DC-WG: 2/15 meeting minutes


Subject: DC-WG: 2/15 meeting minutes
From: Mark Hahn (mhahn@cadence.com)
Date: Mon Feb 21 2000 - 16:11:28 PST


Meeting minutes from the 2/15/00 DC-WG teleconference
-----------------------------------------------------

Attendees:
  Mark Hahn, Cadence
  Tom Dewey, Mentor
  Bob Dilly, IBM
  Vassilios Gerousis, Infineon
  John Paul, IBM

New action items:

      Who When What
      ---------- ------ --------
 1. Tom 2/22 Update the spec with changes
 2. Bob 2/22 Talk with IBM legal about the existing
                           Jikes compiler open source license

Next Meeting:

  The next meeting will be a teleconference on 2/22/00
  from 9-11 am (PDT)

Details:
  1. Review status of golden parser

     John Paul is currently revising the parser and expects to
     finish within about a week.

     We talked about www.sourceforge.com, a web site that offers
     free hosting of a source code development environment and tools
     for open source software. This looks like an ideal way to
     support the parser.

     From this web site, Mark discovered that there is an existing
     open source license for the Jikes compiler from IBM, which could
     provide a useful precedent and speed up the legal process.

  2. Review status of press release

     There was no new information from Kent on this. Vassilios
     noted that we need a catalyst on which to base the press release;
     the availability of the golden parser seems to be the best
     candidate.

  3. Discussion on section 4

     We reviewed the constant, extend_dcdl, history, and
     persistent_comment commands.

     There was some discussion about the relationship between
     constants and mode-dependent timing analysis. We agreed
     that the domain-independent constant command should specify
     constant values that are not mode-dependent, and that there
     may be a separate command in the timing domain. [Ed. note:
     the BuildGates command language has two commands for setting
     constants]

     We talked a lot about extend_dcdl and the persistence of
     the data in the extensions. As a general rule, we agreed
     that it's unreasonable to expect tools to maintain the same
     order between constraints and other persistent data when
     writing out a DCDL file, because the design may be modified
     in ways that require corresponding modifications to constraints.

     However, order within the history commands and extensions is
     important. We agreed that tools must preserve this order within
     a scope. History commands should appear for a scope before
     constraint commands, and extensions should appear after the
     constraint commands.

     We thought that the extension mechanism should be used for
     any persistent comments, rather than introducing a new command,
     because a new command would incorrectly imply that persistent
     comments would show up in the same place in output DCDL files.

Thanks,
Mark

-- 
Mark Hahn                                          phone: (408) 428-5399
Senior Architect                                   fax:   (408) 428-5959
Cadence Design Systems                             email: mhahn@cadence.com



This archive was generated by hypermail 2b28 : Mon Feb 21 2000 - 16:14:13 PST