Services & ResourcesWolfram ForumsMathGroup Archive

[Date Index] [Thread Index] [Author Index]

Kernel init.m File

• To: mathgroup at smc.vnet.net
• Subject: [mg47975] Kernel init.m File
• From: Harold.Noffke at wpafb.af.mil (Harold Noffke)
• Date: Tue, 4 May 2004 01:08:55 -0400 (EDT)
• Sender: owner-wri-mathgroup at wolfram.com

$Version "5.0 for Microsoft Windows [2000] (November 18, 2003)"

MathGroup:

In the ...\5.0\Configuration\Kernel\init.m file, I find the following
template ...


    (** User Mathematica initialization file **)



    (** Decide how to display graphics on this machine **)

    Begin["System`Private`"]
    (* Hide any symbols which are created *)

    Which[
        $Linked || $ParentLink =!= Null,        <<"PSDirect.m",
        Environment["DISPLAY"] =!= $Failed,     <<"Motif.m",
        True,                                   <<"Terminal.m"
         ]

    End[]


It seems to be Wolfram's intent that all user init.m statements
(functions, blocks, modules, packages, etc.) occure BEFORE the
Begin["System`Private`"] statement.  However, I'm confused over
whether user init.m statements should be confined to
...\Kernel\init.m, whether any damage would be caused by placing them
after the System`Private statements, or whether the user is permitted
to create his own init.m file in the $InstallationDirectory (the
...\5.0\ directory).

I've read posts from other users who say they have safely used an
init.m in their 5.0 directory, but it is not clear from $Path that
5.0\init.m will be loaded at startup, because $InstallationDirectory
does not appear there.

If 5.0\ were added to $Path, it would load AFTER Kernel\init.m.  Would
this put 5.0\init.m into the scope of Kernel\init.m's System`Private
statement, and produce umpredicted results?

Please help me clarify where the user's init.m statements can be
placed without causing damage or unpredicted results.

Regards,
Harold

• Follow-Ups:
  • Re: Kernel init.m File
    • From: Omega Consulting <info@omegaconsultinggroup.com>