From owner-sc22wg5@dkuug.dk Tue Nov 19 15:45:47 2002 Received: (from majordom@localhost) by dkuug.dk (8.9.2/8.9.2) id PAA60161 for sc22wg5-domo; Tue, 19 Nov 2002 15:45:47 +0100 (CET) (envelope-from owner-sc22wg5@dkuug.dk) X-Authentication-Warning: ptah.dkuug.dk: majordom set sender to owner-sc22wg5@dkuug.dk using -f Received: from mailhub1.liv.ac.uk (mailhub1.liv.ac.uk [138.253.100.94]) by dkuug.dk (8.9.2/8.9.2) with ESMTP id PAA60155 for ; Tue, 19 Nov 2002 15:45:40 +0100 (CET) (envelope-from j.l.schonfelder@liverpool.ac.uk) Received: from ts134044.dial.liv.ac.uk ([138.253.134.44]) by mailhub1.liv.ac.uk with esmtp (Exim 3.36 #1) id 18E9dH-00001E-00; Tue, 19 Nov 2002 14:45:43 +0000 Date: Tue, 19 Nov 2002 14:44:46 -0000 From: "J.L.Schonfelder" To: j3@j3-fortran.org cc: "SC22/WG5 members" Subject: Re: (j3.2002-752) Re: Public comment on Fortran: ISO/IEC CD 1539-1 Message-ID: <1427509.1037717085@ts134044.dial.liv.ac.uk> In-Reply-To: <200211190134.gAJ1YZsh020998@math.jpl.nasa.gov> References: <200211190134.gAJ1YZsh020998@math.jpl.nasa.gov> Originator-Info: login-id=jls; server=pop1.liv.ac.uk X-Mailer: Mulberry/2.2.1 (Win32) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Sender: owner-sc22wg5@dkuug.dk Precedence: bulk --On 18 November 2002 17:34 -0800 Van Snyder wrote: snip> > I'm sympathetic to deleting the Varying Strings part now that we have > nonkind type parameters that can be given values during allocation, and > LEN is a nonkind type parameter. I think this covers all of the uses of > Varying Strings. This is a comment on this and Keith's original. 1. as a matter of procedure the future of the varying string part 2 standard is not a question that forms part of the public comment on F2K part-1. 2. the existance of LEN typeparameters for derived types is not relevant. They merely allow the user to write a module that implements a type similar to the existing intrinsic character type, length set at declaration. The varying string standard defines a type that is fully dynamic in that the length is determined during execution by the values. 2. F2k would make it possibly to extend the varying string standard and module significantly. The basic type could be replaced with one depending on allocatable components (This I have already done to experiment with compilers that implemented the TR). This does not of course require any change to the standard. However, the addition of DTIO in F2K means that this could be added providing a much better interface to the user. 3. At least one compiler, NASoftware's Fplus, implements the varying string standard as an integral part of the product and hence providing greater efficiency than the illustrative module. I would vote for the part-2 standard's retention and upgrade once F2K is ratified. > snip> -- Lawrie Schonfelder Honorary Senior Fellow University of Liverpool 1 Marine Park, West Kirby, Wirral, UK, CH48 5HN Phone: +44 (151) 625 6986