From j.l.schonfelder@liverpool.ac.uk Fri Mar 31 20:22:51 2000 Received: from mailhub2.liv.ac.uk (mailhub2.liv.ac.uk [138.253.100.95]) by dkuug.dk (8.9.2/8.9.2) with ESMTP id UAA31344 for ; Fri, 31 Mar 2000 20:22:46 +0200 (CEST) (envelope-from j.l.schonfelder@liverpool.ac.uk) Received: from pcmail1.liv.ac.uk ([138.253.252.13]) by mailhub2.liv.ac.uk with esmtp (Exim 2.12 #1) id 12b64B-0006UE-00; Fri, 31 Mar 2000 19:22:43 +0100 Received: from [138.253.134.46] (helo=liv-dialup.liv) by pcmail1.liv.ac.uk with smtp (Exim 2.12 #1) id 12b64A-0000Dc-00; Fri, 31 Mar 2000 19:22:42 +0100 From: Lawrie Schonfelder Reply-To: j.l.schonfelder@liverpool.ac.uk To: "Kurt W. Hirchert" Cc: SC22WG5@dkuug.dk Subject: Re: (SC22WG5.1751) Interpretation 001 In-Reply-To: <3.0.5.32.20000330102640.007ec240@perseus.ccs.uky.edu> Message-ID: Date: Fri, 31 Mar 2000 19:21:19 +0100 (GMT Daylight Time) Priority: NORMAL X-Mailer: Simeon for Win32 Version 4.1.5 Build (43) X-Authentication: IMSP MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII I am entirely resigned to the problem you outline. I still believe we got it wrong in F90 but I think John's proposed fix is the least objectionable way of clarifying the situation. On Thu, 30 Mar 2000 10:26:40 -0500 "Kurt W. Hirchert" wrote: > At 11:02 AM 3/30/00 +0100, Lawrie Schonfelder wrote: > >I would be happy to have the inclusion of a name as an implied DO index > >defined as an explicit declaration of that name as an integer within the > >scope of the implied do. I would also have it independent of any declaration > >of the same name in the containing scope. As in host association within the > >implied do scope the index variable masks any variable of the same name from > >the outer scope. This would have been clean and consistent with every other > >language I know of that has nested scopes. It would be exactly similar to > >host association scoping rules. > > I see two potential traps in this: > 1. Note that I/O implied DO loops _don't_ use a different variable, so be > careful about how generally you state this. (If you were to make I/O > implied DO loops use different variables, you could break existing > standard-conforming f77 and f90 programs.) > 2. There remains the question of which KIND of integer the DO index is > declared to be. Default integer may be inadequate or inappropriate. I > suppose we could have invented some convention such as the DO index taking > its KIND from that of the initial value expression, but given that we > didn't, we must consider the cost of making an incompatible change to the > language (i.e., one that potentially would render conforming programs > nonconforming). > > > I am relatively comfortable with "fixing" the standard to eliminate phantom > exports, because I seriously doubt that anyone has caused one intentionally > (other than in test programs). I am much less comfortable with the > specific changes you propose, because they have other effects that do have > the potential to invalidate real programs. > -- > Kurt W. Hirchert hirchert@ccs.uky.edu > Center for Computational Sciences +606-257-8748 -- Lawrie Schonfelder Director, Computing Services Dept. The University of Liverpool, UK, L69 7ZF Phone: 44(151)794 3716, Fax: 44(151)794 3759