Paul,
The changes look correct except that you linked America/Rainy_River to America/Toronto instead of America/Winnipeg.
Rainy River is west of 90°W and needs to remain on Central Time.

David,
I don't deny that there may have been some small reserves that did not observe daylight saving.
They could still be added to the database if somebody wants to do the research and provide appropriate names.
However, it would be wrong to use the names of existing towns and cities to represent small reserves that exist outside of those towns and cities.

-chris

On Sat, 15 Oct 2022 at 17:11, dpatte <dpatte@relativedata.com> wrote:
My concern about these changes is that I seem to remember that some first nations in those areas did not use daylight-savings time on their reserves, and that they were real but very small 'zones', not the large zones they ended up being assumed to be.

Unfortunately I have no documentation either way on this.


----- Original Message -----
From: Paul Eggert via tz [mailto:tz@iana.org]
To: "Chris Walton" <crj.walton@gmail.com>
Cc: tz@iana.org
Sent: Sat, 15 Oct 2022 12:53:28 -0700
Subject: Re: [tz] Northern Ontario (Canada)

Thank you for doing that investigative work. Proposed patch attached and
installed into the tzdb development repository on GitHub.

This patch follows your suggestions, except that it moves Nipigon's and
Rainy River's dubious data to 'backzone' instead of removing them
entirely, as I worry that some users will want the appearance of
completeness no matter how illusory.