From baker@dad.cs.fsu.edu Fri Sep 25 16:21:38 1998 Received: from dad.cs.fsu.edu (dad.cs.fsu.edu [128.186.121.23]) by dkuug.dk (8.8.7/8.8.7) with ESMTP id QAA25685 for ; Fri, 25 Sep 1998 16:21:37 +0200 (CEST) (envelope-from baker@dad.cs.fsu.edu) Received: by dad.cs.fsu.edu (8.8.8+Sun/SMI-SVR4) id KAA05089; Fri, 25 Sep 1998 10:20:41 -0400 (EDT) Date: Fri, 25 Sep 1998 10:20:41 -0400 (EDT) From: Ted Baker Message-Id: <199809251420.KAA05089@dad.cs.fsu.edu> To: OblingerJT@csd.npt.nuwc.navy.mil, sc22wg15@dkuug.dk Subject: Re: (wg15tag 2129) (SC22WG15.1321) FW: (SC22.1318) REQUEST TO WG CONVENERS [Fwd: FW: Year 2k] This cannot be answered without some better definitions. The subject came up in some working group discussions at the last IEEE PASC meeting. There was confusion over what exactly "year 2000 compliant" means. Suppose there is some agreed-upon definition of what Y2K compliant means for a given system (thought that itself is interesting). There were still two different interpretations of how Y2K compliance extends to a standard like POSIX, that specifies certain properties of a system while leaving other details to the implementor: 1) To be Y2K compliant, the standard must not specify anything that would prevent a conformant implementation from being Y2K compliant. 2) To be Y2K compliant, the standard must specify enough details to guaranteee that all conformant implementations must be Y2K compliant. My own view is that it is not practical to require a standard to go beyond (1) above. There is another question, related to the definition of Y2K compliance in general. Is it sufficient for a system to be able to correctly handle dates and time intervals within some modest sized interval around 1 January 2000? Or, does this issue generalize to other arbitrary date limitations? --Ted Baker