Network Working Group                                        S. Nelson
Request for Comments: 2077                                        LLNL
Category: Standards Track                                     C. Parks
                                                                 NIST
                                                                Mitra
                                                           WorldMaker
                                                         January 1997


                  The Model Primary Content Type for
                Multipurpose Internet Mail Extensions

Status of this Memo

  This document specifies an Internet standards track protocol for the
  Internet community, and requests discussion and suggestions for
  improvements.  Please refer to the current edition of the "Internet
  Official Protocol Standards" (STD 1) for the standardization state
  and status of this protocol.  Distribution of this memo is unlimited.

Introduction

  The purpose of this memo is to propose an update to Internet RFC 2045
  to include a new primary content-type to be known as "model". RFC
  2045 [1] describes mechanisms for specifying and describing the
  format of Internet Message Bodies via content-type/subtype pairs. We
  believe that "model" defines a fundamental type of content with
  unique presentational, hardware, and processing aspects.  Various
  subtypes of this primary type are immediately anticipated but will be
  covered under separate documents.

Table of Contents

     1. Overview.............................................  2
     2. Definition...........................................  2
     3. Consultation Mechanisms..............................  4
     4. Encoding and Transport...............................  5
     5. Security Considerations Section......................  6
     6. Authors' Addresses...................................  7
     7. Expected subtypes....................................  7
     8. Appendix.............................................  9
     9. Acknowledgements..................................... 13









Nelson, et. al.             Standards Track                     [Page 1]

RFC 2077                Model Primary MIME Types            January 1997


1. Overview

  This document will outline what a model is, show examples of models,
  and discuss the benefits of grouping models together.  This document
  will not directly deal with the intended subtypes since those will be
  covered by their separate registrations.  Some immediately expected
  subtypes are listed in section 7.

  This document is a discussion document for an agreed definition,
  intended eventually to form a standard accepted extension to RFC
  2045.  We are also targeting developers of input/output filters,
  viewer software and hardware, those involved in MIME transport, and
  decoders.

2. Definition of a model

  A model primary MIME type is an electronically exchangeable
  behavioral or physical representation within a given domain.  Each
  subtype in the model structure has unique features, just as does each
  subtype in the other primary types.  The important fact is that these
  various subtypes can be converted between each other with less loss
  of information then to that of other primary types.  This fact groups
  these subtypes together into the model primary type.  All of the
  expected subtypes have several features in common and that are unique
  to this primary type.

  To loosely summarize: models are multidimensional structures composed
  of one or more objects.  If there are multiple objects then one
  object defines the arrangement/setting/relationship of the others.
  These objects all have calibrated coordinate systems but these
  systems need not be in the same units nor need they have the same
  dimensionality.  In detail:

  1. have 3 or more dimensions which are bases of the system and
     form an orthogonal system (any orthogonal system is sufficient).

     This system is specifically defined in terms of an orthogonal
     set of basis functions [for a subspace of the L^2 function space]
     over a coordinate system of dimension 3 or more. Note that this
     does not preclude regular skewed systems, elliptical coordinates,
     different vector spaces, etc.

  2. contain a structural relationship between model elements.

  3. have scaling or calibration factors which are related to physical
     units (force, momentum, time, velocity, acceleration, size, etc.).
     Thus, an IGES file will specify a building of non-arbitrary size,
     computational meshes and VRML models will have real spatial/



Nelson, et. al.             Standards Track                     [Page 2]

RFC 2077                Model Primary MIME Types            January 1997


     temporal units. This allows for differing elements to be combined
     non-arbitrarily.

  4. Models can be single objects or composed of a collection of
     objects.  These normally independent objects are arranged
     in a master/slave scenario so that one object acts as the
     reference, or primary object, which defines how the other
     objects interrelate and behave.  This allows for the creation
     of mathematical, physical, economic, behavioral, etc. models
     which typically are composed of different elements.  The key is
     in the description: these types describe how something
     "behaves"; contrasted to typical data types which describe
     how something "is".

     The inclusion of this "collective" system works similar to the
     Email system's multipart/related type which defines the actions
     of the individual parts.  Further specification of the model/*
     subtypes utilizing these properties is left to the subtype
     authors.

