J3/16-011r1 To: J3 From: Malcolm Cohen Subject: Editor's list of unresolved technical issues Date: 2016 May 02 A. Introduction --------------- This document lists the technical issues raised by the editor with the current working document, standing document 007. Issues are separated into two lists; the resolved ones and the unresolved ones. When an issue is resolved, the J3 note in SD-007 will be deleted and the item moved from the unresolved list to the resolved list. Issue numbers will not be reused. The one-line description is just a hint as to the issue, not a comprehensive statement of it. B. Unresolved issues -------------------- 016. (DE)ALLOCATE mess re STAT=. Feature: Failed images. Added in: 16-007r1 as a result of processing TS 18508. 017. Same execution is not sufficiently well-defined. Feature: Teams. Added in: 16-007r1 as a result of processing TS 18508. 018. CRITICAL semantics do not permit image failure. Feature: Failed images. Added in: 16-007r1 as a result of processing TS 18508. 019. Coarray association incompletely specified. Feature: Teams. Added in: 16-007r1 as a result of processing TS 18508. 020. Can expressions in CHANGE TEAM do image control? Feature: Teams. Added in: 16-007r1 as a result of processing TS 18508. 021. Implicit synchronisation badly specified here. Feature: Teams. Added in: 16-007r1 as a result of processing TS 18508. 022. Incompatible change of semantics. Feature: SYNC ALL (not just failed images). Added in: 16-007r1 as a result of processing TS 18508. 023. Glitches in SYNC IMAGES. Feature: Teams and failed images. Added in: 16-007r1 as a result of processing TS 18508. 024. SYNC TEAM shares flaws with SYNC ALL and SYNC IMAGES. Feature: Teams. Added in: 16-007r1 as a result of processing TS 18508. 025. STAT= with a positive value needs to be an error indication. Feature: Failed images. Added in: 16-007r1 as a result of processing TS 18508. 026. There is no such thing as a "whole coarray". Feature: Collective subroutines. Added in: 16-007r1 as a result of processing TS 18508. 027. Involves synchronization with a stopped image? Feature: Collective subroutines. Added in: 16-007r1 as a result of processing TS 18508. 028. Why should FAILED_IMAGES nanny the user? Feature: Failed images. Added in: 16-007r1 as a result of processing TS 18508. 029. Resolving ambiguity in THIS_IMAGE. Feature: Teams. Added in: 16-007r1 as a result of processing TS 18508. 030. There is, and should be, no concept of an "error image". Feature: Failed images. Added in: 16-007r1 as a result of processing TS 18508. C. Resolved issues ------------------ 001. Ambiguous syntax. Feature: Pointer function references as variables. Resolved in 14-007r1 by 14-115r2. 002. ASSOCIATE ok for plain procedure pointer? Resolved in 14-007r1 by 14-114r1. 003. Cannot count. Feature: Editor. Resolved in 14-007r1 by not existing in the first place. 004. Formal parameters in a prototype cannot have any value whatsoever. Feature: Further C interop (TS). Added in: 14-007r1 as a result of processing N1942. Resolved in 15-007 by 14-197r2. 005. An effective argument is not EVER an address. Feature: Further C interop (TS). Added in: 14-007r1 as a result of processing N1942. Resolved in 15-007r1 by 15-113r3. 006. An interoperable procedure written in Fortran does not even have any formal parameter, so it cannot have the value of a null pointer. Feature: Further C interop (TS). Added in: 14-007r1 as a result of processing N1942. Resolved in 15-007 by 14-197r2. 007. Fortran objects do not have a "host instance". Feature: Further C interop (TS). Added in: 14-007r1 as a result of processing N1942. Resolved in 15-007 by 14-197r2. 008. Is the processor required to define a nonnegative type specifier value for the above cases? Feature: Further C interop (TS). Added in: 14-007r1 as a result of processing N1942. Resulted by: 15-200r2 in 15-007r2. 009. shall contain ... macro definitions that expand to integer constants. Feature: Further C interop (TS). Added in: 14-007r1 as a result of processing N1942. Resolved in 15-007 by 14-197r2. 010. RANDOM_INIT example text claim is inconsistent with RNG definition. Feature: Intrinsic function RANDOM_INIT. Added in: 14-007r2 as a result of processing 14-184r4. Resolved in: 15-007r1 by 15-140r1. 011. GET_ENVIRONMENT_VARIABLE has no example. Feature: Intrinsic subroutine GET_ENVIRONMENT_VARIABLE (base F2003). Added in: 15-007 as a side-effect of processing 14-247r3. Resolved in: 15-007r1 by 15-112r4. 012. Term "actual argument" wrong, nonsensical "effective argument". Feature: Further C interop (TS). Added in: 15-007r1 as a result of review of the UTI 005 resolution. Resulted by: 15-200r2 in 15-007r2. 013. Accessibility ambiguity. Feature: Module names in accessibility statements. Added in: 16-007 as a result of processing 15-247r1. Resolved in 16-007r1 by 16-157r1. 014. Another accessibility ambiguity. Feature: Module names in accessibility statements. Added in: 16-007 as a result of processing 15-247r1. Resolved in 16-007r1 by 16-157r1. 015. Attributes wrong for LOCAL and LOCAL_INIT. Feature: Locality specs. Added in: 16-007 as a result of processing 15-246r1. Resolved in 16-007r1 by 16-106r2. ===END OF DOCUMENT===