From 7838484205359439263 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: f78e5,e94695845965b249 X-Google-Attributes: gidf78e5,public X-Google-ArrivalTime: 1992-12-19 15:59:34 PST Newsgroups: comp.std.c++ Path: sparky!uunet!munnari.oz.au!metro!extro.ucc.su.OZ.AU!maxtal From: maxtal@extro.ucc.su.OZ.AU (John MAX Skaller) Subject: Re: Zero-length structures and pointer comparisons Message-ID: <1992Dec19.203242.21236@ucc.su.OZ.AU> Sender: news@ucc.su.OZ.AU Nntp-Posting-Host: extro.ucc.su.oz.au Organization: MAXTAL P/L C/- University Computing Centre, Sydney References: <1992Dec07.222242.18703@microsoft.com> <1992Dec8.103218.27689@lth.se> Date: Sat, 19 Dec 1992 20:32:42 GMT Lines: 31 In article kanze@us-es.sel.de (James Kanze) writes: >|> alternative is of course to dismiss the idea as not worth the trouble. > >I favor the latter case. I think that MANY collections will be collections of pointers. MANY of these types of collections will rely on there existing some total ordering. SOME of these will also require some that p==q <==> ptrcmpeq(p,q)==0 where I used ptrcmpeq to denote the version of ptrcmp that supports the above identity, whereas plain ptrcmp need not, and might always return 0. (A third version, ptrcmppo might also preserve partial ordering of <.) Therefore I think it is worth persuing the issue, especially as the 'ptrcmp' idea is an extension to the library, and thus might be acceptable to the committee, whereas changing the status of < might not. In fact I believe the status of == must be downgraded, even at the expense of C compataibility, and that will be a tough enough battle as it is. -- ;---------------------------------------------------------------------- JOHN (MAX) SKALLER, maxtal@extro.ucc.su.oz.au Maxtal Pty Ltd, 6 MacKay St ASHFIELD, NSW 2131, AUSTRALIA ;--------------- SCIENTIFIC AND ENGINEERING SOFTWARE ------------------