[tz] Regression Tests and Revision Controls

Paul Eggert eggert at cs.ucla.edu
Fri Aug 15 06:46:20 UTC 2014


Lester Caine wrote:
> My reading of the paperwork
> so far is that it is planned to have a full range base,

Sorry, I don't know what a 'full range base' is, but there are no plans 
for a full coverage of all civil time stamps in every era.  My only plan 
for backzone now is to be a repository for questionable/out-of-scope 
(pre-1970) data that we can easily move out of the tz database proper, 
one step at a time.  If people want to donate other data (within reason; 
see below) that'd be fine, but it's not something I want to spend time 
on.  The backzone file is out of scope for this project and if managing 
it becomes a significant burden we should spin it out into a separate 
project.

> what is not
> clear is how new historic data that identifies that a current timezone
> now has two historic routes.

For current backzone entries this can be deduced from pre1970.tab.  If 
backzone gets more complicated I suggest adding commentary to it in case 
a zone's boundaries are not clear.  Or if that's not enough, do a 
database redesign as a separate project.

> The planing seems to be based on tz zones,
> but as with all of these things there is no planning to manage tz
> identifiers.

The current approach will not scale.  It's meant only for relatively 
modest growth, where we can manage identifiers the same way we've always 
managed them.  Sorry, but a database redesign will be needed to cover 
appreciable quantities of pre-1970 data.

> Can we not just describe 'backzone' as historic data.

Sorry, I don't follow.  The tz database is almost all historic data.

> That the
> provenance of some material is 'poorly-sourced' just needs to be
> properly flagged within the file.

It is, in commentary within the file.  If those comments are not clear 
please feel free to submit patches; see the CONTRIBUTING file.


More information about the tz mailing list