From maine@altair.dfrc.nasa.gov Thu Jan 18 18:07:15 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 SAA17585 for ; Thu, 18 Jan 1996 18:07:08 +0100 Received: by altair.dfrc.nasa.gov (5.0/SMI-SVR4) id AA08414; Thu, 18 Jan 1996 09:07:25 +0800 Date: Thu, 18 Jan 1996 09:07:25 +0800 Message-Id: <9601181707.AA08414@altair.dfrc.nasa.gov> From: Richard Maine To: sc22wg5@dkuug.dk In-Reply-To: <199601091848.TAA14184@dkuug.dk> (janshep@watson.ibm.com) Subject: Re: (SC22WG5.991) edits to Fortran 95 draft standard. content-length: 4660 I have reviewed the editorial content of all of the f95 edits that Janice collected and posted in sc22wg5.991. These were f95 edits from interpretations that passed the WG5 ballot in November and then subsequently passed the latest X3J3 letter ballot. My review was restricted to editorial questions. Some of the items were technically controversial, but as editor I of course defer to the majority instructions of x3j3 and wg5 as expressed in the vote outcomes. The instructions were clear and I verified that I could find the referenced text for all of the changes. I have added all of the edits to my list of proposed changes for f95, with the following two changes (well, really four because each occurs twice) to the edits for defect item 27. 1. In the edits for case (iii) and case (v), I propose changing "not...nor" to "neither...nor". Although the "not...nor" is clearly an improvement over the earlier confusing "not...or", standard usage prefers the "neither". Janice said that she found other uses of "not...nor" in the standard. That is certainly possible, but it probably means that I failed to notice them. However, the relatively minor preference for "neither" doesn't justify spending time to track down and eradicate all cases of "not...nor". 2. The edits for case (ii) and case (iv), which I quote in whole below, contain the parenthetical phrases (there may be only one) and (there may be more than one). > "Case (ii): If TARGET is present and is a scalar target, the > result is true if TARGET is not a zero-sized storage > sequence and the target associated with POINTER occupies > the same storage units (there may be only one) as TARGET. > Otherwise the result is false. If the POINTER is > disassociated the result is false. > Case (iv): If TARGET is present and is a scalar pointer, the result > is true if the target associated with POINTER and the > target associated with TARGET are not zero-sized storage > sequences and they occupy the same storage units (there > may be more than one). Otherwise the result is false. > If either POINTER or TARGET is disassociated, the result > is false. I propose to just delete these parenthetical phrases. If they aren't deleted, they at least need some rewording. I tried to reword them, but I found that I couldn't figure out for sure what they were trying to say, which made it hard for me to do the rewording. First, I couldn't understand why the two versions of the parenthetical phrase were different. In both cases, as far as I can see, there might be any number of storage units >= 1 (zero is excluded). Both cases are for scalars, but the scalars might be of a type that has more than one storage unit (complex, double precision, character*x, or a derived type). Does the difference in wording imply some distinction that I don't appreciate? This is what most confused me. Second, the word "may" implies permission, both in standard usage and by ISO directive 3. I might add myself, that the combination of "may" and "only" can be very confusing because it can mean completely different things depending on the exact placement of the words in the sentence. I've been known to give up and rewrite sentences that used them because often find it difficult to make such sentences unambiguously clear. In particular, I could easily read "there may be only one" to imply that it was not allowed for there to be more than one, which I don't think was the intent here. (If it was the intent, then why was "units" plural?) I won't swear that my reading is correct, but I am very sure that it is a natural misinterpretation. There seems little point in having a parenthetical phrase that is meant to clarify but serves to confuse. I found the sentence a lot easier to understand without the phrase at all; the normal use of the phrase "the same storage units" allows for the possibility of one or more than one without elaboration. If thought necessary, I suppose we could say "the same storage unit(s)", but thats a pretty ugly and awkward construct that doesn't seem necessary. If we do keep the parenthetical phrases, could we at least change "may" to "might"? I believe that possibility (might) is intended instead of permission (may). And can someone explain why the two are different? -- Richard Maine maine@altair.dfrc.nasa.gov