[tz] [PATCH 1/5] Holiday Islands does not exist.

J William Piggott elseifthen at gmx.com
Sun Feb 21 20:23:00 UTC 2016

On 02/20/2016 08:43 PM, Robert Elz wrote:
>     Date:        Sat, 20 Feb 2016 17:04:24 -0500
>     From:        J William Piggott <elseifthen at gmx.com>
>     Message-ID:  <56C8E2E8.5080700 at gmx.com>
>   | +# There is no location named Holiday Islands in Queensland Australia, holiday
>   | +# islands is a colloquial term used globally.
> That's true, but it is a widely used nomenclature, and probably better
> known that Whitsunday Islands.

Its wide use is what makes it a poor identifier to use in the zone
tables. It is used in referring to all of the Great Barrier Reef
holiday destinations, e.g., Bribie Island, Fraser Island, Great Keppel
Island, Lizard Island, Magnetic Island. This zone doesn't apply to them.

Most of the commercial websites I've found that cater to this specific
group of holiday islands refer to them by their proper name, ether
'Whitsunday Islands' or simply the 'Whitsundays'. For example:


Before making further arguments, I'd like to make clear that this patch
is about using an unambiguous name in the zone table comments. The zone
table comments are intended for use in applications like tzselect. These
applications should not assume the user has any prior knowledge of the
areas. To help users choose a file from the Time Zone Database it is
better to identify areas with their proper names, than some vague
colloquial term.

The use of 'Holiday Islands' in the zone tables seems to originate from
the comments in zone source file. So I think there should be some
mention of the proper name there. I do not have a strong opinion as to
the exact wording.

>   | Hayman and Lindeman are at the
>   | +# North and South extremes of the Whitsunday Islands archipelago, and
>   | +# Hamalton is in between;
> That would be Hamilton

True, my mistake.

>, and while Lindeman is near the southern extreme,
> Thomas Island is probably the actual limit.

I did not mean to imply that it was _the_ southern most point. However,
replacing 'at' with 'near' would be better.

>   | it is reasonable to believe that this time zone
>   | +# applies to all of the Whitsundays.
> That would require confirmation from someone in that part of the world
> I expect - some of the islands are probably uninhabited, and so have no
> real timezone.  Inhabited ones that are not tourist destinations (if such
> a thing really exists) could just use reqular Qld time (I have no idea.)

This was only a two year period, they're on 'reqular Qld time' now.

It probably is not relevant which islands in the group the database
considers part this zone. Will someone really try to use this
database to ascertain a time in 1993 on an uninhabited Whitsunday island?

Using the proper name to identify the area is relevant.

> kre
> ps: is all of this structure really needed for what are just comments
> intended for (local knowledgeable) humans?

I do not think it should be assumed everyone working on this file will
have local knowledge, especially colloquial terms. I find the source
file comments useful.

More information about the tz mailing list