Re: CVS commit: src/lib/libc/time

[Paul Goyette <paul%whooppee.com@localhost>]

On Thu, 18 Apr 2019, Paul Goyette wrote:

> > > XXX: To be strictly correct, the tzgetname() call should probably take
> > > a time reference point in order to return the appropriate zone name
> > > for the time given.
> >
> > Yes, the timezone name should be associated with the specific point in
> > time.
>
> Of course, the problem is, what specific point in time should be used,
> and how is it specified?  There is no argument in which the point in
> time could be passed.
>
> For now we should probably add a BUGS section to the man page to make
> note of this restriction.

Here's a draft.  If there's no objection, I can commit it in a day or 
two.

Index: tzset.3
===================================================================
RCS file: /cvsroot/src/lib/libc/time/tzset.3,v
retrieving revision 1.39
diff -u -p -r1.39 tzset.3
--- tzset.3     19 Oct 2018 23:05:35 -0000      1.39
+++ tzset.3     18 Apr 2019 02:06:20 -0000
@@ -427,3 +427,10 @@ function conforms to
 .\" @(#)newtzset.3     8.2
 .\" This file is in the public domain, so clarified as of
 .\" 2009-05-17 by Arthur David Olson.
+.Sh BUGS
+The name and GMT offset for a timezone can vary over time.  The
+.Fn tzgetname
+and
+.Fn tzgmtoff
+functions return values are those that apply to the latest point in 
time
+for which values are available, even if that refers to a future time.



+--------------------+--------------------------+-----------------------+
| Paul Goyette       | PGP Key fingerprint:     | E-mail addresses:     |
| (Retired)          | FA29 0E3B 35AF E8AE 6651 | paul%whooppee.com@localhost     |
| Software Developer | 0786 F758 55DE 53BA 7731 | pgoyette%netbsd.org@localhost   |
+--------------------+--------------------------+-----------------------+