(attaching a text version of the minutes for ease of archiving) ====================================================================== IBIS INTERCONNECT TASK GROUP http://www.ibis.org/interconnect_wip/ Mailing list: ibis-interconnect@freelists.org Archives at http://www.freelists.org/archive/ibis-interconn/ ====================================================================== Attendees from April 6 Meeting (* means attended at least using audio) ANSYS Curtis Clark* Cadence Design Systems Bradley Brim Cisco David Siadat Intel Corp. Michael Mirmak* Keysight Technologies Radek Biernacki*, Ming Yan Mentor Graphics Arpad Muranyi Micron Technology Justin Butterfield*, Randy Wolff* SAE ITC Maureen Lemankiewicz, Logen Johnson Signal Integrity Software Walter Katz, Mike LaBonte* Teraspeed Labs Bob Ross* University of Aveiro in Portugal Wael Dghais Michael Mirmak convened the meeting. No patents were declared. Michael called for comments on the minutes of the March 30 meeting. Bob Ross moved to approve the minutes. Radek Biernacki seconded the motion. No objections were raised and the motion carried. No opens were raised. Michael noted that Brad Brim was not in attendance, so his presentation could not be reviewed. The team reviewed Draft 31. Bob commented on table regarding “Aggressor”, noting there is still an open question about capitalization. Should “rail” be lower-case or not? “I/O” refers to “Input/Output”. Do we keep the “R” upper-case? Bob expressed a mild preference for lower-case. Bob notes that “pin_name” should be lower-case. Mike LaBonte noted that “PdRef” or “PDRef” would be more consistent. He otherwise accepts the usage for qualifiers. He noted that Bob addressed this in an earlier document. Radek suggested another vote could be held to resolve this. Michael noted an off-line discussion regarding “sub-params” vs. “subparameters”, and that the text of several pin-related keywords call the headers of columns “subparameters”. He observed that this could be easily corrected in the document text, but that the parser impact was not completely known. Mike stated that most treat column headers as sub-parameters even in the specification. Bob added that headings for the pin list were originally treated as subparameters. Mike formally moved to use an initial upper-case letter and an underscore for “Gc_Ref”, and similar Interconnect terminal names (e.g., “R” for reference with a following underscore, _); he amended this motion to use capitalization for both initial letters, as in GC_Ref, etc. Randy Wolff noted that the document uses labels that refer to three different locations (pin, pad, buffer), but now we have another set assumed to be at the buffer but aren’t labeled that way. He suggested using the full word “buffer”, with capitalization, as in “Buffer_PD_Ref”, etc. Radek added that it’s the type of the terminal used as the initial set of letters, which doesn’t change between PU_Ref, PD_Ref, etc. Is it needed? Bob replied yes, that it is; we make reference to a pin name and then a terminal type, and then we connect to PU_Ref, but we might connect in a different way to PC_Ref. Radek replied that he would like to come back to this issue after consideration of where connections actually attach. Mike expressed concern about line length with the full word “buffer”. Bob noted that 120 characters per line are allowed, so this extension of the name should not be an issue. Mike accepted Randy’s proposed alteration. Bob noted that this is totally inconsistent with is elsewhere in the document and specification. A more consistent usage would be “Buf”. He suggested against using “PD” by itself, as it’s a new convention and an abbreviation for a single word. Mike replied that making the format inconsistent makes the terms more recognizable as terminal types. Bob replied this was a good argument. Curtis Clark suggested “Buf” instead of “Buffer”. Radek added that even “Buf_PU” would be acceptable; the use of “Ref” is unnecessary. Randy replied that “Buf_Ext” would be a bit odd. Mike suggested he would almost be willing to head in direction of keeping Ref for all terminal names *except* “Ext”. He could accept having all use “Ref” or none use “Ref”. Mike withdrew his original motion, noting that the “Allowed Terminal_Type_Associations” table needs scrubbing. Mike made a new motion to change the table to use a format where “Buf” leads all terminal_types, and is followed by _ and where double-capitalized letters are used (as in “Buf_PU_Ref”). Bob seconded the motion. The attending organizations voted as follows: ANSYS - yes Intel Corp. - yes Keysight Technologies - abstains Micron Technology - yes Signal Integrity Software - yes Teraspeed Labs - yes The motion carried. Mike accepted the action item to post a revised document, draft 32, with this change. Next time, the team will take up Brad’s proposal and, time permitting, review Draft 32. Curtis moved to adjourn. Randy seconded the motion. The meeting adjourned.