  With these assumptions:

  a. the default dimensionality will be spatial and temporal (but
     any are allowed).

  b. it is presumed that models will contain underlying structure
     which may or may not be immediately available to the
     user. (fluid dynamics vector fields, electromagnetic
     propagation, interrelated IGES dimensional specifiers, VRML
     materials and operators, etc.)

  c. it is assumed that basis set conversion between model domains
     is lossless.  The interpretation of the data may change but
     the specification will not.  i.e. convert the model of the
     U.S.A.  Gross Domestic Product into a VRML model and navigate
     it to explore the variances and interrelationships.  The model
     has many dimensions but also "passages" and "corridors"
     linking different parts of it.  A similar situation is true
     for meshes and CAD files. The key is identifying the basis set
     conversion which makes sense.

  d. models are grouped to assure LESS loss of information between
     the model subtypes than to subtypes of other primary
     types. (i.e.  converting a chemical model into an image is
     more lossy than concerting it into a VRML model).






Nelson, et. al.             Standards Track                     [Page 3]

RFC 2077                Model Primary MIME Types            January 1997


  Items c and d above define the grouping for model similar to the way
  that "images" and "videos" are grouped together; to assure less loss
  of information.  Obviously converting from a GIF image to a JPEG
  image looses less information than converting from a GIF image to an
  AU audio file.

3.  Consultation Mechanisms

  Before proposing a subtype for the model/* primary type, it is
  suggested that the subtype author examine the definition (above) of
  what a model/* is and the listing (below) of what a model/* is not.
  Additional consultations with the authors of the existing model/*
  subtypes is also suggested.

  Copies of RFCs are available on:

                       ftp://ftp.isi.edu/in-notes/

  Copies of Internet-Drafts are available on:

                   ftp://ftp.ietf.org/internet-drafts/

  Similarly, the VRML discussion list has been archived as:

                       http://vrml.wired.com/arch/

  and discussions on the comp.mail.mime group may be of interest.
  Discussion digests for the existing model/* subtypes may be
  referenced in the respective documents.

  The mesh community presently has numerous different mesh geometries
  as part of different packages.  Freely available libraries need to be
  advertised more than they have been in the past to spur the
  development of interoperable packages.  It is hoped that by following
  the example of the VRML community and creating a freely available
  comprehensive library of input/output functions for meshes [11] that
  this problem will be alleviated for the mesh community.  A freely
  available mesh viewer conforming to these standards is available now
  for various platforms.  Consulations with the authors of the mesh
  system,

           http://www-dsed.llnl.gov/documents/tests/mesh.html

  will be beneficial.

  The IGES community has a suite of tests and conformance utilities to
  gauge the conformance to specifications and software authors are
  encouraged to seek those out from NIST [14].



Nelson, et. al.             Standards Track                     [Page 4]

RFC 2077                Model Primary MIME Types            January 1997


4. Encoding and Transport

  a. Unrecognized subtypes of model should at a minimum be treated
     as "application/octet-stream".  Implementations may optionally
     elect to pass subtypes of model that they do not specifically
     recognize to a robust general-purpose model viewing
     application, if such an application is available.

  b. Different subtypes of model may be encoded as textual
     representations or as binary data.  Unless noted in the
     subtype registration, subtypes of model should be assumed to
     contain binary data, implying a content encoding of base64 for
     email and binary transfer for ftp and http.

  c. The formal syntax for the subtypes of the model primary type
     should look like this:

     Media type name:          model
     Media subtype name:       xxxxxxxx
     Required parameters:      none
     Optional parameters:      dimensionality, state
                               (see below)
     Encoding considerations:  base64 encoding is recommended when
                               transmitting model/* documents through
                               MIME electronic mail.
     Security considerations:  see section 5 below
     Published specification:  This document.
                               See Appendix B for references to some of
                               the expected subtypes.
     Person and email address to contact for further information:
                               Scott D. Nelson <[email protected]>
                               7000 East Ave.
                               Lawrence Livermore National Laboratory
                               Livermore, CA  94550

  The optional parameters consist of starting conditions and variable
  values used as part of the subtypes.  A base set is listed here for
  illustration purposes only and will be covered in detail as part of
  the respective subtypes:

 dimension := string ; a number indicating the number of dimensions.
                       This is used as a "hint" in selecting
                       applicable viewer programs.








