J3/03-243r1 Date: 21-aug-2003 To: J3 From: Toon Moene Subject: Partial response to 03-214/N1557 ------------- Page and line references are to 03-007. Paper 03-214/N1557 contains WG5 comments and recommendations concerning the issues raised in paper 03-201/N1524, which in turn contains comments from the Editor on several papers from J3 meeting 164. This paper replies to the comments and recommendations concerning papers 03-119r1, 03-120r1, and 03-172r1. ======================================== Part I (reply to comments in 03-201/N1524 on 03-119r1, excluding those covered in 03-213) The sentence in [393:8-9] ends with "and enumerations (4.6) that correspond to C types". The C standard defines enumerations to be a set of named integers, the same as the Fortran definition. Since the C standard states "Each distinct enumeration constitutes a different .", a set of named constants can be considered a C type in this circumstance. The current text is OK. No edit. The C code in [486:37-38] provides a C prototype that is supposed to interoperate with a Fortran subroutine interface on the next page. However, the arguments do not correspond. There is additional C code and descriptive text on pages 488 and 489 that is non-conforming to C rules. Edits are supplied to fix these problems. Edits: [486:37-38] Replace by: "int C_Library_Function(void *sendbuf, int sendcount, int *recvcounts);" [488:41] Add semicolon at the line end. [489:2] Replace "void *" by "void ". [489:3] Add semicolon at the line end. [489:7] Replace "function" by "void function". [489:7] Replace "with the name simulation, which returns a pointer to void" by "\cf{simulation}". ================================================================== Part II (reply to comments in 03-201/N1524 on 03-120r1) There is a grammar error in Note 15.12, [400:8+2]. Edit: [400:8+2] Replace "has" by "have". ================================================================== Part III (reply to comments in 03-201/N1524 on 03-172r1) The comments propose to replace the following text at [182:12-13]: "Input/output to and from external files can be performed without interference, as long as procedures defined by different means do not do input/output on the same external file." with a copy of the text at [405:27-28]: "If a procedure defined by means of Fortran and a procedure defined by means other than Fortran perform input/output operations on the same external file, the results are processor dependent." The replacement text is a better statement of the case of input/output operations on the same file. However, the replacement text does not mention anything about the case of different files, which is covered in the current text. The edit below addresses both cases. Edit: [182:12-13] Replace "Input/output...external file." by "If a procedure defined by means of Fortran and a procedure defined by means other than Fortran perform input/output operations on the same external file, the results are processor dependent. A procedure defined by means of Fortran and a procedure defined by means other than Fortran can perform input/output operations on different external files without interference."