Re: Happy 1234567890 everyone!

2009-02-14 Thread Mark Andrews

In message ef3daad70902131706x2277ac5fp5cc243354...@mail.gmail.com, 
Nathan Malynn writes:
 Question about 2k38: Aren't most Unixoid systems using 64-bit clocks now?

No.  Even if they were you have 32 bit timestamps in lots
of things that have to be handled even if time/time64 returns
a 64 bit timestamp.
 
 On Fri, Feb 13, 2009 at 8:03 PM, Chris Adams cmad...@hiwaay.net wrote:
  Once upon a time, Ravi Pina r...@cow.org said:
  Yes... that is more like the y2k38 problem on 03:14:07 UTC
  2038-01-19...
 
  Oddly enough, the end of the current Unix epoch is a prime.  Not only
  that, it is a Mersenne prime, 2^31 - 1.  Even more, it is the largest
  known Mersenne prime where its Mersenne number (31) is also a Mersenne
  prime (2^5 - 1).
 
  You can always count on numerology.  This means something!
  --
  Chris Adams cmad...@hiwaay.net
  Systems and Network Administrator - HiWAAY Internet Services
  I don't speak for anybody but myself - that's enough trouble.
 
 
 
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: mark_andr...@isc.org



Re: Happy 1234567890 everyone!

2009-02-14 Thread Florian Weimer
* Mark Andrews:

 In message ef3daad70902131706x2277ac5fp5cc243354...@mail.gmail.com, 
 Nathan Malynn writes:
 Question about 2k38: Aren't most Unixoid systems using 64-bit clocks now?

   No.  Even if they were you have 32 bit timestamps in lots
   of things that have to be handled even if time/time64 returns
   a 64 bit timestamp.

Those values can often be interpreted as unsigned, so we get another
68 years.  By then, hopefully, the required protocol updates can be
fully automated.



Re: Happy 1234567890 everyone!

2009-02-13 Thread Wayne E. Bouchard
You haven't lived until you've lived through an epoch.

On Fri, Feb 13, 2009 at 06:54:54PM -0500, Ravi Pina wrote:
 On Fri, Feb 13, 2009 at 06:49:49PM -0500, Steve Church wrote:
  Just in case you missed it.
  
  date -d Fri Feb 13 23:31:30 UTC 2009 +%s
  
  It's like a really geeky y2k without the potential cataclysm.  :)
  
  Steve
 
 Yes... that is more like the y2k38 problem on 03:14:07 UTC
 2038-01-19...
 
 ...by then I can only hope I am out of this profession. :)
 
 -r
 
 

---
Wayne Bouchard
w...@typo.org
Network Dude
http://www.typo.org/~web/



Re: Happy 1234567890 everyone!

2009-02-13 Thread Chris Adams
Once upon a time, Ravi Pina r...@cow.org said:
 Yes... that is more like the y2k38 problem on 03:14:07 UTC
 2038-01-19...

Oddly enough, the end of the current Unix epoch is a prime.  Not only
that, it is a Mersenne prime, 2^31 - 1.  Even more, it is the largest
known Mersenne prime where its Mersenne number (31) is also a Mersenne
prime (2^5 - 1).

You can always count on numerology.  This means something!
-- 
Chris Adams cmad...@hiwaay.net
Systems and Network Administrator - HiWAAY Internet Services
I don't speak for anybody but myself - that's enough trouble.



Re: Happy 1234567890 everyone!

2009-02-13 Thread Nathan Malynn
Question about 2k38: Aren't most Unixoid systems using 64-bit clocks now?

On Fri, Feb 13, 2009 at 8:03 PM, Chris Adams cmad...@hiwaay.net wrote:
 Once upon a time, Ravi Pina r...@cow.org said:
 Yes... that is more like the y2k38 problem on 03:14:07 UTC
 2038-01-19...

 Oddly enough, the end of the current Unix epoch is a prime.  Not only
 that, it is a Mersenne prime, 2^31 - 1.  Even more, it is the largest
 known Mersenne prime where its Mersenne number (31) is also a Mersenne
 prime (2^5 - 1).

 You can always count on numerology.  This means something!
 --
 Chris Adams cmad...@hiwaay.net
 Systems and Network Administrator - HiWAAY Internet Services
 I don't speak for anybody but myself - that's enough trouble.





