🚀 go-pugleaf

 In article <827777192_OASIS_@swampdog.demon.co.uk>
  guy@swampdog.demon.co.uk (Guy Harrison) said:-

> In article <199603230101.AA01823@moroder.demon.co.uk>
>  Stephen Mulrine <robotron@moroder.demon.co.uk> said:-
>
>>In demon.ip.support.atari, guy@swampdog.demon.co.uk (Guy Harrison) wrote:
>>>The first line of nntp.dat looks like this ...
>>>news.demon.co.uk 960320 150222
>>
>>It's just occurred to me..  This strategy is only going to be good for
>>another four years, isn't it?  I wouldn't like to think what will happen
>>when the server thinks it's 000000 or later.  The very least you could
>>expect would be to lose all the news which arrived on the server between
>>the last time you dialled up in 1999 and the first time you dialled up in
>>2000.  Time for a fix, perhaps?
>
>Good point that. I bet an awful lot of software is going to break and 
>I have no idea what the solutions will be. Does anybody else know?
>
The solution is for the original programmers to have taken the fact into
account. Having said that how many programmers have done that is open to
speculation... We will all find out in year 00!

I once worked on the help desk of a large computer company, the bulk of the
software from which all fell over when we hit the 32768th day of the
century. In fact the bug had been clearly identified and the fix (setting a
variable to be a long integer rather than a short one) easy to do. They had
four years advance warning to do this. For many customers they didn't,
preferring to wait for the problem to hit, then charge them for fixing it.
Needless to say I handed in my resignation shortly after the 32768th day
when I realised the charges were being made. Hopefully the start of year 00
is a far more easy to recognise date than the 32768th day of the century,
so less people will get ripped off as a result of programmers lack of
foresight. 
--   
Brian