This is the mail archive of the
libc-locales@sourceware.org
mailing list for the GNU libc locales project.
day/abday in POSIX vs. ISO 14652
- From: "Paul Marcher" <flyer dot REMOVE dot SPAM at gsmpaaiml dot com>
- To: libc-locales at sourceware dot org
- Date: Wed, 13 May 2009 18:43:40 +0200
- Subject: day/abday in POSIX vs. ISO 14652
Dear readers,
I am unsure about the current policy of the GNU C library maintainers
concerning day/abday for countries, where the first day of the week is not
Sunday.
(Although a search in the archives gave some results, those results were
rather vague statements than a clear commitment to either POSIX or ISO
14652.)
Some locales are being moved to day/abday field _not_ starting in Sunday.
ISO/IEC 14652.2002(E) requires in section 4.8 LC_TIME to start day/abday
with the first day of a week (which might be Sunday or not).
Therefore, changing the day/abday fields to reflect the first day of a
week *seems* sound.
However, the above section is marked as âcontroversialâ in the document.
For example the âcomments from USâ part of the document SC22/WG20 N956D
explains why: Such changes are incompatible with POSIX.
And from [1], I am now under the strong impression, that the GNU C library
strifes for POSIX compatibility. On [1] we can for example read
It [ The GNU C library ] follows all relevant standards (ISO C 99,
POSIX.1c, POSIX.1j, POSIX.1d, Unix98, Single Unix Specification).
However, the âSingle Unix Specificationâ (v3) requires (Chapter 7.3.5
LC_TIME; definition of day and abday) that day and abday start in Sunday.
day and abday *have* to start on Sundays. Unconditionally. Regardless of
the first day of a week.
So while the GNU C library claims to try to follow the Single Unix
Specification, it breaks locales in the sense of Single Unix
Specificationâonly to meet a section of an ISO/IEC Technical report, that
is marked as controversial.
Does the GNU C library trade adherence to the Single Unix Specification to
following a controversial section of an ISO Technical Report?
Kind regards,
Paul
P.S.: I can see the limitations of day/abday starting in Sundays.
I can understand the urge to set the week modifier as hinted in the
controversial section of above technical report.
But I'd rather adhere to the Single Unix Specification and be sloppy on
the controversial ISO chapter than the other way round.
[1]
http://www.gnu.org/software/libc/
--
Do _not_ remove the ".REMOVE.SPAM" in the address.
It's a catch for too smart address harvesters