1394 PWG - Issue No.970514-008

1394 PWG - Issue No.970514-008
Back to the Issues Page
ISSUE NO.:970514-008NAME:Need to use consistent terminology in our discussionsSTATUS:In Process
RESOLUTION:DATE:TBD
ORIGIN:05/14/97 Meeting in San Diego
DESCRIPTION:Difficulty in understanding various proposals and ideas came from people using different terminology. This issue attempts to create a single repository for the various terminology used in discussions.
DISCUSSIONS:
  • Device Discovery
    Items which pertain to the entire functional unit defined as the device connected to a unique 1394 node address.
      3)TERMINOLOGY....,my understanding of the definition of "device discovery"
      was more clearly defined as the following functions;
      1:device class identification(printer,scanner,....)
      2:manufacture, model name
      3:availability of lowest layer right above 1394 transport
      4:unique ID
      
    • Physical Device - Manufacturer, Model, Version
    • Unique ID
    • Class(es)
    • Printer
    • Scanner
    • Fax
    • Camera
    • etc.….
    • Mechanism
    • Broadcast
    • Poll
    • Location
    • ROM
    • RAM
    • Info
  • Datalink(s)
    • Should the datalink be paired with each individual Unit or refer to the group of Units that are supported or managed by that DL.
    • MFP devices identify individual Units that can be managed and accessed as a unique function. Units that cross class boundaries and must be managed as a composite whole would be identified as a Vendor specific Unit.
      4)TERMINOLOGY..........I thought the word "datalink" was defined for the
      layer right above the 1394 transaction layer.
      (No relationship with the same word used in OSI)
      
  • Service Discovery
    Services provided by the particular Unit for the printer class. A functional unit may consist of multiple Units, each providing some set of services.
    • Transport protocols
      5)TERMINOLOGY..........I thought the word "transport" was defined for the
      (whole) set of layer(s) above the 1394 transaction layer. Slim(Thin) transport and
      Thick transport were defined as well. (No relationship with the same word used in OSI)
      Under this definition, a monolithic "transport" will be the "datalink" itself.
      
    • Data transfer protocols
    • Push
    • Pull
    • Remote initiated
    • Control protocol
    • Login(?)
    • Job Submission
    • Job Management
    • 1284.1
    • Device Attributes
    • Feature set
    • Data formats
    • Raw24
    • Postscript
    • PCL
    • etc.
  • Security and authentication
    • Login: This is whatever the process is that allows a device to initiate communication or submit a job. A session initialization. OR… is this the connection to the next layer 'above' your layer. Granting of access to this layer of communication.
  • From January 26th, 1998 Meeting in Maui, HI
    • The requirements list includes the term "symmetrical connections."
      • Defined as "either endpoint can open or close a connection." Confirmed that this means either side (any node) can initiate a transport-level connection.
  • From March 2nd, 1998 Meeting in Austin, TX
    • Symmetry
      • There seems to be the necessity to provide a better understanding of what is meant by "Symmetry". The service can be connected from any host and has no knowledge of the potential client space. While a server is waiting for a connection it uses minimal resources on it's host. If a target is a service then there is no problem. This is going to require more discussion. This topic will be sent to the reflector.
  • From May 18th, 1998 Meeting in Crystal City, VA
    • Functions: capabilities like printing, faxing, modem communication;
      • "Units": one-to-one mapping with device drivers
    • Service: "A capability provided by one protocol layer entity for use by a higher layer or by a management entity."
      • "Logical Unit": "A subcomponent of a unit that operates on a set of tasks independently of other services, and is accessed by a the device driver software."
Back to the Top of the Page
Back to the Issues Page / Back to the Top of the Page

Go To 1394 Home


If you would like to contribute any input, contact Greg LeClair.
Last modified May.18.1999.