Nelson, et. al.             Standards Track                     [Page 5]

RFC 2077                Model Primary MIME Types            January 1997


 state     := string ; "static" or "dynamic".  In "static", the
                       observer may move about, thus effecting
                       translations, rotations, pans, zooms, etc.
                       but the data does not change.  In "dynamic",
                       the data itself is manipulated via
                       skews, elongations, scales, etc.  Note that
                       time evolution is still a static operation
                       since it is just a translation along one of
                       the principal dimensions while the elongation
                       of a cube or object deformation are dynamic
                       operations.

     Note that this optional parameter list does not limit those
     specified by the various subtypes.

  d. The specific issues relating to the various subtypes are covered
     as part of the description of those specific subtypes.  The
     following is an example of a typical MIME header used for mail
     transport purposes:

        To:   [email protected]
        From: [email protected]
        Date: Fri, 30 Aug 96 13:33:19 -0700
        Content-Type: model/mesh; dimension="4"; state="static"
        Content-Transfer-Encoding: base64
        MIME-Version: 1.0
        Subject: model data file

        I1ZSTUwgVjEuMCBhc2NpaQojIFRoaXMgZmlsZSB3YXMgIGdlbmVyY...
        byBDb21tdW5pY2F0aW9ucwojIGh0dHA6Ly93d3cuY2hhY28uY29tC...
        IyB1c2VkIGluIHJvb20gMTkyICh0ZXN0IHJvb20pCiAgIAojIFRvc...
        .
        .
        .

5.  Security Considerations Section

  Note that the data files are "read-only" and do not contain file
  system modifiers or batch/macro commands.  The transported data is
  not self-modifying but may contain interrelationships.  The data
  files may however contain a "default view" which is added by the
  author at file creation time.  This "default view" may manipulate
  viewer variables, default look angle, lighting, visualization
  options, etc.  This visualization may also involve the computation of
  variables or values for display based on the given raw data.  For
  motorized equipment, this may change the position from the hardware's
  rest state to the object's starting orientation.




Nelson, et. al.             Standards Track                     [Page 6]

RFC 2077                Model Primary MIME Types            January 1997


  The internal structure of the data files may direct agents to access
  additional data from the network (i.e. inclusions); the security
  limits of whom are not pre-supposed.  Actions based on these
  inclusions are left to the security definitions of the inclusions.
  Further comments about the security considerations for the subtypes
  will be contained in each subtype's registration.

6. Authors' Addresses

     S. D. Nelson
     Lawrence Livermore National Laboratory,
     7000 East Ave., L-153,
     Livermore CA 94550, USA.
     E-Mail: [email protected]

     C. Parks
     National Institute of Standards & Technology
     Bldg 220, Room B-344
     Gaithersburg, MD 20899, USA.
     E-Mail: [email protected]

     Mitra
     WorldMaker
     1056 Noe
     San Francisco, CA 94114
     E-Mail: [email protected]

7. Expected subtypes

  Table 1 lists some of the expected model sub-type names.  Suggested 3
  letter extensions are also provided for DOS compatibility but their
  need is hopefully diminished by the use of more robust operating
  systems on PC platforms.  The "silo" extension is provided for
  backwards compatibility.  Mesh has an extensive list of hints since
  the present variability is so great.  In the future, the need for
  these hints will diminish since the files are self describing.  This
  document is not registering these subtypes.  They will be handled
  under separate documents.













Nelson, et. al.             Standards Track                     [Page 7]

RFC 2077                Model Primary MIME Types            January 1997


Table 1.

  Primary/sub-type           Suggested extension(s)    Reference

  model/iges                         igs,iges              [8]
  model/vrml                         wrl                   [9]
  model/mesh                         msh, mesh, silo       [10]

  It is expected that model/mesh will also make use of a number of
  parameters which will help the end user determine the data type
  without examine the data.  However, note that mesh files are self-
  describing.

     regular+static, unstructed+static, unstructured+dynamic,
     conformal+static, conformal+dynamic, isoparametric+static,
     isoparametric+dynamic

  The sub-types listed above are some of the anticipated types that are
  already in use.  Notice that the IGES type is already registered as
  "application/iges" and that RFC states that a more appropriate type
  is desired.  Note that the author of "application/iges" is one of the
  authors of this "model" submission and application/iges will be re-
  registered as model/iges at the appropriate time.

