[6459] in Kerberos
Re: Kerberos 5 and the new year
daemon@ATHENA.MIT.EDU (Darren Reed)
Mon Jan 8 18:42:23 1996
Date: Tue, 9 Jan 1996 10:13:40 +1100
From: darrenr@nabaus.com.au (Darren Reed)
To: jik@annex-1-slip-jik.cam.ov.com, tytso@MIT.EDU
Cc: kerberos@MIT.EDU
I realise that it is highly unlikely that the same code will be running in
the year 2100 (which isn't a leap year), perhaps not even in the year 2000,
but the year 2000 *IS* a leap year too, AFAIK (the century is divisible by
400). Somehow, the following still looks incorrect, as did the patch posted
for Kerberos, in how it handles "leap-centuries".
darren
> *** nntp.c_orig Sat Nov 19 01:01:23 1994
> --- nntp.c Wed Jan 3 01:41:34 1996
> ***************
> *** 320,324 ****
> day += maxdays[month];
>
> ! ss = ((((year-1970) * 365 + (year-1968)/4 + day - 1) * 24L + hh) * 60
> + mm) * 60 + ss;
>
> --- 320,324 ----
> day += maxdays[month];
>
> ! ss = ((((year-1970) * 365 + (year-1969)/4 + day - 1) * 24L + hh) * 60
> + mm) * 60 + ss;
>
> (Looks a lot like the problem with the KDC error message on dialup servers)
