Phone Conference P1450.1 Working Doc Subgroup Thurs Jan 17, 10:00 am PST --------------------------------------------- Attendees: ---------- Jim Teisher Doug Sprague Tony Taylor Greg Maston Peter Wohl Agenda ------ - Review 1450-1999 (dot0) clarification additions for placement on website. - Continue resolution of internal review cycle and finalize draft 13 document containing reviewer resolutions. Focus on red text in resolution document (indicating open issues). Issues ------ DM1 - Doug would like hierarchical scan in .1; Tony identified that this was addressed by putting chain_names into the cell_name space. This construct is identified in section 6.3, however in getting there we identified section 14.3 is incorrect for this new definition. Doug will review and confirm that this extension is sufficient. Tony took [AI1] to correct section 14.3 for the new constructs. Discussion about the concept of defining a STILKeywords block, that allows "partial selection" of the constructs in a dotted standard. Greg is concerned that to support this properly, a parser would need to now scope language support not at the dot-level but at each construct in the dot, which would complicate parsing for "what is legitimate" immensly. Also, each dot is designed to be consistent only over the set of all definitions in that standard; by attempting to load a subset of constructs each construct must be logically independent -- and that's not the case in general. This discussion was tabled because the issue it was in response to (DM1) appears to have been addressed. Discussion about the use of commas in this extension. Currently there is inconsistant handling between integer_list and cellname_list constructs, one of which supports commas and one that doesn't. After review of dot-zero (which does not support commas at all), it was decided to be consistant across all the standards, so resolution is to remove commas from this spec. [AI2] to Tony to remove from existence the use of commas in dot1, as if they never were. JC2 - Final resolution: response from the Working Group is to support all new capabilities of Vectors in a Shift, again for consistancy's sake. The last paragraph in the resolution here, about what is allowed inside a Shift, Greg took [AI3] to resolve. This is still open from [AI11 last meeting]. DS4 - greg's [AI4] still open from [AI4 last meeting] DS7 - doug's [AI5] still open from [AI5 last meeting] DS14 - still open from [AI12 last meeting]. Doug's email. Issue of localization of an X statement during pattern execution has to do with execution sequence of a pattern set, and how to uniquely identify the failing location of a pattern set given the execution. This is a general traceback problem of execution flow; while some support can be defined here the general problem of execution-traceback IS a general problem. Tony took [AI6] to put Doug's email on this topic into the resolution doc. DS15 - greg's [AI7] still open from [AI7 last meeting] Greg to distribute current dot1 recommendation doc. DS18 - This issue resolved with Doug's update. --- End of identified open issues from the resolution doc. Dot-0 clarification review -------------------------- The dot-0 clarification doc had two modifications during the week, due to comments from Peter and Doug. Greg [AI8] will distribute the latest version of this document for resolution next meeting. Other open issues from last meeting ----------------------------------- [AI6 last meeting] Tony and Greg to clarify behavior of variable assignment. The 'sticky behavior' issue is still open, assigned [AI9] to Tony. New issues ---------- None raised. Next Meeting ------------ Doug wants a face-to-face in Hawaii, warmth in exchange for snow. Next meeting January 31, 2002. On the phone. Sorry, Doug. AIs --- [AI1] Tony correct section 14.3 to remove InheritScanStruct. [AI2] Tony remove commas from dot1. [AI3] Greg clarify what goes inside a Shift (dot-0 issue). [AI4] Greg to rewite last paragraph of section 6.1 [AI5] Doug to provide expression-type-clarification diagram or text. [AI6] Tony to put Doug's X-stmnt-position-add't into resolution doc. [AI7] Greg to distribute current dot1 recommendation doc. [AI8] Greg to distribute current dot0 clarification doc. [AI9] Tony to specify 'sticky behavior' of STIL variables during execution.