  The VRML type is gaining wide acceptance and has numerous parallel
  development efforts for different platforms.  These efforts are
  fueled by the release of the QvLib library for reading VRML files;
  without which the VRML effort would be less further along.  This has
  allowed for a consistent data type and has by defacto established a
  set of standards. Further VRML efforts include interfaces to other
  kinds of hardware (beyond just visual displays) and it is proposed by
  those involved in the VRML effort to encompass more of the five
  senses.  Unlike other kinds of "reality modeling" schemes, VRML is
  not proprietary to any one vendor and should experience similar
  growth as do other open standards.

  The mesh type is an offshoot of existing computational meshing
  efforts and, like VRML, builds on a freely available library set.
  Also like VRML, there are other proprietary meshing systems but there
  are converters which will convert from those closed systems to the
  mesh type.  Meshes in general have an association feature so that the
  connectivity between nodes is maintained.  It should be noted that
  most modern meshes are derived from CAD solids files.








Nelson, et. al.             Standards Track                     [Page 8]

RFC 2077                Model Primary MIME Types            January 1997


8. Appendices

8.1 Appendix A -- extraneous details about expected subtypes

VRML Data Types

  The 3D modeling and CAD communities use a number of file formats to
  represent 3D models, these formats are widely used to exchange
  information, and full, or lossy, converters between the formats exist
  both independently and integrated into widely used applications. The
  VRML format is rapidly becoming a standard for the display of 3D
  information on the WWW.

Mesh Data Types

  For many decades, finite element and finite difference time domain
  codes have generated mesh structures which attempt to use the
  physical geometry of the structures in connection with various
  physics packages to generate real world simulations of events
  including electromagnetic wave propagation, fluid dynamics, motor
  design, etc.  The resulting output data is then post processed to
  examine the results in a variety of forms.  This proposed mesh
  subtype will include both geometry and scalar/vector/tensor results
  data.  An important point to note is that many modern meshes are
  generated from solids constructed using CAD packages.

  Motivation for mesh grew out of discussions with other communities
  about their design requirements.  Many CAD or scene descriptions are
  composed of a small number of complex objects while computational
  meshes are composed of large numbers of simple objects.  A 1,000,000
  element 3D mesh is small.  A 100,000,000 element 3D structured mesh
  is large.  Each object can also have an arbitrary amount of
  associated data and the mesh connectivity information is important in
  optimizing usage of the mesh.  Also, the mesh itself is usually
  uninteresting but postprocessing packages may act on the underlying
  data or a computational engine may process the data as input.

  Meshes differ principally from other kinds of scenes in that meshes
  are composed of a large number of simple objects which may contain
  arbitrary non-spatial parameters, not all of whom need be visible,
  and who have an implicit connectivity and neighbor list.  This latter
  point is the key feature of a mesh. It should be noted that most
  meshes are generated from CAD files however.  The mesh type has
  association functions because the underlying physics was used to
  calculate the interaction (if you crash a car into a telephone pole,
  you get a crumpled car and a bent pole).  Most interesting
  computational meshes are 4D with additional multidimensional results
  components.



Nelson, et. al.             Standards Track                     [Page 9]

RFC 2077                Model Primary MIME Types            January 1997


IGES CAD Data Types

  (The following text, reproduced for reference purposes only, is from
  "U.S. Product Data Association and IGES/PDES Organization Reference
  Manual," June 1995; by permission.)

  IGES, the Initial Graphics Exchange Specification, defines a neutral
  data format that allows for the digital exchange of information among
  computer-aided design (CAD) systems.

  CAD systems are in use today in increasing numbers for applications
  in all phases of the design, analysis, and manufacture and testing of
  products. Since the designer may use one supplier's system while the
  contractor and subcontractor may use other systems, there is a need
  to be able to exchange data digitally among all CAD systems.

