On 08/14/13 08:51, Derick Rethans wrote:
Nobody has mentioned any significant problems yet.
I can think of three already, without doing any further research into this.
1. Debian's "dpkg-reconfigure tzdata".
This isn't a problem, since dpkg-reconfigure operates by directly inspecting the tree of installed tz binary files, so it's unaffected by the proposed changes to zone.tab.
2. Debian's/Raspian's "tzselect" lists for "Asia":
Yes, that's the tz package's tzselect program. With the proposed changes the resulting TZ setting would work as before, the only difference being a political difference how the setting itself is spelled (whether the setting is spelled TZ='Europe/Zagreb' or TZ='Europe/Belgrade', say). This objection to this change is essentially political. Whether one considers the change to be a significant problem depends on whether one is trying to make the best technical change independent of politics (as I was) or is trying to defer to political considerations of users who look at internal TZ settings and might object to how they're spelled.
3. PHP's DateTime support
This is the same issue as the tzselect program.
And I can assure you people will get upset by this.
As I think we both agree, people will get upset no matter what, even if we do nothing; it's more a question of how do we lessen the amount of future ire and work for everybody. All this being said, it does seem that the recent proposed changes to zone.tab do not have consensus, so as described in <http://mm.icann.org/pipermail/tz/2013-August/019546.html> I've reverted them in the experimental github version. I'll try to think of a gentler way of insulating this project from future political issues.
You also need to realise that the ISO 3166-1 alpha-2 codes are *not* just for countries anyway. Wikipedia states: "represent countries, dependent territories, and special areas of geographical interest." —
Thanks, although that's not enough to insulate us from politics (it certainly wasn't last time ...), it can't hurt to mention this in zone.tab and in iso3166.tab, so I added that to the commentary while reverting the changes.