Date: Wed, 26 Apr 2000 17:05:30 -0400 From: Rob Furphy <rcf@ox.com> To: freebsd-java@freebsd.org, glewis@trc.adelaide.edu.au Subject: *BETTER* Pre-alpha JDK 1.2.2 TimeZone (time) bug fix Message-ID: <39075A1A.70DBE0F1@ox.com> References: <39061716.2E69B654@ox.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello all,
In my fix from yesterday although the wall clock time was correct, the
timezone
was incorrect and the JVM was defaulting to a generic timezone which set
the
jvm's clock to the best guess of what a 'wall clock' would say the time
was.
Needless to say this eventually fails if you ever cross a daylight savings
time boundry
while the jvm is running.
I have a better code fix now which I think will work for all cases and give
the
correct time and timezone. I now get the correct time and the timezone is
"America/New_York" as I expect.
Could everyone else please try this new change to see if it works in your
timezone?
My informal way to get this into the next patchset is here:
in src/freebsd/native/java/lang/java_props_md.c at line 283 you'll find
#ifdef __FreeBSD__ /*** XXX ***/
hasDST = local_tm->tm_isdst;
offset = local_tm->tm_gmtoff;
#else
hasDST = daylight == 0 ? kStandard : kDaylight;
offset = timezone / -60; /* Conventions reversed between UNIX &
Java */
#endif
change this to:
#ifdef __FreeBSD__ /*** XXX ***/
hasDST = (local_tm->tm_isdst == 0) ? kStandard :
kDaylight; /*** changed line ***/
offset = (local_tm->tm_gmtoff / 60) - (local_tm->tm_isdst == 0 ? 0
: 60); /*** re-changed line ***/
#else
hasDST = daylight == 0 ? kStandard : kDaylight;
offset = timezone / -60; /* Conventions reversed between UNIX &
Java */
#endif
Rob F
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-java" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?39075A1A.70DBE0F1>