bug#63349: Bug in date when using UTC/GMT timeszones in the TZ variable

2023-05-07 Thread Eiríkur Hjartarson via GNU coreutils Bug Reports
On 7.5.2023 14:52, Andreas Schwab wrote: On Mai 07 2023, Eiríkur Hjartarson via GNU coreutils Bug Reports wrote: Now the "bug": It's not a bug. Thanks for the explanation, in my defense, I did read the date info page and the FAQ at https://www.gnu.org/software/coreutils/faq/coreutils-faq

bug#63349: Bug in date when using UTC/GMT timeszones in the TZ variable

2023-05-07 Thread Paul Eggert
On 5/7/23 07:52, Andreas Schwab wrote: Thus TZ=UTC+2 means two hours before UTC. Yes, and this mistake seems to be common enough that I installed the attached patch into Gnulib, so that it'll propagate into the Coreutils manual, which should help people who read the 'date' documentation (ad

bug#63349: Bug in date when using UTC/GMT timeszones in the TZ variable

2023-05-07 Thread Andreas Schwab
On Mai 07 2023, Eiríkur Hjartarson via GNU coreutils Bug Reports wrote: > Now the "bug": It's not a bug. > $ TZ=Europe/Riga date --iso-8601=minutes -d "2024-01-01T00:00-05:00" > 2024-01-01T07:00+02:00 > > $ TZ=UTC+2 date --iso-8601=minutes -d "2024-01-01T00:00-05:00" > 2023-01-01T03:00-02:00 > >

bug#63349: Bug in date when using UTC/GMT timeszones in the TZ variable

2023-05-07 Thread Eiríkur Hjartarson via GNU coreutils Bug Reports
Hi, I'm on Fedora-38 GNU/Linux and the version string of 'date' is "date (GNU coreutils) 9.1". $ ls -l /etc/localtime lrwxrwxrwx. 1 root root 40 jan 11 2022 /etc/localtime -> ../usr/share/zoneinfo/Atlantic/Reykjavik Now the "bug": $ TZ=Europe/Riga date --iso-8601=minutes -d "2024-01-01T00