[tz] [PROPOSED] Use "PST/PDT" for Philippine time

[Paul Eggert]

Michael H Deckers wrote:
>     Fine if this were a generally applied guideline -- but it is not,
>     see eg America/Adak, Indonesia, and many others. When is this guideline
>     applied, and when not?

The guideline is supposed to be applied reasonably consistently for clear-cut 
cases such as the Philippines. Alaska and Indonesia are more complicated, since 
they are not merely a question of what's the name of a time zone; they are also 
a question of time zone boundaries moving, and where it's not so clear what is 
being named.  For example, a historian today would be unlikely to write "Sitka 
observed Alaska Standard Time in 1984" because in 1984 the Alaska time zone 
boundaries were quite different and Sitka was then in the geographically small 
part of Alaska that was in the same time zone as the Yukon.

Unfortunately I don't see a good solution for these more-complicated cases, and 
am reluctant to change the data if it's likely this would introduce as much or 
more confusion than it would alleviate.

>     And all that does not justify the resurrection of an invented abbreviation.

First, for the Phillippines it's no longer an abbreviation that exists only 
because we invented it. It's taken on a life of its own in reliable sources and 
in national law. And second, I doubt whether these sources got "PST" from tzdb, 
since tzdb has been using "PHST" for over two decades and used "PST" only 
briefly before that. Most likely it was independent reinvention.