  The databases of CAD systems from different vendors often represent
  the same CAD constructs differently. A circular arc on one system may
  be defined by a center point, its starting point and end point, while
  on another it is defined by its center, its diameter starting and
  ending angle. IGES enables the exchange of such data by providing, in
  the public domain, a neutral definition and format for the exchange
  of such data.

  Using IGES, the user can exchange product data models in the form of
  wireframe, surface, or solid representations as well as surface
  representations. Translators convert a vendor's proprietary internal
  database format into the neutral IGES format and from the IGES format
  into another vendor's internal database. The translators, called pre-
  and post-processors, are usually available from vendors as part of
  their product lines.

  Applications supported by IGES include traditional engineering
  drawings as well as models for analysis and/or various manufacturing
  functions. In addition to the general specification, IGES also
  includes application protocols in which the standard is interpreted
  to meet discipline specific requirements.

  IGES technology assumes that a person is available on the receiving
  end to interpret the meaning of the product model data. For instance,
  a person is needed to determine how many holes are in the part
  because the hole itself is not defined. It is represented in IGES by
  its component geometry and therefore, is indistinguishable from the
  circular edges of a rod.

  The IGES format has been registered with the Internet Assigned
  Numbers Authority (IANA) as a Multipurpose Internet Mail Extension
  (MIME) type "application/iges". The use of the message type/subtype



Nelson, et. al.             Standards Track                    [Page 10]

RFC 2077                Model Primary MIME Types            January 1997


  in Internet messages facilitates the uniform recognition of an IGES
  file for routing to a viewer or translator.

  Version 1.0 of the specification was adopted as an American National
  Standards (ANS Y14.26M-1981) in November of 1981. Versions 3.0 and
  4.0 of the specification have subsequently been approved by ANSI. The
  current version of IGES 5.2 was approved by ANSI under the new
  guidelines of the U.S. Product Data Association. Under these
  guidelines, the IGES/PDES Organization (IPO) became the accredited
  standards body for product data exchange standards. This latest
  standard is USPRO/IPO-100-1993.

8.2 Appendix B -- References and Citations

  [1] Freed, N., and N. Borenstein, "Multipurpose Internet Mail
  Extensions (MIME) Part One: Format of Internet Message Bodies", RFC
  2045, Innosoft, First Virtual, November 1996.

  [2] Fitzgerald P., "Molecules-R-Us Interface to the Brookhaven Data
  Base", Computational Molecular Biology Section, National Institutes
  of Health, USA; see http://www.nih.gov/htbin/pdb for further details;
  Peitsch M.C, Wells T.N.C., Stampf D.R., Sussman S. J., "The Swiss-3D
  Image Collection And PDP-Browser On The Worldwide Web", Trends In
  Biochemical Sciences, 1995, 20, 82.

  [3] "Proceedings of the First Electronic Computational Chemistry
  Conference", Eds. Bachrach, S. M., Boyd D. B., Gray S. K, Hase W.,
  Rzepa H.S, ARInternet: Landover, Nov. 7- Dec. 2, 1994, in press;
  Bachrach S. M, J. Chem. Inf. Comp. Sci., 1995, in press.

  [4] Richardson D.C., and Richardson J.S., Protein Science, 1992, 1,
  3; D. C. Richardson D. C., and Richardson J.S., Trends in Biochem.
  Sci.,1994, 19, 135.

  [5] Rzepa H. S., Whitaker B. J., and Winter M. J., "Chemical
  Applications of the World-Wide-Web", J. Chem. Soc., Chem. Commun.,
  1994, 1907; Casher O., Chandramohan G., Hargreaves M., Murray-Rust
  P., Sayle R., Rzepa H.S., and Whitaker B. J., "Hyperactive Molecules
  and the World-Wide-Web Information System", J. Chem. Soc., Perkin
  Trans 2, 1995, 7; Baggott J., "Biochemistry On The Web", Chemical &
  Engineering News, 1995, 73, 36; Schwartz A.T, Bunce D.M, Silberman
  R.G, Stanitski C.L, Stratton W.J, Zipp A.P, "Chemistry In Context -
  Weaving The Web", Journal Of Chemical Education, 1994, 71, 1041.

