On Mon, 28 Nov 2022 at 20:25, Paul Eggert via tz <tz@iana.org> wrote:
On 2022-11-28 03:31, Stephen Colebourne via tz wrote:
perhaps you'd like to think whether your choice to be deliberately awkward here
Actually it's less awkward for me, since if you've made the patch I needn't worry about writing the change, checking it, checking that the comments match the data, etc. And even though running "git format-patch" is a tiny bit more work for you, that's OK if it makes the overall project more scalable. This sort of thing is reasonably common in software projects developed by collaborators.
I think it is important to document a key difference. I think it is reasonable to ask those who wish to actually maintain data that is *only* in backzone to write patches (although the process is acane because GitHub PRs are switched off). By contrast, the Singapore case is just general maintenance: * you had been made aware of the need to update Singapore in advance * you were already making a change * making the change to Singapore would have resulted in a more logical changeset * making the change to Singapore would have resulted in less churn and emails The goal should be that if you are making a change that affects one of the non-backzone files, but the change also impacts one of the backzone locations, you should also be responsible for updating backzone. Anything else is simply noise and annoyance for the project, and against the long-term health of tzdb. Stephen