Phone Conference P1450.1 Working Doc Subgroup Thurs Dec 13, 10:00 am PST --------------------------------------------- Attendees: ---------- Tony Taylor Greg Maston Doug Sprague Peter Wohl Jason Doege Agenda ------ Review the p1450.1 spec, draft D13, Dec. 3, available on the dot1 website, with respect to the internal-review-resolutions doc also found on the web. Question about dot4 ------------------- Doug asked if any progress had been made on the dot4 restart. Tony took [AI1] to ping Dave to see what's going on. Issues ------ CW1 - look to be clarifications; uncertain how to clarify... Peter took [AI2] to respond on the clarification doc and to identify IF any changes to doc. CW2 - add footnote on headers to table 5 to refer to *_expr. ([AI3]) CW3 - complete TT1 - complete. Examples in doc available in plain-text; Tony will distribute to WG. Also need to check with IEEE again to put dot0 examples on website ([AI4]); asked before but never got a response. TT2 - complete TT3 - complete TT4 - scoping of constants, integers, and signalvars; constants and integers in the signals only, signalvars in groups only. Doug identified a concern that having the constants and ints in the groups does allow for the potential of incremental definitions or even an override mechanism, and that perhaps this was part of the motivation, but is very comfortable with this simplification. Jason asked what the motivation was for the original definitions (with constructs in both places); motivation was primarily a very-old notion of attempting to maintain a simple object model for STIL where the information in the groups and signals blocks are identical. Also a question was raised about whether the datastructures are sufficiently robust, for example, whether arrays of these constructs need to be supported. After this discussion the WG felt comfortable with the changes proposed. Tony mentioned that another location for this information is as part of the Spec constructs. Greg took [AI5] to make sure that this interaction is clearly identified in the spec - interaction of constants and integers (in particular) with spec constructs needs to be identified in this doc. TT5 - complete TT6 - extend the identifiers to indicate purpose of variables. Need to update the description on page 24 to reflect this proposal. Jason identified this as an extension to a concept of "type" - if the appearance of the variable does not reflect the usage specified, then it's an "error". Greg raised his standard objection with the "User" option on this statement, that functionally this construct removes any checking opportunity performed by the other fields and is in fact identical to having the user identify a separate Userkeywords statement with this information. After some discussion on whether this is (or can be) checked statically or dynamically, Peter proposed the syntax: Usage (< List of STIL block names >)+; Doug questioned the value-added of this construct (in general, not limited to Peter's extensions); original motivation was to identify variables that are "meant for simulation" and therefore other contexts could "ignore" them. Tony identified that the blocks syntax above satisfies everything except potentially "test" vs "non-test" applications, but we could not identify any differentiation that is not covered by the change above. Tony took [AI6] to implement this change. TT7 - complete DS14 - Doug raised this is a concern for him still and something that needs to be resolved. Discussion of testbench replacement ----------------------------------- Jason asked about the concerns of a "testbench replacement" scenario/application. He was referred to the individual who caused the removal of another proposed definition in the language with the justification that this effort was not meant to be a "testbench replacement", to gain some understanding of that perspective. Next Steps ---------- Meeting next week, Dec 20., same time and phone#. Continue working to the same Draft 13 for next meeting. AIs --- [AI1] Tony will ping Dave in regards dot4. [AI2] Peter complete CW1 responses in resolution doc. [AI3] add footnote on headers to table 5 to refer to _expr. [AI4] Tony check again w/IEEE copyrights to put dot0 examples on web. [AI5] Greg validate spec-block interactions with integer/cosntants. [AI6] Tony to change Usage statement per discussions.