MathGroup Archive 2008

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

Search the Archive

Re: "The requested evaluator is not currently defined"

  • To: mathgroup at smc.vnet.net
  • Subject: [mg94645] Re: "The requested evaluator is not currently defined"
  • From: David Reiss <dbreiss at gmail.com>
  • Date: Thu, 18 Dec 2008 07:24:04 -0500 (EST)
  • References: <200812141240.HAA10363@smc.vnet.net> <gi5jsa$ptq$1@smc.vnet.net>

There are a number of places in the Mathematica system where the name
of the Kernel (the Evaluator Option for a notebook, for example: see
below for a particular case) is hard wired to the name "Local".
Wherever this occurs, and if you have renamed your Kernel to something
other than "Local" you will get the "The requested evaluator is not
currently defined" error message.  So perhaps this is what you folks
are observing.  I have come across this in a number of situations
since I have 4 or 5 versions of Mathematica on my system for testing
purposes.  Each Kernel is named differently.  (I am using OS X, but
also use Windows XP for development and testing through Parallels on
my Mac.)

An example of where you will get the "The requested evaluator is not
currently defined" message for exactly this reason is in the
Documentation Center.  The Documentation Center notebooks all use the
Wolfram Reference.nb Stylesheet  This is not a publicly accessible
style sheet via the Format>Stylesheet menu.  However if you look
inside the Mathematica distribution and locate that Stylesheet and
take a look at the StyleData at the Notebook level you will see that
it has

Evaluator->"Local"

explicitly set (be careful not to break this Stylesheet: it's best if
you play with it as a copy).  If you are using a Kernel renamed to
anything other than "Local", then the Documentation Center will not
work properly.

I believe that there are other places in the system where behavior is
hard wired to have Evaluator->"Local", but I can't remember them at
the moment.  Several of these I have submitted as bugs to WRI in the
past.   (I believe that one ohter place that it occurs in has to do
with the SldeShow enviornment...).

I suspect that there needs to be a design change somewhere so that,
for example, when the Evaluator needs to have a particular version of
Mathematica for some reason (e.g., 7 rather than 6) it checks the
Kernel's version number, rather than just its name.  For example,
rather than an Evaluator option, there might be a
EvaluatorMinimumVersionNumber Option for notebooks and Cells that
takes on a numerical value.

--David





On Dec 17, 6:35 am, Michael Weyrauch <michael.weyra... at gmx.de> wrote:
> Well, I had that error message, when I had set the kernel for a
> particular notebook to a specific kernel, then copied that notebook to
> a machine, where this kernel was not in the list of available kernels,
> and tried to evaluate an expression in that notebook.
>
> This may have happend accidentally and inadvertantly in your case. So, I
> suggest, you explicitly set "Notebook's kernel" (using the menu) to
> "Local" and then evaluate. This should work and did in my case.
>
> Michael
>
> Murray Eisenberg schrieb:
>
> > Some additional information: This is under Windows XP (SP3).  And each
> > version of Mathematica has its own mathpass file.
>
> > Murray Eisenberg wrote:
> >> I just tried to evaluate an expression in Mathematica 5.2 and got an
> >> error message "The requested evaluator is not currently defined."
>
> >> The problem likely arises from Mathematica 6.0.3 and Mathematica 7.0.0
> >> both being installed (and they are working OK).  But the preceding
> >> happens even with both of them closed and after I open 5.2 with a clean
> >> cache.



  • Prev by Date: Re: Mathematica 7 weirdness
  • Next by Date: Re: NIntegrate Problem
  • Previous by thread: Re: Re: "The requested evaluator is not currently defined"
  • Next by thread: Re: "The requested evaluator is not currently defined"