Minutes of ISAC meeting held via telecom on 02 May 2007 Present: Peter Ashenden, Jim Lewis, Larry Soule, Chuck Swart, Ajay Verikat Absent: Lance Thompson Next Meeting Wednesday June 13, 2007, 8 pm Pacific Daylight Time (Thursday, June 14, 2007, 3 am GMT ) TOPIC: IR 2112 Can attributes be applied to a signal on the entity within the architecture for that entity? There was one negative vote from Jim, arguing that the language needs to be extended to allow attribute specifications to be used in a wider set of cases than is now the case. All other votes were positive. The ISAC agreed that this issue should be forwarded to the requirements committee. ACTION: Chuck to incorporate Jim's comments into the IR and forward it to requirements group. TOPIC: IR 2113 FILE Input and ROM Inference and IR 2114 1076.6-2004 section 8.6.5 Slice names is too restrictive. Both of these IR are requirements for the now-defunct 1076.6. It was decided to forward these requirements to the DASC and recommend that the DASC give responsibility for 1076.6 to the VASG. The ISAC will also recommend that the VASG have oversight on 1076.4, VITAL. The various folks responsible agree with this plan. ACTION: Chuck to contact DASC with this proposal, to cc Ajay (chair of VITAL) and Jim (chair of VASG) TOPIC: IR 1070 VPI Issue 14 -- Prefixes in USE clauses No activity. TOPIC: IR 2099 Alias declarations introduce homographs John Ries was contacted, and he more or less agrees with the current analysis. The IR has been recently revised to include one problematic case (EXAMPLE 5b). There was a related issue for P1076-2006/D3.0 which was entered as Bugzilla 131. It was agreed to separate the two issues and treat IR 2099 independently.ISAC members should carefully review the latest analysis to see if any unexpected conditions occur in practical cases. ACTION: ALL to review and vote. TOPIC: IR 2110 Implicit subtype conversions not defined. The two main requirements are to update the clause on type conversions to include implicit conversions, and to change the clause on variable array assignments to properly cover slices and unconstrained subtypes. Chuck asked whether wording should be changed for composite signals as well. Peter pointed out that signal assignments are decomposed into scalar assignments of matching elements, so this case is already adequately covered. ACTION: Chuck to analyze. TOPIC: Review of IRs incorporated into D3.0 2083 2006-042 Lance OK 2084 2006-043 Lance Possible issue 2086 2006-047 Lance OK Lance pointed out that some VHPI changes use terminology such as "NumFields" where "NumElements" would be preferable. It was decided that the terminology differences were very minor, that the intent was clear, and that backward compatibility issues were more important than very minor editorial issues. Therefore, the ISAC decided that IR 2084 was properly incorporated into D3.0. TOPIC: Review of Pre-2002 IRs 1057 Lance All Issues Resolved 1058 Lance Open Issues 1071 Ajay 1072 Ajay 1074 Ajay 1075 Chuck Open Issues 1076 Chuck Open Issues 1077 Chuck All Issues Resolved 1079 Peter All Issues Resolved 1081 Peter All Issues Resolved 1082 Peter Open Issues 1085 Larry All Issues Resolved 1086 Larry All Issues Resolved 1088 Larry Open Issues ACTION: Pre-2002 IRs to be examined for next meeting: 1071 Ajay 1072 Ajay 1074 Ajay 1089 Larry 1090 Larry 1091 Larry 1092 Peter 1093 Peter All Issues Resolved 1094 Peter 1095 Chuck 1096 Chuck 1097 Chuck Also, Peter discovered documents dealing with IRs 1040,1081,1086, 1117, 1124. These include votes and apparently reflect changes which were not incorporated into the eda.org repository. ACTION: Chuck to update the repository to reflect these changes.