tcptrace-bugs [ Re: TSN as double instead of unsigned...]

From: Wesley Eddy (
Date: 12/06/02

  • Next message: Julie Scott: "tcptrace-bugs"

    Date: Fri, 6 Dec 2002 12:51:04 -0500
    From: Wesley Eddy <>
    Subject: tcptrace-bugs [ Re: TSN as double instead of unsigned...]
    Message-ID: <>

    ----- Forwarded message from -----

    Date: Fri, 6 Dec 2002 18:08:52 +0100 (CET)
    Subject: Re: TSN as double instead of unsigned...
    To: wesley Eddy <>
    X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by id gB6H90R19805
    X-Spam-Status: No, hits=-7.1 required=5.0


     I write to you since I do not want to make some noise in the
    tcptrace-bugs archive. I thought about using tcptrace-maintainers, but
    your web site says its purpose is not to discuss about tcptrace
    technical details. Please forward this message to the appropriate
    destination, thanks in advance.

    On 5 Déc, Pour: wrote:
    > tcptrace/plotter.c
    > /* Header for all other plots */
    > /* graph coordinates... */
    > /* X coord is timeval unless graph_time_zero is true */
    > /* Y is signed except when it's a sequence number */
    > /* ugly hack -- unsigned makes the graphs hard to work with and is
    > only needed for the time sequence graphs */
    > /* suggestion by Michele Clark at UNC - make them double instead */
    > Mfprintf(f,"%s %s\n",
    > graph_time_zero?"dtime":"timeval",
    > ((strcmp(ppi->ylabel,"sequence number") == 0)&&(!graph_seq_zero))?
    > "double":"signed");
    > Hi,
    > If I write a "uint64.c" new type of "coord_type" for xplot, would you
    > consider including the code in your xplot releases ?
    > Cheers,

    When I wrote the message above, I thought tcptrace was unwrapping the
    32bit TSN, that's why it was advertising it as a double. Later I
    realized that it was not unwrapped. So the question is : why unsigned
    does make the graphs "hard to work with" ?

    Then there is this wrapping TSN problem. I see two options: move it to
    64bits both in tcptrace and xplot, with unwrapping in tcptrace, or
    another, maybe lighter solution, would be to shift the Y axis in xplot,
    in order to put 2^32 in the middle of the Y axis when needed, instead of

    Thanks for feedback.


    PS: would a read-only access to your CVS server be foreseeable?

    Marc Herbert
    Ph.D intern
    INRIA / Sun Labs Europe

    ----- End forwarded message -----

    This archive was generated by hypermail 2b30 : 12/07/02 EST