> > btw, how stale is 3.2.5? Would it be appropriate to declare pre-v4.1 AIX
> > as "unsupported"?
> IMHO that's inappropriate, as even IBM hasn't done that yet. (They plan
> to in December, last I heard.) It's certainly more supported (tho less
> widely used) than SunOS.
> This problem isn't just limited to AIX anyway -- any system that doesn't
> have an int timezone would be affected, as the bug is in the code that
> handles timezone information when the timezone int (or the tm_gmtoff
> element in struct tm) is not available. Of course, NeXTstep seems to
> be the only other one that fits that category, and it's probably going
> away. While it would be convenient to drop AIX 3.2.5 and make this
> datetime bug disappear, I'll have it cleaned up shortly and then of
> any new (old) systems come along that need it, the code will be there.
I'm not aware of other USE_POSIX_TIME/!HAVE_INT_TIMEZONE machines seeing
this problem. Is AIX4.1 in that same features category?
I thought that AIX3.2.5 has _some_ timezone support; if not then you
really are the only machine I've found other than NextStep which is in
that category.
- Tom
------------------------------
