[tz] Value of data ...

[Lester Caine]

Norbert supplied a link to a ecma document which makes interesting reading ...

 From the section on 'Time Zone Names'
> All registered Zone and Link names are allowed. Implementations must recognize all such names, and use
> best available current and historical information about their offsets from UTC and their daylight saving time
> rules in calculations.

Note the use of 'and historical information' here.

One has to bare in mind the target data set when deciding what data is useful 
and what can be thrown away, and for much day to day work a TZ database that 
only provides future 'estimated' data is probably more than adequate. Banks may 
be able to get away with 'it's ten years old so we can throw it away' and take 
control of dormant funds, but for some of us the history is as important as the 
present day. In fifty years time a large proportion of the data we are creating 
today will still be available. I dug out material going back 200+ years last 
night which has now been scanned and archived ... material on 'time' in the Isle 
of Man for the record ... which helps to clarify calculations for UTC time 
offsets for material in the early 1900's, which if now stored as UTC normalised 
data need to be 'corrected' using an historic version of the TZ data. If someone 
tries to use that data with a 'winnowed' data set, then the wrong times will be 
returned? Who is to say what an end user is going to be doing with their machine?

All right you can reduce 440 zones down to 300 or so, but all of the original 
440 names still need to be available, and in these days does saving a few bytes 
really matter? 'Simplification' of access is more to do with showing the 
appropriate subset of options by the application and historical ones may 
actually be the correct answer. So the data should be as complete a record of 
the 'best available current and historical information' that we have! Providing 
a version of the data that will be WRONG for some users is not an option?

-- 
Lester Caine - G8HFL
-----------------------------
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk