Re: State Machine

From: Jonathan NAYLOR (terhi.victor@logonet.com)
Date: Tue Feb 05 2002 - 11:51:48 EET

  • Next message: Arno Verhoeven: "Where is source of dhlog-0.1.4 ?"

    Hi Andy

    As the person responsible for the Linux AX.25 implementation (1995 to
    1998) I can tell you that the original state machine in the AX.25 v2.0
    spec is way too complex. This was spotted by a number of guys in the US
    (including Phil Karn KA9Q) and a simplified version was presented in
    one of the Computer Networking Conference proceedings. They used 4
    states and a number of flags. The five states chosen were:

    0. Disconnected (not really a state in the Linux implementation)
    1. Connecting
    2. Disconnecting
    3. Information Transfer
    4. Timer Recovery (from 3)

    If AX.25 v2.0 did not use <RR> frames to poll the other end when an
    information frame retry was occuring, then we could remove state 4 above,
    this was done in NET/ROM which simply repeats the information frame, and
    for ROSE also.

    This makes a lot of sense as it captures the essence of what is happening,
    it allows for a much easier visualisation of the state machine. It is
    probable that this state machine will change in the future as different
    people now maintain the AX.25 code.

    Jonathan HB9DRD/G4KLX
    -
    To unsubscribe from this list: send the line "unsubscribe linux-hams" in
    the body of a message to terhi.victor@logonet.com
    More majordomo info at http://vger.kernel.org/majordomo-info.html



    This archive was generated by hypermail 2b30 : Tue Feb 05 2002 - 12:14:30 EET