Yet another option is to wait until after the March 29/30 transition and then release 2023c as a duplicate of 2023a. This would have a different set of problems, which I hope are obvious; surely this option would be worse than rolling back to 2023a today.
I think it would be fine to release 2023c right now as a duplicate of 2023a and then add historical detail about the period between Saturday and Wednesday in a future version which doesn’t need to be released immediately or even soon. Debbie
On Mar 28, 2023, at 12:22 PM, Paul Eggert <eggert@CS.UCLA.EDU> wrote:
On 2023-03-28 12:11, Saadallah Itani wrote:
Can you just release 2023-c based on the government decision yesterday which is the official and align with the government statement yesterday and comment its description for history records?
Yes, that is an option. The problem, though, is that there is genuine disagreement on the ground as to what rules should be followed, whether the prime minister's announcement last week was legal, how today's timestamps in hospitals and banks and etc. should be interpreted, and so forth. And generating such a 2023c would involve making further decisions, such as what abbreviation to use for these timestamps - would they be "BMT" or "EET"? and if they're "BMT" wouldn't that break some existing software?
Yet another option is to wait until after the March 29/30 transition and then release 2023c as a duplicate of 2023a. This would have a different set of problems, which I hope are obvious; surely this option would be worse than rolling back to 2023a today.