====================================================================== IBIS EDITORIAL TASK GROUP http://www.ibis.org/Editorial_wip/ Mailing list: ibis-editorial@freelists.org Archives at http://www.freelists.org/archive/ibis-editorial/ ====================================================================== Attendees from May 19 Meeting (* means attended at least using audio) ANSYS Curtis Clark Intel Corp. Michael Mirmak* Keysight Technologies Radek Biernacki*, Ming Yan Mentor, a Siemens Company Arpad Muranyi* Micron Technology Justin Butterfield, Randy Wolff* Signal Integrity Software Walter Katz*, Mike LaBonte* Teraspeed Labs Bob Ross* Michael Mirmak convened the meeting. No patents were declared. Mike LaBonte moved to approve the May 5 minutes; Bob Ross seconded. The minutes were approved with no objections. During the review of ARs, Mike noted that he had completed his AR to update Figure 1 to indicate "directory". Bob stated that he had completed his AR to fold three new definitions into the most recent draft. Bob added that work is needed on Sections 3.3 and 3.4, and not all cumuluative changes have been folded into the new draft. A lot of changes are directives, but the final BIRD might show the final full language. Mike reviewed Draft 7. The "Summary of Proposed Changes" section identifies the parts that are changing. Mike added that he is specifically looking for clashes, overlaps, etc. Radek Biernacki noted that the phrase "parts of file" may be confusing. "Parts of file name" is less so. Mike suggested that new filename definitions can be placed in a new section 3.1; an entry for "directory" has been added. Michael objected to the description of "directory" as a file; Mike suggested "container", and Radek agreed. Michael noted that "directory" is a location to almost every user, but the technical definition is correct for the system. Bob noted that, in the boost.org approach, "filename" is one word, but IBIS uses two words. The two word format is being maintained in this draft. "File" was added to the list in the draft BIRD, which was not present in boost.org. Mike added that a reader has to visit several locations on boost.org in order to obtain the full list in the BIRD. Radek observed that the definition of file uses "expressed" but is not actually identified. Arpad Muranyi asked whether the definition lists what the file *is* or how it's named. Mike noted that there is a similar issue in Java. The reason here is because we are using the word "file" when we are talking about the *name*. The New Figure 1 shows the dividers in red boxes and shows the directory names. However, the Section numbering needs correction in the final BIRD. Arpad asked whether absolute paths do not need to include root names. Mike noted that this complicates things, but is true. Randy Wolff added that it's illegal, so we could omit it. Bob stated that, in BIRD186, the item 3 definition is very long. Mike replied that this has been reorganized in the new BIRD. Bob answered that would like to remove the stem requirement at some point, in favor of having the complete, absolute path limited to 256 characters. Radek observed that we have to have room for porting data to a physical system, which will consume some part of the total path. Mike answered that we could simply call for common sense. Bob added that the document had that in an earlier version and should state that; we don't want to check path names along with file names. VHDL has some limitations related to extensions, with more than three characters. Offline work is needed to work the extension issue out. The treatment of slash and period characters needs working out as well. Mike noted that several sections change 120 to 1024, which needs to be done globally throughout the document. Walter Katz asked whether this needs to be done as an Editorial pass or through a BIRD. Mike replied that this is not that big a deal, as it could take only a few minutes to cover everything in the BIRD. Bob shared his changes: [File Name] now uses adjusted language featuring "shall", and file_name language for [External Circuit] has been changed to "file" where appropriate. We may not check stem length for Touchstone, for example. This may mean restricting character lengths to accommodate different operating systems. This occurs in three locations, including ami files and multi-lingual twice. Arpad suggested the change from file_name to file; object and object name. Are we enforcing something on the object file, or something on the name. We are usually talking about its name, not the file. Walter noted that some of the confusion is between "file name" and "file_name". "File name" is the more logical version. Radek suggested that "file" should be the object, as we are really talking about a "file reference", which can include the name and the path. We are free to introduce such a term. Bob agreed, suggesting it can be added to the list. Bob added that one can reference SPICE directories and information separately from the actual SPICE subcircuit. Michael asked whether file references affect ISS or other external files. Walter suggested including a statement in IBIS that ISS may be affected. For AMI, parameter file and file name were being misused in the keyword description. This will be corrected. Michael asked whether these changes had been folded into the draft sent out. Mike confirmed that they had. Bob moved to adjourn. Mike seconded the motion. The meeting adjourned without objection.