Re: Happy 1234567890 everyone!

2009-02-13 Thread Chris Adams
Once upon a time, Nathan Malynn ne...@nerdramblingz.com said:
 Question about 2k38: Aren't most Unixoid systems using 64-bit clocks now?

Unix/POSIX systems use time_t to store the base time counter, which is
seconds since the epoch (1970-01-01 00:00:00 UTC).  Most platforms still
use a 32 bit time_t for compatibility.

However, it does appear that at some point, 64 bit Linux systems
switched to a 64 bit time_t, so I can only assume others are switching
as well.  Hopefully, the 32 bit systems (at least that have to count
seconds) will be mostly gone in another 29 years.

-- 
Chris Adams cmad...@hiwaay.net
Systems and Network Administrator - HiWAAY Internet Services
I don't speak for anybody but myself - that's enough trouble.



Re: Happy 1234567890 everyone!

2009-02-13 Thread Eric Gearhart
On Fri, Feb 13, 2009 at 6:06 PM, Nathan Malynn ne...@nerdramblingz.com wrote:
 Question about 2k38: Aren't most Unixoid systems using 64-bit clocks now?


Exactly! What are we going to do when we're at the end of the 2^64
epoch?? (after the sun burns out and.. oh wait)

--
Eric
http://nixwizard.net



Re: Happy 1234567890 everyone!

2009-02-13 Thread Joe Greco
 Once upon a time, Nathan Malynn ne...@nerdramblingz.com said:
  Question about 2k38: Aren't most Unixoid systems using 64-bit clocks now?
 
 Unix/POSIX systems use time_t to store the base time counter, which is
 seconds since the epoch (1970-01-01 00:00:00 UTC).  Most platforms still
 use a 32 bit time_t for compatibility.
 
 However, it does appear that at some point, 64 bit Linux systems
 switched to a 64 bit time_t, so I can only assume others are switching
 as well.  Hopefully, the 32 bit systems (at least that have to count
 seconds) will be mostly gone in another 29 years.

FreeBSD used a 64-bit time_t for the AMD64 port pretty much right away.
On the flip side, it used a 32-bit time_t for the Alpha port.  I guess
someone predicted it wouldn't be a problem.

Nowhere near as annoying a problem as the variability of the size of
size_t.

... JG
-- 
Joe Greco - sol.net Network Services - Milwaukee, WI - http://www.sol.net
We call it the 'one bite at the apple' rule. Give me one chance [and] then I
won't contact you again. - Direct Marketing Ass'n position on e-mail spam(CNN)
With 24 million small businesses in the US alone, that's way too many apples.



Re: Happy 1234567890 everyone!

2009-02-13 Thread Chris Adams
Once upon a time, Joe Greco jgr...@ns.sol.net said:
 FreeBSD used a 64-bit time_t for the AMD64 port pretty much right away.
 On the flip side, it used a 32-bit time_t for the Alpha port.  I guess
 someone predicted it wouldn't be a problem.

Tru64 on Alpha uses a 32 bit time_t (they have their own time64_t and
time64() call), so I expect *BSD and Linux on Alpha stayed with 32 bit
time_t for compatibility (Linux at least could run many Tru64 binaries).
-- 
Chris Adams cmad...@hiwaay.net
Systems and Network Administrator - HiWAAY Internet Services
I don't speak for anybody but myself - that's enough trouble.



Re: Happy 1234567890 everyone!

2009-02-13 Thread Steven M. Bellovin
On Fri, 13 Feb 2009 21:08:12 -0600
Chris Adams cmad...@hiwaay.net wrote:

 Once upon a time, Joe Greco jgr...@ns.sol.net said:
  FreeBSD used a 64-bit time_t for the AMD64 port pretty much right
  away. On the flip side, it used a 32-bit time_t for the Alpha
  port.  I guess someone predicted it wouldn't be a problem.
 
 Tru64 on Alpha uses a 32 bit time_t (they have their own time64_t and
 time64() call), so I expect *BSD and Linux on Alpha stayed with 32 bit
 time_t for compatibility (Linux at least could run many Tru64
 binaries).

NetBSD has just converted its -current branch to 64-bit time_t; I'm
pretty sure that that includes the Alpha port.


--Steve Bellovin, http://www.cs.columbia.edu/~smb