These minutes are also available on the website. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. Minutes of ISAC meeting held via telecom on 10 October 2007 Present: Peter Ashenden, Larry Soule, Chuck Swart, Ajay Verikat Absent: Jim Lewis, Lance Thompson Next Meeting Wednesday, November 07, 8 pm Pacific Standard Time (Thursday, November 08, 4 am GMT) TOPIC: IR 2122 Protected method has implied object parameter? The analysis is accepted. The IR is ISAC-APPROVED. ACTION: Chuck to forward to VASG. TOPIC: IR 2123 Process resumption and callbacks Support for VHPI in 1076C resulted in a finer granularity in the simulation cycle. The LRM contains certain conflicting statements in the LRM about when and whether a process resumes. These conflicts had no meaningful effect until the VHPI changes were added. So the resolution of this issue depends, at least in part, on the needs of VHPI. This issue has uncovered a problem in resolving LRM issues. The LRM now contains many areas which are not part of the VHDL "core" and in which the ISAC does not necessarily have expertise. Such areas include VHPI, encryption, PSL, and the various packages in the IEEE library. It would be unreasonable to draft experts in these areas into the ISAC for the relatively infrequent occurrences of problems in these specialized areas. Instead, the ISAC recommends that the VASG keep standing committees with expertise in each specialty area. The ISAC will serve as a clearing house to refer specialty issues to the appropriate group or groups and to resolve conflicting technical solutions proposed by various groups. In addition, each specialty group will be expected to examine each new IR and identify those IRs which have an impact in its area. The ISAC will continue to resolve issues involving core language features. ACTION: Chuck to contact VHPI chair about this issue. Chuck to pass "specialty group" recommendation to VASG chair. TOPIC: IR 2124 Ordering of process execution and callbacks This IR identifies certain points in the simulation cycle which overly restrict process ordering from the perspective of simulation performance. These restrictions were added to support VHPI, and it may be the case that they were not intended by the VHPI, so response from that group is necessary to resolve this issue. ACTION: Chuck to contact VHPI chair about this issue. TOPIC: IR 2125 Resolved for std_ulogic is broken for '-' Peter has sent an email in response to this issue and Jim has emailed a reply. The submitter claims that a change to the resolution function to allow better '-' propagation would improve testbench writing. The ISAC is concerned that a backwards incompatible change to such a stable and long-lasting language feature could have undesirable consequences. In addition, for the particular application he seeks, the submitter could write his own resolution function with few changes needed in his code. New VHDL-200X features allow type-compatible resolution function changes to vectors as well as scalars. The submitter's comments show that he is trying to implement something like bundles or Verilog mod ports. This requirement has not yet been met by a language revision. The ISAC believes that significant input is needed from the user community before the change to the resolution function can be made, but that the requirements committee is the correct forum to gather and synthesize this information. ACTION: Chuck to prepare a preliminary analysis reflecting the emails and this discussion, circulate that analysis to the ISAC for review by 10/17, then forward the IR as a requirement. ISSUE: IR 2099 Alias declarations introduce homographs This IR has been extensively reviewed and analyzed. Its time for an ISAC vote. ACTION: ALL to vote on this IR by Wednesday 10/17. TOPIC: IR 2110 Implicit subtype conversions not defined This IR has been superseded by Bugzilla #174 as it currently stands. ACTION: Chuck to mark this "Superseded" when Bugzilla #174 is approved. TOPIC: IR 2120 How to access objects in higher level nested protected type This is ready for an ISAC vote. ACTION: ALL to vote on this by Wednesday, 10/17/2007. TOPIC: IR 2121 Allow for vectors to have assigns and opens in the port map Chuck's analysis of problems, especially those involving use of default expressions, is considered appropriate for the VASG but too confusing for comp.lang.vhdl. Ajay did not receive the analysis until a copy was sent at the meeting, so he was unable to review it adequately. Therefore, Chuck will incorporate the analysis into the IR, submit it to the ISAC for review, then forward it to the requirements committee. ACTION: Chuck to revise the IR, ALL to review by 10/17/2007, Chuck to forward to requirements.Received on Fri Oct 12 10:33:39 2007
This archive was generated by hypermail 2.1.8 : Fri Oct 12 2007 - 10:33:43 PDT