-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 06/06/2012 11:51 AM, Ian Abbott wrote:
I guess the main problem here is that the week-day names and month names (in the Gregorian/Julian calendar system) are known internally by glibc, whereas timezone names and abbreviations are external to glibc (and are also a bit of a moving target).
Yup. Not trivial. However, GLIBC and coreutils are already picking up timezone information and abbreviations (eg. date +%z_%Z) from the tzif files, and they already pick up localizations from localedef files, so it's not too far a jump to suggest extending it to an arbitrarily sized list of unique zone-names (and with Tobias's suggestion, unique zone-abbreviations). BTW, a look at the wikipedia list of tz abbreviations shows a case to confound some translators for unique zone descriptions, because I suspect that some languages use non-unique words for Arab, Arabian, Arabic: AST Arab Standard Time (Kuwait, Riyadh) UTC+03 AST Arabian Standard Time (Abu Dhabi, Muscat) UTC+04 AST Arabic Standard Time (Baghdad) UTC+03 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iQIcBAEBAgAGBQJPz4LhAAoJEDvrUfDmCx9L/z0P/jP3M9mWGm9EsjWv5Ra96Fri GkjBUXVHPoTwYFKXJK5gg9ZjqxVSDfTIyTBruzSg/szkhCNEwSS56LL5j0AOtKwm zyyW/2MTVBRRWQ5W+HST9O5i6oR1kEutHXNkCCCNknsiH9HmdjK7jBcMankHre/R 7+SN18opNCwef6/ZTJJY95FAE2D1tDbJPdKTXK9tqpM0wNWkxeTjq9yvfPXbgm5r B/slkCyJ8Ef8cAhnr6lEMQI6KWz2bFWt+PZBfp07kzyUbK1ZSmqTlUFF+gFx1Ttb KHJ/dhgHecGHu8lwZoywscIVbwDRN3iOmGU1jCpRBzOKjYh468hxGLbiVTVXdcLs G+6DmVkF4XbZDIdvkJk3TLs5EUzLqaXgJ3VNSK6thfjiS9vqE9sDUzTEsDNgxUj0 qQL9ugRx2DLOcu4zVjD8syh5opCFvKPN6psCmHCxFOnlWrnGvvuzeBDHSzPoi40c 2ps2KhDBf2DXWH0ZAh8+oe8ZEsN4SQsmGjx8BZX3Xm/76FHG3rU+vk/8jtESZ+DB FHRcGUAIU/b7G/55pRyd819aJGu2Utgio9NVFxivcxfOsNLbJ3i6uuNhQ4XNbfmR sN0Tz1+eMcjaC4lHaVNOKGYyV1MZBEiXvAJe6YFuFn2YNRIrSj4A3cFU35Jko2AG OJ1XHLLGw0vO05ihsABD =egOy -----END PGP SIGNATURE-----