Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Double-Ended Atom Token Translation

IP.com Disclosure Number: IPCOM000116787D
Original Publication Date: 1995-Nov-01
Included in the Prior Art Database: 2005-Mar-31
Document File: 4 page(s) / 116K

Publishing Venue

IBM

Related People

Johnson, VM: AUTHOR

Abstract

A method is described for allowing message passing using self-defining data structures that does not require the same data names to be used by the sending and receiving programs.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 49% of the total text.

Double-Ended Atom Token Translation

      A method is described for allowing message passing using
self-defining data structures that does not require the same data
names to be used by the sending and receiving programs.

      Some distributed object messaging software uses self-defining
data structures for passing messages between programs.  This has the
benefit of freeing the sending and receiving programs from a static
parameter list structure but has the disadvantage of requiring more
data to be transmitted.  The messaging software reduces this cost by
using translation tables to convert text strings to/from numeric
values to reduce the volume of data transmitted in the distributed
environment.  The tables are built at execution time and may be
partitioned to speed searching for table entries.  The entries in the
table and the partitioning are based on text strings hard-coded in
the application programs which use the communication software.  The
text strings represent items such as message/method names or data
structure names and are designed to include both a table partition
designation and the method/message/data structure name (identifier
for the data).  The first transmission of a string includes both the
string and the associated numeric token.  The remaining transmissions
include only the token, which is converted to the string at the
receiving node.  However, since the communication software can not
control the text values used in the underlying programs their current
technique has limitations:
  1.  Translation tables have the same partitioning across nodes.
  2.  Sending and receiving programs must know and use the same
method
       and data structure names.

      The first limitation is important because it will not always be
possible to predict and control which partitions each application
uses for its strings.  Thus, a partition at a node may become
overloaded if several programs run at that node happen to choose to
use the same partition.  The second limitation means that the
application programs are not as independent from one another as is
desirable in a 'plug and use' software environment.  They are in
effect connected at program development time (i.e., when they are
coded).  Publishing each program's Application Program Interface's
(API) /data structures is one possible solution, but still results in
a code time linking of two programs which wish to interact.  An
illustration of the current technique follows:
        Sender                             Receiver
        Table                              Table
  String    Token    Initial send      Token    String
  A | SYS    11       A|SYS->11          11     A | SYS
  B | SYS    12       B|SYS->12          12     B | SYS
  C | MKT    21       C|MKT->21          21     C | MKT
  D...