Re: I'm working on libgcj...

["Robert Collins" <robert dot collins at itdomain dot com dot au>]

----- Original Message -----
From: "DJ Delorie" <dj@delorie.com>
To: <robert.collins@itdomain.com.au>
Cc: <cygwin-apps@sources.redhat.com>
Sent: Monday, April 16, 2001 12:00 PM
Subject: Re: I'm working on libgcj...


>
> > What spec is timezone() derived from?
>
> Timezone is one of those "traditionally different" things.  I don't
> recall exactly which OS defined it the cygwin way, but it's not
> uncommon to define it that way.
>

Thanks. The bit of code I'm fighting with is

#ifdef STRUCT_TM_HAS_GMTOFF
  tzoffset = -(tim->tm_gmtoff); // tm_gmtoff is secs EAST of UTC.
#elif HAVE_TIMEZONE
  tzoffset = timezone;          // timezone is secs WEST of UTC.
#else
  // FIXME: there must be another global if neither tm_gmtoff nor
timezone
  // is available, esp. if tzname is valid.
  // Richard Earnshaw <rearnsha@arm.com> has suggested using difftime to
  // calculate between gmtime and localtime (and accounting for possible
  // daylight savings time) as an alternative.  Also note that this same
  // issue exists in java/util/natGregorianCalendar.cc.
  tzoffset = 0L;
#endif

And I was hoping to avoid putting a #ifdef __CYGWIN__ in there. I
presume the long term fix is to add tm_gmtoff to the timezone struct?

Rob