Minutes of the December 2, 2002 IEEE 1364 working group meeting ---- Preamble ---------------------------------------------------- NEXT MEETING DATE: -----> January 13, 2003 <----- NEXT MEETING TIME: 08:30 PST 11:30 EST 18:30 IST ------------------------------------------------------------------ Attendence: Paaaaa Anders Nordstrom P aaa Clifford E. Cummings P aaaa Dennis Marsa - aa Erich Marchner Paa aa Gordon Vreugdenhil Paaaaa James A. Markevitch Paa aa Karen Pieper Paaaaa Michael McNamara Paa aa Steven Sharp Paa a- Charles Dawson Paaaa- Shalom Bresticker P aaa- Stefen Boyd Pa aa- Stuart Sutherland -a aa- Tom Fitzpatrick - -a Peter Flake P a Brad ------------------------------------------------------------------ came to order at 9:19 am Updates: 1: Sent 1364-1995 frame to Shalom 2: AR/Mac: Can we have a live Errata page? Ask Paul, Andy. 3: discussion on split standard into two parts: LRM, PLI. Decice to wait to see how discussions with Accellera proceed. 4: System Verilog/1364: DIscussion on how we see working together. Observation: Shalom: 2001 has taken _forever_ for vendors to implement; SV will take even longer. Steve Sharp: There are subsets of SV that are more useful to mainline hardware designers; perhaps folks will implement these partially. Charles Dawson: We should focus on preventing divergence. Issues are coming up in SV where they are going differently than 1364. Because 1364 & System Verilog have run concurrently, there are small differences that have naturally occured which need to be addressed as soon as possible. Specifically: In SV: initialization happens before time 0, but in 1364 it is at time zero. Also: 1364 has @*; but SV has always_comb These are very similar, but different. To serve our users, we need to focus on these quickly. Some pieces valuable; some not very useful. If we bring it all on, we will overwhelm the users and the vendors. Perhaps SV can exist as a roadmap? Stefen Boyd: As quickly as possible, finish SV; then 1364 picks through it for kernels of value, and add to later revs of 1364? What areas should we use to communicate? There will come a day when we will have to address the 1364/SV integration. Is this Accellera? or is this IEEE? If it is Accellera, they need to start. It appears that this is not in Accellera's plans. So if they expect IEEE to do this, then they need to do this now. (as soone as possible). AR/mac: Draft a letter asking for all of SV3.0 so that we can begin the work. Vendors can not take this amount of stuff. We want users input on features; but figuring out what we Get PAR for generic update of 1364-2001 that allows Karen proposes the following: ETF Issue 50 VSG approved unanimously, with and amendment of adding the following: First sentence of the example in sec 19.5 reads The examples of legal comments are as follows: replace with: Examples of include compiler directives are as follows: ETF Issue 155 VSG approved with one abstention. Abstain: shalom ETF Issue 171 VSG approved unanimously. Next meeting will set agenda for operations in 2003, based on feedback from Accellera. Set next meeting of Working Group to be Jan 13th. At 10:22 meeting of Working Group finished. ETF resumed