From JANSHEP@torolab2.vnet.ibm.com Fri Jan 6 06:25:59 1995 Received: from vnet.ibm.com by dkuug.dk with SMTP id AA19201 (5.65c8/IDA-1.4.4j for ); Fri, 6 Jan 1995 17:24:51 +0100 Message-Id: <199501061624.AA19201@dkuug.dk> Received: from TOROLAB2 by VNET.IBM.COM (IBM VM SMTP V2R2) with BSMTP id 5177; Fri, 06 Jan 95 11:24:17 EST Date: Fri, 6 Jan 95 11:25:59 EST From: "Janice Shepherd" To: sc22wg5@dkuug.dk Subject: defect item 185 X-Charset: ASCII X-Char-Esc: 29 Lawrie wrote: > >It is my view that with the publication of corrigendum 2 the iterpretation >processing is of limited relevance qua interpretation. We should now be treating >defects as things that we fix in f95. We do it right in 95 and stop worrying >about what we thought or intended in 198x when designing F90 except to the >extent that that might be relevant to getting the design right now.. > >I am happy to say this again to Jerry and/or WG5. If we are serious in intending >an F95 revision and I believe WG5 is, a third corrigendum to F90 is essentially >pointless. > Continuing the interpretation process for Fortran 90 is very important. I believe the planned schedule for Fortran 95 shows it reaching publication sometime in the second half of 1996. To stop processing Fortran 90 interpretations would in essence be the same as saying to all Fortran users "Please stop using Fortran 90 and just wait for Fortran 95 when all outstanding interpretations would have been made addressed appropriately". The point is the people are using Fortran 90 now; developers are encountering questions in the semantics of the language (either with existing implementations or yet to be shipped implementations). We need to answer the interpretations on Fortran 90 now; not in a version of Fortran that won't be available for almost two years (and how much longer for implementations?) On the other hand we are not blind to the fact that Fortran 95 is being developed. When working on an interpretation we always keep in mind: What is the appropriate answer for Fortran 90? Is that the same as what we want for Fortran 95? Can we apply a small edit now (or perhaps no edit at all) but give a better set of edits for Fortran 95? The continued processing of interpretations has been discussed extensively at the last WG5 meeting, and at the last two X3J3 meetings. So the "process" is being monitored. I believe the majority of members of the two committees continue to be in support of the interpretation processing. As to whether a third corrigendum to F90 will be developed, it has been proposed that a third corrigendum be developed if there are edits agreed to by X3J3 and WG5 that can affect user's programs or implementations. WG5 agreed to that proposal. I imagine this will be discussed again once we have a set of approved edits. Thanks. Janice C. Shepherd