  [6] Rzepa H.S., "WWW94 Chemistry Workshop", Computer Networks and
  ISDN Systems, 1994, 27, 317 and 328.





Nelson, et. al.             Standards Track                    [Page 11]

RFC 2077                Model Primary MIME Types            January 1997


  [7] S.D. Nelson, "Email MIME test page", Lawrence Livermore National
  Laboratory, 1994. See http://www-dsed.llnl.gov/documents/WWWtest.html
  and http://www-dsed.llnl.gov/documents/tests/email.html

  [8] C. Parks, "Registration of new Media Type application/iges",
  ftp://ftp.isi.edu/in-notes/iana/assignments/media-types/
  application/iges, 1995.

  [9] G. Bell, A. Parisi, M. Pesce, "The Virtual Reality Modeling
  Language",
  http://sdsc.edu/SDSC/Partners/vrml/Archives/vrml10-3.html, 1995.

  [10] S.D. Nelson, "Registration of new Media Type model/mesh",
  ftp://ftp.isi.edu/in-notes/iana/assignments/media-types/model/
  mesh, 1997.

  [11] "SILO User's Guide", Lawrence Livermore National Laboratory,
  University of California, UCRL-MA-118751, March 7, 1995,

  [12] E. Brugger, "Mesh-TV: a graphical analysis tool", Lawrence
  Livermore National Laboratory, University of California,
  UCRL-TB-115079-8, http://www.llnl.gov/liv_comp/meshtv/mesh.html

  [13] S. Brown, "Portable Application Code Toolkit (PACT)", the
  printed documentation is accessible from the PACT Home Page
  http://www.llnl.gov/def_sci/pact/pact_homepage.html

  [14] L. Rosenthal, "Initial Graphics Exchange Specification
  (IGES) Test Service",
  http://speckle.ncsl.nist.gov/~jacki/igests.htm

8.3 Appendix C -- hardware

  Numerous kinds of hardware already exist which can process some of
  the expected model data types and are listed here for illustration
  purposes only:

     stereo glasses, 3D lithography machines, automated manufacturing
     systems, data gloves (with feedback), milling machines,
     aromascopes, treadmills.











Nelson, et. al.             Standards Track                    [Page 12]

RFC 2077                Model Primary MIME Types            January 1997


8.4 Appendix D -- Examples

  This section contains a collection of various pointers to examples of
  what the model type encompasses:

  Example mesh model objects can be found on this mesh page:
     http://www-dsed.llnl.gov/documents/tests/mesh.html

  Various IGES compliant test objects:
     http://www.eeel.nist.gov/iges/specfigures/index.html

  VRML Test Suite:
     http://www.chaco.com/vrml/test/

  An image of a model of a shipping cage crashing into the ground:
     http://www.llnl.gov/liv_comp/meiko/apps/dyna3d/cagefig2.gif

  An image of a 100,000,000 zone mesh:
     http://www.llnl.gov/liv_comp/meiko/apps/hardin/PMESH.gif

  A video of a seismic wave propagation through a computational mesh:
     http://www.llnl.gov/liv_comp/meiko/apps/larsen/movie.mpg

9. Acknowledgements

  Thanks go to Henry Rzepa ([email protected]), Peter Murray-Rust
  ([email protected]), Benjamin Whitaker
  ([email protected]), Bill Ross ([email protected]),
  and others in the chemical community on which the initial draft of
  this document is based.  That document updated an IETF Internet Draft
  in which the initial chemical submission was made, incorporated
  suggestions received during the subsequent discussion period, and
  indicated scientific support for and uptake of a higher level
  document incorporating physical sciences[2-7].  This Model submission
  benefited greatly from the previous groundwork laid, and the
  continued interest by, those communities.

  The authors would additionally like to thank Keith Moore
  ([email protected]), lilley ([email protected]), Wilson Ross
  ([email protected]), hansen ([email protected]), Alfred Gilman
  ([email protected]), and Jan Hardenbergh ([email protected])
  without which this document would not have been possible.  Additional
  thanks go to Mark Crispin ([email protected]) for his comments
  on the previous version and Cynthia Clark ([email protected]) for
  editing the submitted versions.






Nelson, et. al.             Standards Track                    [Page 13]