An interesting situation cropped up. Seems in one system, someone configured a point as timezone "GMT+10" to get Brisbane time (just in case some foolish government introduces DST like they trialled in the 90's).
The Olson database defines a number of timezones that match the GMT±XX name pattern, and the sign convention is opposite to that of normal UTC conventions:
RC=0 stuartl@rikishi ~ $ TZ=Etc/GMT+10 date
Tue May 3 13:41:23 GMT+10 2016
RC=0 stuartl@rikishi ~ $ TZ=Etc/GMT-10 date
Wed May 4 09:41:27 GMT-10 2016
RC=0 stuartl@rikishi ~ $ TZ=Etc/GMT date
Tue May 3 23:41:36 GMT 2016
RC=0 stuartl@rikishi ~ $ TZ=Australia/Brisbane date
Wed May 4 09:41:44 AEST 2016
Is Project Haystack meant to follow these same conventions, or are we going to "fix" it?
Brian FrankTue 3 May 2016
We follow the Olson database exactly - in fact we generate our database straight from the Olson database on a periodic basis. The GMT names are indeed extremely confusing - not sure why they did that
Stuart Longland Tue 3 May 2016
Hi all,
An interesting situation cropped up. Seems in one system, someone configured a point as timezone "GMT+10" to get Brisbane time (just in case some foolish government introduces DST like they trialled in the 90's).
The Olson database defines a number of timezones that match the GMT±XX name pattern, and the sign convention is opposite to that of normal UTC conventions:
Is Project Haystack meant to follow these same conventions, or are we going to "fix" it?
Brian Frank Tue 3 May 2016
We follow the Olson database exactly - in fact we generate our database straight from the Olson database on a periodic basis. The GMT names are indeed extremely confusing - not sure why they did that