J3/97-247 Date: 21 October 1997 To: J3 From: Richard Maine Subject: J3/97-007r2 I was hoping to avoid this. Its going to be confusing. Best laid plans and all that. I thought I/we was/were done with the f95 draft (except for "details" like interp processing) when I sent it out in June. But ISO had a few trivial pieces of editorial stuff that they wanted. I took care of them after some discussion, including some consultation with Miles, and I gave the (hopefully final this time) copy to our secretary to mail off to Switzerland today or tomorrow. The changes are pretty small and none of them have any technical impact, but they do mean that 97-007r0 is no longer a 100% accurate copy of the content of the official standard. Therefore, I've made a 97-007R2. PLEASE NOTE that 97-007r2 is a revision of 97-007r0, the previous f95 draft. It is *NOT* a revision of 97-007r1, which is the first working draft of f2k. When the next revision of f2k comes out (and it will incorporate these changes), it will presumably be 98-007r0. Have I confused you enough yet? I had hoped to avoid this overlap of the f95 and f2k drafts. Unless you particularly like killing trees, there will be no need to reprint most of the document. I'll describe the changes and the parts that I recommend printing here. There are *NO* changes, not a single word, in sections 3-14 or annex A-C. If you want to print them to get the revised page headings, feel free, but that's all that changed. Section (um, I mean clause) 2 has exactly 1 word changed. On page 13, in the first sentence of section 2.3.2, "Section 2.1" changed to "subclause 2.1". (They originally wanted to change it to just "2.1", but relented when I pointed out the confusion that would result from this being right below table 2.1 - we had used a different form for this citation specifically to avoid this confusion). This changes a few linebreaks in the rest of the para, but nothing of consequence. I'd suggest just marking the 1 word change. Section 1 has a 1-word change in 1.9 ("upon"->"on"), some reformatting and a title correction to the references in 1.9, and some trivial changes in the style of self-reference in section 1.1 (things like always spelling out ISO/IEC 1539 instead of just saying 1539). Of more consequence, the formal title was added at the top of page 1, and this takes enough space to change pagination throughout section 1. I'd suggest reprinting section 1, not for the wording changes, but because of the changed pagination (or you'll end up telling people to look on page 3 for something that is really on page 4). The introduction has a few places where "this standard" was changed to "this part of ISO/IEC 1539" (this being before we define the term "this standard" in section 1.1 - they wanted to make a simillar change globally, but I objected to that on numerous grounds; they acceded to my objection, but as part of a compromise, I use their form in the intro, which is prior to where the term "this standard" is defined in 1.1). The intro was also pushed to a new page separate from the forward, which changes pagination throughout the intro. I'd recommend reprinting the intro to get the page numbers right. They completely rewrote the forward, which I didn't like, but didn't feel worth arguing about. (The part I didn't like is that it no longer mentions either WG5 or J3; the "lowest" level that gets mentioned is SC22; other than that, there's nothing wrong with their rewrite). Its still included in the same source file as the intro, so if you reprint the intro, you'll also get the new forward. And you'll want to reprint the table of contents and index to correctly reflect the pagination changes in the intro and section 1. (I didn't bother to check for specific changes in that regard, but there could be some). P.S. I asked them to spell it "Fortran" instead of "FORTRAN" on the cover; the draft cover that they sent me had it all caps. -- Richard Maine maine@altair.dfrc.nasa.gov