Cliff's Config Proposal #4 (simplification)

From: Clifford E. Cummings (cliffc@sunburst-design.com)
Date: Sun May 08 2005 - 11:08:48 PDT

  • Next message: Mark Hartoog: "RE: Last 2 email votes for 1364"

    PROPOSAL - Mantis #680 - 5/8/2005 by Cliff Cummings

    Observations:
    - We do not have complete agreement on the discussed configurations
    proposal
    - There are already at least two implementations that recognize the
    keywords: config, endconfig, cell
    - The two implementations have minor use-model differences that will
    have no consequence to 99%+ of all Verilog designs
    - Most engineers will separate library files from config files from
    design files, and the divergent use-models are actually very similar when
    engineers make the separations
    - config, endconfig and cell are already passed keywords in the
    Verilog-2001 Standard
    - We do not want to set a precedent of reviewing keywords after they
    have become part of an IEEE Standard and after they have been implemented
    by one or more vendors
    - The current proposed changes are too significant to be added to a
    re-circulation ballot of the standard and do not promote consensus.
    - The current proposed changes will change at least one yes-vote to a
    no-vote on both the 1364 and 1800 re-circulation ballots. The no-votes will
    be related to changes in the standard and will have to be addressed in a
    third re-circulation ballot.
    - There is already at least one book and multiple copies of training
    materials that show the use of the config-endconfig implementation of
    configurations. All of these published materials become obsolete if the
    keywords change.
    - There are other config-issues that should be addressed in a
    more-encompassing proposal, including the issues of wildcard directories
    and file-naming to wildcard matching restrictions (both hinted at by Steve
    Sharp).

    Proposal options (I vote yes to either proposal under the conditions noted
    below).

    (1) Do nothing (not even BNF or typo fixes) and respond to the negative
    ballot with the template form: "Although the proposed changes have merit,
    they are beyond the scope (etc., etc.) ..."

    This option is the safest because no new negative votes can be registered
    against the 1364 and 1800 standards based on any configuration issues.

    (2) Make the minor BNF changes and typo fixes shown (in the attached
    file) and still respond with the above template form.

    I am willing to support this option only if a straw-poll show that there
    will be no new configuration section objections raised by the changes. If
    Cadence or any other entity indicates that they would register a new
    negative vote due to the BNF and typo changes, then I will vote no to even
    these minor changes.

    See attached proposal file for details.

    Regards - Cliff

    ----------------------------------------------------
    Cliff Cummings - Sunburst Design, Inc.
    14314 SW Allen Blvd., PMB 501, Beaverton, OR 97005
    Phone: 503-641-8446 / FAX: 503-641-8486
    cliffc@sunburst-design.com / www.sunburst-design.com
    Expert Verilog, SystemVerilog, Synthesis and Verification Training


    This archive was generated by hypermail 2.1.4 : Sun May 08 2005 - 10:48:57 PDT and
    sponsored by Boyd Technology, Inc.