinks is Alberta time on that day. Roozbeh
On Tue, 29 Sept 2026 at 21:16, Roozbeh Pournader via tz <tz@iana.org> wrote:
I was given this list of Canadian time zones to pick from. I think many people would agree it's very hard for a normal user to figure out what they should choose:
And a lot of apps will only ever show you the *current* string and offset for the zone in a dropdown, even if a different one would be in effect on the date that's actually in question for a given operation. Though some edge cases around transition times make this hard to generalize with 100% accuracy, I've not seen many handle it particularly well. (I even draw the line at various places for my own handful of personal projects, considering expected audience, seasonality, etc.) -- Tim Parenti
participants (22)
-
Adam Thompson -
Aoun Khan -
Brian Inglis -
Brooks Harris -
Doug Ewell -
Garrett Wollman -
Guy Harris -
Jake Zukowski -
James Bellaire -
james.epp@fusioncu.com -
john@wilcock.fr -
Kim Davies -
Magnus Fromreide -
Mark Davis Ⓤ -
Michelle Bastian -
Paul Eggert -
peter@tmcconsulting.ca -
Robert Bastian -
Robert Elz -
Roozbeh Pournader -
Steven R. Loomis -
Tim Parenti