From Miles.Ellis@etrc.ox.ac.uk Fri May 17 11:17:38 1996 Received: from oxmail3.ox.ac.uk (oxmail3.ox.ac.uk [163.1.2.9]) by dkuug.dk (8.6.12/8.6.12) with SMTP id LAA23921 for ; Fri, 17 May 1996 11:17:24 +0200 Received: from vax.ox.ac.uk (actually host vax) by oxmail3 with SMTP (PP); Fri, 17 May 1996 10:17:01 +0100 Received: from 163.1.85.1 by vax.ox.ac.uk (MX V4.2 VAX) with SMTP; Fri, 17 May 1996 10:16:57 +0100 X-Sender: MELLIS@vax.ox.ac.uk Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 17 May 1996 10:17:46 +0100 To: "j.l.schonfelder" From: Miles.Ellis@etrc.ox.ac.uk (Miles Ellis) Subject: Re: (SC22WG5.1080) Varying string standard CC: sc22wg5 Lawrie, Thank you for your comments. > >It would appear that the extensions made to the Language to produce Fortran 95 >do not neccessitate any changes to the actual normative sections of the >standard. I am glad to hear that, since we only have approval to revise the anexe and not the normative text! >They would, however, allow modifications to be made to the example >module which would make it more memory efficient. These are changes to the >definition of the type so that by default all strings are created with the >pointer component diassociated. The "defining" procedures such as the >assignment and input subroutines can then safely test for the association >status of the argument which is to become defined in the invocation and free >any memory by dallocation if the string is currently allocated. > >The required changes amount to about 10 lines of changed code in the source of >the module. In addition there are a small number of errors in the published >document which should be corrected. Again I am aware of only a couple of code >errors in the example program VOCAB.F90 that need correcting. > >Given the small number of changes required it is recommended that the base >standard is not republished but an addendum document is produced indicating >the necesary editorial changes to the existing published document. The >existing unammended document will continue to be the Fortran 90 auxilliary >standard and the existing document plus addendum will provide the Fortran 95 >version of the auxilliary standard. > That is an interesting concept. The problem from an ISO perspective, of course, is that once Fortran 95 has been approved then Fortran 90 ceases to exist. Thus the existence of, in effect, a twin standard, one half of which refers to a non-existent language (from a pedantic perspective) might cause problems. I shall take advice from the SC22 Secretariat. >An updated version of the source code can be found on the web. >http://www.liv.ac.uk/~jls/iso_vst.f95 > I shall put a copy of this on the WG5 server in the next few days and ask WG5 members to review it. As you imply, we probably shouldn't send it to SC22 (in whatever form) until you have been able to test it on a Fortran 95 compiler - after all, it is only the informative part of the standard that is changing, and it is rather important that when a new example module is published that it is correct! Miles ---------------------------------------------------------------------- Dr Miles Ellis Director: Educational Technology Resources Centre University of Oxford, 37-41 Wellington Square, Oxford OX1 2JF, ENGLAND Telephone: +44 1865 270528 Fax: +44 1865 270527 Email: Miles.Ellis@etrc.ox.ac.uk WWW: http://www.etrc.ox.ac.uk/Personal/Miles_Ellis.html