My confusion with Asia/Aqtau
Chuck Soper
chucks at lmi.net
Tue Jul 22 21:53:17 UTC 2003
At 10:42 PM -0700 7/21/03, Paul Eggert wrote:
>Chuck Soper <chucks at lmi.net> writes:
>
>> Does this indicate that I'm using a different version of the
>>timezone package?
>
>Most likely, yes.
I would like to be able to verify what version of the timezone
package (data and tools) is installed on my operating system. If I
had evidence that an old version of the timezone package was being
used in a current operating system then I could file a problem report
with the OS vendor. This mechanism could facilitate OS vendors to
stay current with the time zone package (I don't know if this is an
issue).
I think that adding version information to each time zone information
file might be a way to verify version information for these data. In
a previous post, Martin Smoot suggested that the sixteen bytes
reserved for future use in the header might be used for version
information. Would both a package version and a file version be
needed?
> > How can I determine what version of the timezone package is installed
>> on my system? I hope that I'm using version 2003a.
>
>There should be a way to do this, at least for the timezone code like
>zdump.c. Perhaps you could add a "--version" option for zdump and
>zic? If you donated code to do that, I suspect that Arthur David
>Olson would buy it.
>
>zdump has this at the start:
>
>#ifndef lint
>#ifndef NOID
>static char elsieid[] = "@(#)zdump.c 7.29";
>#endif /* !defined NOID */
>#endif /* !defined lint */
>
>and zic.c has something similar, but they don't work well.
>Nowadays many compilers optimize away these unused strings.
>
>Basicall, all the "--version" option needs to do is to print
>the elsieid.
A "--version" option for zdump would also be useful.
More information about the tz
mailing list