Re: When random isn't random

From: John Herbster \(TeamB\) ("John)
Date: 12/14/03

  • Next message: John Remy: "Re: DLL with string params callable via D4 and VB" 
    Date: Sat, 13 Dec 2003 19:25:11 -0600

    "Dr John Stockton" <spam@merlyn.demon.co.uk> wrote

    > >Would you be kind enough to add your very good
    > >comments and suggestions to the report #6619 in
    > >Quality Central where they will not be forgotten.

    > I will not. I see no need to spend my connect time and money in
    > repeating information which is already held at Borland. Feel free to
    do
    > so yourself, however, if you wish; but it should not be necessary.

    Dr. John, Thanks! With your permission I will post the
    following comments to Quality Central Report #6619
    which is titled "Expand the Random function".
    Regards, JohnH

    --------------------
    The following are suggestions by John Stockton, Surrey, UK
    posted by John Herbster, with permission:

    IMHO, the Random and Randomize functions should not be changed, for
    compatibility reasons.

    A new Unit should instead be provided, containing thoughtfully- named
    and -written routines including versions of Random, Deal, and Shuffle
    (and, if there is not one already, a Sorting Unit). The task could be
    subcontracted to skilled volunteer labour, such as is found here.

    The documentation for the existing routines should be changed to say
    that they are OK for most normal uses but that users should look at the
    new stuff before deciding.

    New randomisers should be capable of initialising their seeds to any
    value within range, at least if called a small but arbitrary time after
    the start of a program which was started at an arbitrary time of day on
    a Windows booted an arbitrary time ago.

    If a reasonably efficient reverse algorithm is known, it should be
    provided; it is necessarily an alternative, and reversibility might be
    of use.

    The new routines should not use a specific RandSeed; instead, each
    should have a first parameter var seed : <type> .

    For completeness, the unit should contain the old algorithms but with
    the new parameter.
    --------------------------

  • Next message: John Remy: "Re: DLL with string params callable via D4 and VB"

    Relevant Pages

    • Re: FOCUS application, written in?
      ... John "REBOOT" Brandon ... It's fairly common practice in the OpenVMS world to use routines ... I have programs that use a routine in the BASIC RTL that contain ...
      (comp.os.vms)
    • Re: DESCryptoServiceProvider, CryptoStream and InitialisationVector
      ... > RandomNumberGenerator class. ... >> I have written a couple of routines which encrypt and decrypt data ... >> encrypt, the key and the IV to the routine. ... >> John Young ...
      (microsoft.public.dotnet.framework)
    • Latest John Herbster Rounding Routines
      ... Does anyone here know the latest version of John Herbster's rounding ... routines and where to find it? ... Jeff ... Prev by Date: ...
      (borland.public.delphi.language.basm)
    • Re: Fastcode Challenge Status 2005
      ... I ask because I have improved routines for many of the 2003/2004 challenges ... John ... The Fastcode Project: ... Prev by Date: ...
      (borland.public.delphi.language.basm)