On Fri, 24 Sept 2021 at 23:12, Paul Eggert <eggert@cs.ucla.edu> wrote:
On 9/22/21 11:06 AM, Stephen Colebourne via tz wrote: So, instead of the 2021a1-vs-2021b compromise I suggested earlier (which turns out to have important compatibility issues just with the name "2021a1"!) I now plan to compromise along the lines I suggested a dozen hours ago, here:
https://mm.icann.org/pipermail/tz/2021-September/030686.html
I.e., the idea is to revert most (but not all) of the objected-to changes. In particular, this will revert the changes to Europe/Oslo and Europe/Stockholm, which have drawn the most objections. The idea is to take the first step now, and to take more steps in future releases (which should not be distant-future releases, as we need to continue to make and exhibit a good-faith effort to fix the problem). This will let us generate just one version, 2021b.
Although your followup <https://mm.icann.org/pipermail/tz/2021-September/030689.html> didn't agree with the proposed compromise, it didn't reject it in quite the strong terms that I saw in earlier emails. And indeed, the proposed compromise is far closer to what you suggested than to what I suggested, as it replaces only 9 Zones with Links instead of 30-odd Zones. So it's a reasonable way forward, even if both of us dislike the compromise.
No. No, no, no. I've talked about Oslo and Stockholm as examplars - I reject the whole concept of link merging without a bigger plan. Merging ANY zones in major world business centres will immediately impact any Joda-Time user that updates their time-zone data themselves. 9 changes instead of 30 changes is not helpful. The way out of this mess is zero merges today and a rational discussion next week. Stephen