Re: [PATCH] FAT timezones [was Re: updating the RTC automagically]

Alexander Viro (viro@math.psu.edu)
Sat, 20 Nov 1999 22:36:31 -0500 (EST)


On Sat, 20 Nov 1999, Peter Samuelson wrote:

>
> [Alexander Viro]
> > Ouch... Timezone is much nastier than just one offset. First of all,
> > you can't use the _current_ offset - you should take offset for the
> > moment of file creation.
>
> I know that. But MS-DOS doesn't know that. It puts the current
> *local* time right on the platter, and what it reads off the platter it
> still interprets as *local* time even if it was created in the "other"
> half of the year. Even NT is that stupid -- I checked.
>
> My patch (if I'm not mistaken) allows Linux to match those semantics,
> which I agree are quite broken. I don't know any better way to do so,
> and I *have* heard from people who want this "feature".

Hold on. Why not use sys_tz? settimeofday(NULL, &tz) from userland utility
and there you go. Nothing except the sucking filesystems will be affected.
Why do you need per-fs granularity?

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/