From maine@altair.dfrc.nasa.gov Sat Feb 3 19:17:10 1996 Received: from altair.dfrc.nasa.gov (altair.dfrc.nasa.gov [130.134.34.72]) by dkuug.dk (8.6.12/8.6.12) with SMTP id TAA24738 for ; Sat, 3 Feb 1996 19:17:05 +0100 Received: by altair.dfrc.nasa.gov (5.0/SMI-SVR4) id AA20237; Sat, 3 Feb 1996 10:17:19 +0800 Date: Sat, 3 Feb 1996 10:17:19 +0800 Message-Id: <9602031817.AA20237@altair.dfrc.nasa.gov> From: Richard Maine To: Miles.Ellis@etrc.ox.ac.uk Cc: sc22wg5@dkuug.dk In-Reply-To: (Miles.Ellis@etrc.ox.ac.uk) Subject: Re: WG5 requested corrections to Fortran 95 CD (FOR X3J3/136) content-length: 2894 Miles, I got the subject email and will bring at least an electronic copy to Vegas. I haven't yet had time to carefully review it, and I probably won't have time until I get to the Vegas meeting, but I did notice one thing that I have a question about. Alla Gorelik had some comments about the section headings in Chapter 13. The misprints that he noted were clearly errors and have been fixed. However, I disagree with his more general comment on the headings. The former comments on this area were not overlooked, but were discussed extensively. Section 13.3 describes the syntax used in these headings, and notes 13.2 and 13.3 give examples and elaboration. This approach was intentionally choosen over either of those mentioned by Gorelik. In particular > The first variant is: > SUM (ARRAY, DIM [,MASK]) or SUM (ARRAY [,MASK]) > The second variant is: SUM (ARRAY [[,DIM] [,MASK]]) There are significant semantic differences between these 2 variants. This was the subject of extensive discussion at the WG5 San Diego meeting and elsewhere. With the first form, DIM is not an optional argument; there are two specific interfaces, one with DIM and one without; neither has DIM as an optional argument. For SUM, this is the desired semantics. For most of the intrinsics, this is not the correct semantics; the subject arguments are optional. I won't try to go into the details of why the difference matters here; I'll restrict myself to pointing out that the subtle technical question involved was considered at length by WG5 and was reflected in the papers passed at the San Diego meeting. The second form is more or less technically correct for most of the intrinsics (but not for SUM). However, I feel that this syntax is too complicated for a section heading. For all of the complexity, it still does not explicitly reflect all of the possibilities (namely the use of keyword forms). There was an intentional decision to avoid this syntax, but to instead keep the simpler syntax and explain its interpretation in section 13.3. I'd have to research a bit to find exactly where this decision was made - I think it might have been done at the X3J3 Maui meeting. My question (finally) is whether these are comments to be considered or whether I/we are specifically directed to make the changes. The body of your message implies to me that these are comments to be considered. My personal feeling is that this issue has already been adequately considered and can be answered by pointing out the explanation in section 13.3. I'm also worried about getting some of the subtle points technically wrong - some of them are subtle. However, the subject line "WG5 requested corrections..." sounds like a more specific direction to make these corrections. I'm probably just reading too much into the exact words of the subject line. -- Richard Maine maine@altair.dfrc.nasa.gov