MathGroup Archive 2002

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

Search the Archive

Re: RE: Corrupted Files & Version 4.2 Problems

  • To: mathgroup at
  • Subject: [mg35292] Re: [mg35263] RE: Corrupted Files & Version 4.2 Problems
  • From: Andrzej Kozlowski <andrzej at>
  • Date: Sat, 6 Jul 2002 05:45:31 -0400 (EDT)
  • Sender: owner-wri-mathgroup at

You can not only run a different kernel but in fact two kernels at the 
same time, though I am not quite sure  if the usual license  allows 
that. You do it as follows (I am describing how it works under MacOS X, 
but it ought to be the same for other OS's).  Choose "Kernel 
Configuration Options" from the Kernel menu. Click on the Add button. 
Name your new kernel. Make sure that Launch On Local machine is chosen 
(it is the default). Click on the Advanced Options tab. In the field 
"Arguments to MLOpen" insert the proper path to your new kernel. Now if 
you go again to the Kernel menu you will find that under Start Kernel, 
and Notebook Kernel items you have two kernels. Create a new notebook. 
 From the Notebook Kernel menu choose the new kernel. If you did not make 
a mistake in inserting the full path to the kernel in "Arguments to 
MLOpen" all computation in this notebook will be performed by the new 
kernel while those in other notebooks by the default kernel. Of course 
you need enough RAM for this to work. Other than that, you can just 
quite the default kernel first.

I have never really found out whether the owners of single licenses are 
licensed to run their own old kernels together with new ones. It is 
actually a very reasonable thing to do when new kernels (or new 
FrontEnds) contain bugs not present in previous versions.

Andrzej Kozlowski
Toyama International University

On Saturday, July 6, 2002, at 11:06  AM, DrBob wrote:

> I have no idea how to run the 4.1 FrontEnd with the 4.2 kernel, but I do
> have them both, as I just upgraded from one to the other.  I haven't
> lost a file since Tuesday, but that's meaningless.
> I've also had NO RESPONSE from WRI on this.  Don't believe whatever they
> tell you about "Premier" service.
> Bobby
> -----Original Message-----
> From: Andrzej Kozlowski [mailto:andrzej at]
To: mathgroup at
> Sent: Friday, July 05, 2002 4:35 AM
> Cc: mathgroup at
> Subject: [mg35292] Re: [mg35263] RE: Corrupted Files & Version 4.2 Problems
> It would be interesting to know if the corruption problem can be avoided
> by running the 4.1 FrontEnd with the 4.2 kernel. Since as far as I can
> tell all the interesting enhancements are in the kernel that would be a
> fairly reasonable work around, if it works of course. Can anyone who has
> got both versions test it?

> On Friday, July 5, 2002, at 03:19  PM, David Park wrote:
>> Bobby,
>> Thanks for the information. That is now two people who have complained
>> about
>> 4.2 on MathGroup. It doesn't look good.
>> I am certainly disappointed that the Graphics3D rendering errors are
> not
>> fixed in Version 4.2. Thank you for testing them for me. (These
> concern
>> loss
>> or transfer of Line graphics directives when there are both Lines and
>> Polygons in the Graphics3D.) I complained to WRI about it a long time
>> ago
>> and sent them test notebooks. I also complained on MathGroup and had
>> other
>> people verify the rendering errors. I seem to recall that WRI said it
>> would
>> be fixed in the next version.
>> David Park
>> djmp at
>>> Danger, Will Robinson, danger, danger!
>>> Don't start depending on version 4.2 yet -- it corrupts files.
> Version
>>> 4.0 was bad about that, but not THIS bad.
>>> Version 4.1 was almost entirely innocent of that particular problem.
>>> I'll be uninstalling this thing immediately.
>>> Bobby
>>> Dr. Bobby Treat
>>> Survey Sampling:
>>> OR Models:
>>> Austin INFORMS:

  • Prev by Date: RE: How to suppress the mesh in ParametricPlot3D ?
  • Next by Date: Re: How to suppress the mesh in ParametricPlot3D ?
  • Previous by thread: RE: RE: Corrupted Files & Version 4.2 Problems
  • Next by thread: Re: RE: Corrupted Files & Version 4.2 Problems