====================================================================== IBIS INTERCONNECT MODELING AD HOC TASK GROUP MEETING MINUTES AND AGENDA http://www.eda.org/ibis/adhoc/interconnect/ Mailing list: ibis-interconn@freelists.org ====================================================================== Next Meeting Wednesday, January 28, 2008 8 AM US Pacific Time Telephone Bridge Passcode 916-356-2663 3 297-7421 LiveMeeting: https://webjoin.intel.com/?passcode=2977421 Agenda: - Call for patents - Opens - Finish review of Rob Elliott's list of suggest changes ====================================================================== Minutes from January 21: Attendees: ---------- (* denotes present) Agilent - Radek Biernacki*, John Moore, Ken Wong Ansoft - Denis Soldo Cadence Design Systems - Terry Jernberg*, Brad Griffin Green Streak Programs - Lynne Green Hewlett-Packard - Rob Elliott Intel - Michael Mirmak* Mentor Graphics Corp. - John Angulo, Vladimir Dmitriev-Zdorov* Micron Technology - Randy Wolff Sigrity - Sam Chitwood, Raymond Y. Chen, Tao Su, Brad Brim SiSoft - Walter Katz* Teraspeed Consulting Group - Bob Ross* ======================================================================== No patents were announced. The team continued reviewing Rob Elliot's suggested changes. The proposed changes were incorporated in draft 14 as shown below, with comments from the team: Changes made: - ASCII properly defined - use "shall" instead of "may" as appropriate to context globally { Bob stated that "may" can be changed to "is permitted" Michael Will check on IEEE/ANSI/GEIA rules. Radek and Bob support changing "may" on a case-by-case basis } - rules missing for Version 2.0, Version 1.0 under Comments { Bob stated that perhaps we might add rule mentioning Version 2.0 follows Version 1.0 rules without change } - "option line" instead of "options line" globally { option line is used in original document } - use 0.5 instead of ½ in Appendix A { this is due to line spacing; Radek suggests using r0/2 } - log10 should have 10 as a subscript - "specification" instead of "document" globally - shall, will and must usage rules enforced globally - will be, will consist, is, contains, etc. made consistent - "bang" replaced as name for ANSI standard ! character - change last magnitude-angle example to MHz/dB - p. 1: 2009 instead of 2008 - p. 1: use (R) only on p. 1 and not thereafter - p. 3: incorrect placement of - and , characters - p. 4: 20h instead of hexadecimal 20; same with 07E - p. 4: is instead of IS - p. 5: The [Version] keyword instead of [Version], under Rules - p. 5: "keyword" instead of "line" on p. 5 for [Version] { Radek noted that we never actually define "line" } - p. 6: reorder MA|DB|RI to DB|MA|RI - p. 6: reorder units of frequency order - p. 6: change n-port to 3-port - p. 6: use GHz at end of frequency unit order - p. 7: remove note regarding Version 1.0 syntax - p. 7: remove "Note that" - p. 12: or instead of and/or - p. 12 lines instead of 'lines', remove "after the newline character" - p. 13: 1-port instead of ort - p. 15: extra period - p. 15: Re instead of Rel - p. 19: check spacing on n x n - p. 23: change Noise Parameters to Noise Parameter Data { Radek: either is fine; Noise Parameter Data is consistent with the rest of the document; Bob noted this was called Data Lines in original document} - p. 24: add !NETWORK DATA PARAMETERS comment to both examples { dates back from version 1.0; either take originals out or add in new ones; consensus is to add in new comments } - p. 27 passive voice instead of active voice (WE) on this page and globally - p. 28 consists instead of will consist - p. 33 capitalization of Theory Not Made Yet: - p. 5: no example for Comments - "network data" instead of "data" globally - Touchstone should also mandate that only one line termination sequence be used throughout the entire file { Radek suggested this was possible, but may not make any difference; Bob noted that it could be a rule, but we don't necessarily have to check it; if it's a current practice, we can follow it. It also represents a possible change for IBIS and ICM } - "data lines" should be "network parameter data pairs" globally { Bob suggested watching for frequency blocks; Radek added that "network data" is fine, as is "keyword data"; also, watch for "argument" vs. "entry" } - entry/entries vs. network data lines, data, etc. globally - use explicit multiplication symbol X { disagreement about use of x versus "splat"; rule to be discussed } - standardize use of ^ for power/exponential - add examples showing [Reference] across multiple lines { Michael suggested we could do this with existing examples with small change; Bob added that we put in comment line regarding ports, then put in references; modify one existing example as grouping will be often implemented } Will Not Make (Per Team Agreement): - remove unnecessary keywords from examples { team agreed that inconsistency is preferable to starting from scratch } - Need Version example { team disagrees; consensus is to maintain style of current examples; note presence of [End] without example } - Mixed-Mode data should go above Network Data, to keep keywords ordered consistently - [Mixed-Mode Order] should be above [Network Data] { Radek noted that the only way to address this consistent with Rob's suggestion is to move entire mixed-mode introduction to an appendix } - p. 4: Comment definitions present on p. 4 and duplicated on p. 5 { Bob observed that the text on p. 4 is an overall summary, not a detailed statement of rules; problem is use of different rules/wording; Radek added that a multi-line comments rule may need to be re-stated with ! symbol; problem is also that comments are permitted before [Version] definition; team consensus was to keep repeated rules } - p. 16: explain what's special about 5-port data vs. the others { rules for 5 ports are the same as 4 with one additional wrapping pair; 1 and 2 are special; 3 is special and so is 4; Radek added that it's not that 5 is special; only that Version 1 makes 1-4 special; not critical to address, but it isn't explained directly in the doc; thought it was clear from existing text } - p. 25: move Information block wrapper higher in the spec, with the other keywords as properly ordered { this was a conscious choice when adding the Information block } - p. 15: line up etc. - p. 15: make 7 GHz data consistent in terms of alignment, but add example of unaligned data - p. 23: use network parameter data instead of individual call-out of G, H, Z, etc. ARs - Team: Vladimir stated that the document should include two complete separate examples, with as many functions as possible. - Michael to obtain GEIA and/or IEEE rules for may, shall, will, must in standards - DONE - Michael to list out Rob's suggested changes as a bullet list, so that global principles can be settled before deciding. - DONE